Integrated DM Care - Multidisciplinary Approach

Integrated
Diabetes Care
A Multidisciplinary
Approach
David Simmons
Helmut Wenzel
Janice C. Zgibor
Editors
123
Integrated Diabetes Care
David Simmons
Helmut Wenzel Janice C. Zgibor
Editors
Integrated Diabetes
Care
A Multidisciplinary Approach
Editors
David Simmons Helmut Wenzel
School of Medicine Health Economist
Western Sydney University Konstanz, Germany
Sydney, NSW, Australia
Janice C. Zgibor
Department of Epidemiology
and Biostatistics, College
of Public Health
University of South Florida
Tampa, FL, USA
ISBN 978-3-319-13388-1 ISBN 978-3-319-13389-8 (eBook)

DOI 10.1007/978-3-319-13389-8
Library of Congress Control Number: 2016955045
Springer International Publishing Switzerland 2017

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or
part of the material is concerned, specifically the rights of translation, reprinting, reuse of
illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way,
and transmission or information storage and retrieval, electronic adaptation, computer software,
or by similar or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are
exempt from the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in
this book are believed to be true and accurate at the date of publication. Neither the publisher nor
the authors or the editors give a warranty, express or implied, with respect to the material
contained herein or for any errors or omissions that may have been made.
Printed on acid-free paper
This Springer imprint is published by Springer Nature

The registered company is Springer International Publishing AG Switzerland
Foreword
If one wished to choose a health condition that requires an integrated multi-

disciplinary approach to management, then you cannot go past diabetes. And
given the spectacular growth of diabetes over the last 20 years, the number of
people with diabetes worldwide has more than doubled. The International
Diabetes Federation (IDF) today estimates there are at least 415 million peo-
ple with diabetes. The IDF also estimates there will be over 620 million with
diabetes by 2040. Concerning features in relation to the current epidemic are
the increases in both type 1 and 2 diabetes, gestational diabetes, and the emer-
gence of type 2 diabetes in children and adolescents. This scenario poses
huge social, public health, and economic problems to most nations and
stretches their capacity for optimal diabetes care.
The very nature of diabeteswith its issues relating to day-to-day man-
agement and the risk and burden of future complications such as cardiovascu-
lar disease, retinopathy, and kidney disease and comorbidities which include
liver disease and obstructive sleep apnoearaises huge issues for an inte-
grated approach to management.
People with diabetes often have multiple comorbidities and see a number
of different health professionals across primary, community, and specialist
care services. This is where integrated care may have an important game-
changing role. It has been demonstrated that best-practice, high-quality dia-
betes care can only be achieved when healthcare professionals work
seamlessly. This involves partnerships across primary healthcare, community
care, and specialist care services delivering integrated diabetes care to the
consumer, the person with diabetes.
Establishing an effective, integrated system will require a transformation
in the way care is delivered, making it more consumer focused and team
based. The person with diabetes may be looked after in primary, secondary,
and also tertiary care. Their healthcare team includes the general practitioner,
the diabetologist, diabetes educator, the dietician, and other healthcare pro-
viders. There is a requirement for coordination across the relevant agencies,
encompassing the whole diabetes care pathway. Consumer engagement, self-
management, and empowerment will be major contributing factors in achiev-
ing this goal.
Unless we embrace this approach to management of diabetes, health sys-
tems in both developed and developing countries will be swamped by the
numbers of people with diabetes. This will be associated with the increased
v
vi Foreword
direct cost of diabetes, and it also has indirect costs nationally in terms of
premature morbidity and impacts on the workforce.
So this brings us to this timely book Integrated Diabetes Care: A
Multidisciplinary Approach. It addresses integrated care and also the many
barriers for improving diabetes care across the globe. So, what do we really
need to understand about integrated care and how to overcome these hur-
dles? This excellent book edited by David Simmons, Helmut Wenzel, and
Janice C. Zgibor gives comprehensive coverage of these issues and provides
examples of approaches that could improve care while reducing costs.
Integrated Diabetes Care: A Multidisciplinary Approach collates worldwide
evidence of how integrated care works both across disciplines and across
organisations to improve diabetes care.
The integrated approach prioritises the needs of the individual and recog-
nises the many interactions required between the person with diabetes, the
range of health professionals needed for their care, and the various levels of
the health system. Diabetes is for life, and this book provides guidance for all
those involved in diabetes to bring seamless and optimal care for the person
with diabetes.
Professor Paul Zimmet, AO MD PhD FRACP FRCP (London) FTSE

Honoris Causa Doctoris (Complutense University, Madrid)
Doctor of Laws Honoris Causa (Monash University, Melbourne)
Doctor Philosophiae Honoris Causa (Tel Aviv University, Tel Aviv)
Professor, Department of Medicine, Monash University
Honorary President of International Diabetes Federation
Formerly Director, International Diabetes Institute, and Director Emeritus,
Baker IDI Heart
Preface
Why a book on integrated diabetes care? Over the last 2030 years, there
have been a plethora of projects and policies putatively designed to bring
together all the different health workers and health services for defined groups
of people with diabetes. Some have sat behind grandiose broader integrated
care initiatives. Others have sat within a single health service. Over this time,
there have been enormous improvements in the way we can manage type 1
and type 2 diabetes. There has been the growing recognition of the impor-
tance of personalised medicine including the ability to diagnose rare forms of
diabetes (such as monogenic diabetes). Behind this diabetes clinical evolu-
tion has been, perhaps, an even greater revolution in the work behind the
scenes, especially in the way we handle health data and clinical governance,
and in our understanding that there is a chasm behind what we can do and
what is actually happening. Why is care not as good as we know it can be?
Why do avoidable complications still happen? It is clear that there is much
more that can be done to facilitate and enable those with diabetes: right care,
right time, and right place.
This book came about to provide greater depth than possible in academic
publications on what worked and what did not from the clinicians and devel-
opers points of view. This in turn can inform future developers, managers,
and clinicians on how best to structure their next attempt to move towards a
more united and seamless approach to the way that those with diabetes
receive their care.
Sydney, Australia David Simmons

Konstanz, Germany Helmut Wenzel
Tampa, FL, USA Janice C. Zgibor
vii
Contents
1 An Introduction to Integrated Care and Diabetes

Integrated Care ........................................................................... 1
Helmut Wenzel and David Simmons
2 Integrating Outpatient Care the Toyota Way: An
Individualized Multidisciplinary Team-Care Model
for Diabetes Care Delivery ......................................................... 11
R. Harsha Rao and Peter Perreiah
3 Approaches to Integrated Diabetes Care: United States:
San Francisco .............................................................................. 31
David H. Thom and Thomas Bodenheimer
4 A Primary Health-Care System Approach to Improving
Quality of Care and Outcomes in People with Diabetes:
The University of Pittsburgh Medical Center Experience ...... 51
Janice C. Zgibor, Francis X. Solano Jr., and Linda Siminerio
5 Integrated Diabetes Care in Hong Kong: From
Research to Practice to Policy .................................................... 65
Roseanne O. Yeung, Junmei Yin, and Juliana C.N. Chan
6 Approaches to Integrated Diabetes Care: A South
African Approach ........................................................................ 87
Larry A. Distiller and Michael A.J. Brown
7 English Approaches to Integrated Diabetes Care:
The East Cambridgeshire and Fenland Diabetes
Integrated Care Initiative: A Multiple Provider
Approach...................................................................................... 107
David Simmons, Dahai Yu, and Helmut Wenzel
8 UK Approaches to Integrated Diabetes Care:
DerbyA Joint Venture Model Under the NHS ...................... 131
Paromita King
9 Integrated Diabetes Care: Coventry and Warwickshire
Approach...................................................................................... 147
Ponnusamy Saravanan, Vinod Patel, Joseph Paul OHare,
and Sudhesh Kumar
ix
x Contents
10 Integrated Diabetes Care in Germany: Triple Aim

in Gesundes Kinzigtal ................................................................. 169
Caroline Lang, Elisa A.M. Kern, Timo Schulte,
and Helmut Hildebrandt
11 Approaches to Integrated Diabetes Care in the
Netherlands.................................................................................. 185
Harold W. de Valk and Helmut Wenzel
12 Integrated Diabetes Care in Sweden ......................................... 201
Helmut Wenzel, Stefan Jansson, and Mona Landin-Olsson
13 Integrated Diabetes Care for Adults with Diabetes:
A Patient Organisation Perspective ........................................... 215
Heather Bird and Bridget Turner
14 Training for Diabetes Integrated Care:
A Diabetes Specialist Physician Perspective
from the English NHS ................................................................. 227
Anne Dornhorst
15 Diabetes Integrated Care: Are We There Yet? ......................... 233
David Simmons, Helmut Wenzel, and Janice C. Zgibor
Index ..................................................................................................... 249

Contributors
Heather Bird Diabetes UK, London, UK

Thomas Bodenheimer Department of Family and Community Medicine,
University of California, San Francisco School of Medicine and San Francisco
General Hospital, San Francisco, CA, USA
Michael A.J. Brown Centre for Diabetes and Endocrinology (Pty) Ltd,
Johannesburg, South Africa
Juliana C.N. Chan Department of Medicine and Therapeutics, The Chinese
University of Hong Kong, Hong Kong, China
Hong Kong Institute of Diabetes and Obesity, The Chinese University of
Hong Kong, Hong Kong, China
International Diabetes Centre of Education, The Chinese University of Hong
Kong, Prince of Wales Hospital, Satin, Hong Kong, China
Harold W. de Valk Internist-endocrinologist, Department of Internal
Medicine, University Medical Center Utrecht, Utrecht, The Netherlands
Larry A. Distiller Centre for Diabetes and Endocrinology (Pty) Ltd,
Johannesburg, South Africa
Anne Dornhorst Department of Diabetes and Endocrinology, Imperial
College Healthcare NHS Trust, Hammersmith Hospital, London, UK
Helmut Hildebrandt OptiMedis AG, Hamburg, Germany
Stefan Jansson Department of Family Medicine, Brickebacken Primary
Health Care Center, rebro, Sweden
Elisa A.M. Kern Gesundes Kinzigtal GmbH, Haslach, Germany
Paromita King Department of Diabetes and Endocrinology, Medical
Specialties, Derby Teaching Hospitals NHS Foundation Trust, Royal Derby
Hospital, Derby, UK
Sudhesh Kumar Warwick Medical School, The University of Warwick,
Coventry, UK
University Hospitals Coventry and Warwickshire NHS Trust, Coventry, UK
xi
xii Contributors
Mona Landin-Olsson Department of Endocrinology, Lund University

Hospital, Lund, Sweden
Caroline Lang Department of Internal Medicine, Prevention and Care of
Diabetes, University Hospital Carl Gustav Carus at the Technische Universitt,
Dresden, Dresden, Germany
Joseph Paul OHare Warwick Medical School, The University of Warwick,
Coventry, UK
Vinod Patel George Eliot Hospital NHS Trust, Coventry, UK
Warwick Medical School, The University of Warwick, Coventry, UK
Peter Perreiah Sapience Technologies, Alpharetta, GA, USA
R. Harsha Rao VAPHS-UD, University Drive, Pittsburgh, PA, USA
Ponnusamy Saravanan George Eliot Hospital NHS Trust, Coventry, UK
Warwick Medical School, The University of Warwick, Coventry, UK
Timo Schulte OptiMedis AG, Hamburg, Germany
Linda Siminerio Division of Endocrinology, School of Medicine, University
of Pittsburgh, Pittsburgh, PA, USA
David Simmons School of Medicine, Western Sydney University, Sydney,
NSW, Australia
Francis X. Solano Community Medicine Inc, University of Pittsburgh
Medical Center, Pittsburgh, PA, USA
Physician Services Division, University of Pittsburgh Medical Center,
Pittsburgh, PA, USA
Division of General Internal Medicine, School of Medicine, University of
Pittsburgh, Pittsburgh, PA, USA
David H. Thom Department of Family and Community Medicine,
University of California, San Francisco School of Medicine and San Francisco
General Hospital, San Francisco, CA, USA
Bridget Turner Diabetes UK, London, UK
Helmut Wenzel Health Economist, Konstanz, Germany
Contributors xiii
Roseanne O. Yeung Department of Medicine, Division of Endocrinology,

University of Alberta, Edmonton, Alberta, Canada
Hong Kong Institute of Diabetes and Obesity, The Chinese University of
Hong Kong, Hong Kong, China
Junmei Yin Department of Medicine and Therapeutics, The Chinese
University of Hong Kong, Hong Kong, China
Dahai Yu Research Institute for Primary Care and Health Sciences, Keele
University, Newcastle Under Lyme, UK
Janice C. Zgibor Department of Epidemiology and Biostatistics, College of
Public Health, University of South Florida, Tampa, FL, USA
An Introduction to Integrated Care
and Diabetes Integrated Care 1
Helmut Wenzel and David Simmons
Background delivery is now fragmented [2]. Such fragmenta-

tion is a manifestation of organisational and finan-
Modern health care is a wonderful thing. We have cial barriers, which divide providers at the
medications, investigations, procedures, equip- boundaries of primary and secondary care, physi-
ment and health-care approaches that can do more cal and mental health care, and between health
than ever before. Half of the years of life gained and social care. Diverse specific organisational
19501995 were due to medical [1] intervention, and professional cultures, and differences in terms
and benefits have continued further over the last of governance and accountability also contribute
two decades. However, the knowledge and exper- to this fragmentation [2].
tise required to provide these interventions is so Evaluation demonstrates suboptimal care pro-
vast, that it requires different degrees of speciali- cesses resulting in both poorer health and finan-
sation, increasingly grouped into a range of ser- cial outcomes. Many of these deficiencies are
vices. Efficiencies have been introduced through caused by organisational problems (barriers, silo
the centralisation of the more expensive equip- thinking, accountability for budgets) and are
ment and expertise into dedicated buildings often to the detriment of all of those involved:
including hospitals. Access to such facilities has patients, providers and funders in extreme
been limited through a range of mechanisms cases leading to lose-lose-lose-situations [3].
including out of pocket expense, distance, patient Patients observe missing coordination, ran-
knowledge, capacity/waiting times and a variety dom care, chaotic care or even negative
of gatekeeper functions (e.g., primary care in the coordination and state:
UK NHS). The way such services have devel- We are sick of falling through gaps. We are tired of
oped, and the mechanisms for health-care access, organisational barriers and boundaries that delay
has varied between nations, but it is clear that in or prevent our access to care. We do not accept
general, wherever you are in the world, service being discharged from a service into a void. We
want services to be seamless and care to be con-
tinuous. [4, p. 11]
Negative coordination is characterised by

H. Wenzel (*) participants caring solely for their own interests.
Health Economist, Konstanz, Germany They do not see the common goal and they do not
e-mail: [email protected] work jointly for an optimal solution. Negative
D. Simmons coordination can result from specialisation and a
School of Medicine, Western Sydney University, lack of orientation across the broader population,
Sydney, NSW, Australia
e-mail: [email protected] its health and social care and to some extent
Springer International Publishing Switzerland 2017 1

D. Simmons et al. (eds.), Integrated Diabetes Care, DOI 10.1007/978-3-319-13389-8_1
2 H. Wenzel and D. Simmons
reflects a problem with the division of labour. economic or financial issue. Such a large number
One important aspect, particularly within a com- of people needing complex care calls for appro-
petitive health economy, that must not be under- priate delivery systems that bring together pro-
estimated, is an unfavourable organisational fessionals and skills from both social and
culture [5]. Negative Coordination only allows health-care sectors. There is some evidence that
policy changes that are pareto-superior to the sta- integrated care does improve the quality of
tus quo, i.e., there will be no losers. If disadvan- patient care and leads to improved health or
tages occur, the existence of negative coordination patient satisfaction [10, 11], but evidence of eco-
prevents new policies as long as there are still nomic benefits remain an issue for further
parties who are better off without the changes [6]. research [10].
Some call it coordination by avoidance, a kind Failure to improve integration and coordina-
of self-protection where conflicting rules and tion of services along a care continuum can
excessive regulation exists [7]. Negative coordi- result in suboptimal outcomes (health and cost),
nation exists in vertical and horizontal structures such as potentially preventable hospitalisation,
and it applies to both civil society and govern- avoidable death, medication errors and adverse
mental units [8]. drug events [3, 12, 13]. Clearly, if our health sys-
Politicians and governments criticise the wid- tems are currently fragmented, and, as a result,
ening gap between expenditure and available are not delivering the best care using the resources
financial resources. The discussions centre on that are available, care processes and the associ-
expenditure drivers like the ageing population ated organisational principles of care need to be
and medical progress. Governments try to close reappraised and restructured. A broad spectrum
the gap between rapidly increasing demand and of instruments and organisational options are
slower growth in financing opportunities, by now available for application to facilitate care
applying four classical administrative measures: provision [14] in a more integrated way, but first-
what is integrated care?
1. cutting down expenditure (various budgets),
2. excluding services from being reimbursed (a
problem for patients), thus increasing the size What Is Integrated Care?
of the patients contribution, i.e., out-of-
pocket payments, Integrated care may be described best by
3. by raising the contribution fees from the involves the provision of seamless, effective and
insured, and efficient care that responds to all of a persons
4. last but not least, by bargaining with providers health needs, across physical, social and mental
and forcing the setting of fixed prices. health, in partnership with the individual, their
carers and family [15, p. 1]. Or, like the WHO
Of course, there is some variation, depending put it another way: The organization and man-
on the health-care model tax financed vs contri- agement of health services so that people get the
bution fee financed. Rationing of services is also care they need, when they need it, in ways that
an option. This situation has been aggravated by are user friendly, achieve the desired results and
the recent so-called global financial crisis. provide value for money[10, p. 5].
However, it is unlikely that there will be a Nolte et al. [10] point out that integrated
change for the better in the budgeting and care is often used in the context with concepts or
global economic situations in the near future. terms like case management, care coordina-
From an economic viewpoint the efficient use of tion, collaborative care or combinations of
resources has to be a major focus for any future these. However, Integrated health services
health system changes [9]. means different things to different people [16].
Of course, the rising burden of chronic disease From the patients perspective it is primarily
is about the lives of millions of people, not just an continuity of care. In a survey with 254 health
1 An Introduction to Integrated Care and Diabetes Integrated Care 3
expert and 670 patients Juhnke et al. [17] anal- the highest degree of working together.
ysed the needs and expectations with regard to Coordination is somewhat in between and stands
patient-centred health care (coordinated care). for partial cooperation [22]. Nolte et al. [10]
Coordinated care is made up by seven dimen- report on a more extended analytical grid.
sions: access, data and information, service and Looking closer at possible attributes of the term
infrastructure, professional care, interpersonal integration, three dimensions have been anal-
care, individualised care, continuity and coordi- ysed: the type, the breadth (horizontal, vertical
nation. For diabetic patients continuity of care integration), the degree (from full integration to
is a comprehensive and complex concept. The collaboration.), and the process of integration.
abstract notion comprises five components: For further categorising of the integrated diabetes
access to services, interactions with physician, care projects described in this book, we will
and interactions with other health-care providers, apply this typology.
personal self-responsibility, and communication. The dimension types of integration can be
This concept exceeds what is traditionally associ- subdivided even further into four categories:
ated with continuity of care and attaches particu- Functional, organisational, professional, and
lar importance to personal self-responsibility clinical [23]. Functional integration means that
[18], possibly through patient education and self- support functions, such as financial management,
management approaches. For example, Haggerty information management, strategic planning and
et al. [16] identify three types of continuity: human resource management are the target area.
information continuity, management continuity Organisational integration has to do with the
and relation continuity. creation of new entities, e.g., mergers or joint
From the point of view of providers provider ventures, either by the creation of a new organisa-
connectedness has to be added to Haggertys tion or by absorption by one of the others, con-
concept. Providers felt that the communication tracting issues, and creation of networks (aiming
was more effective, and it was easier to provide at the integration of planning and delivery of ser-
continuity of care, when they already had rela- vices). Professional integration deals with stra-
tionships with other providers [19]. tegic alliances of health-care professionals within
Operationalisation and implementation of conti- or between institutions or contracting. Clinical
nuity of care solutions should therefore consider integration deals with the level to which patient
all perspectives, exploring how these come care services are coordinated across the various
together to enhance the patient-centredness of personnel, functions, activities and operating
care [20]. units of a system [23, p. 71]. With breadth of
integration a further distinction is made between
integration on the same level, e.g., general prac-
Typology of Integrated Care tices and community care (horizontal integra-
Approaches tion), and integration at different levels, e.g.,
primary care, secondary care and tertiary care
There have been many proposals on how to struc- (vertical integration) [10, 22].
ture the various concepts of integrated care. Integrated care is often described as a contin-
Schrijvers et al. [21] report on classifications uum [10, 24], actually depicting the degree of
according to target groups or field of integrated integration. This degree can range from linkage,
care. However, most of the integrated care litera- to coordination and integration [10], or segrega-
ture distinguishes between different ways and tion (absence of any cooperation) to full integra-
grades of working together, depicting a cooperation [25], in which the integrated organisation is
tion scale, which is based on three attributes like responsible for the full continuum of care, includ-
autonomy, coordination and integration. ing funding, to collaboration, which describes
Autonomy represents the one end of the scale separate structures in which organisations retain
with least cooperation, integration the end with their own service responsibility and funding
criteria [10]. Following Nolte et al. [10] link- ments were determined that can characterise such
age works on the basis of the existing separate fully integrated delivery systems. This includes
structures of health and social services systems. a population defined by enrolment, contractual
See Fig. 1.1. The corresponding organisations responsibility for a defined package of compre-
preserve their own service responsibilities, way hensive healthcare services, financing on the
of funding and operating rules. In the case of basis of pooling multiple funding streams, a
coordination, additional structures and pro- closed network (i.e., a selected group of con-
cesses are added, such as routinely shared infor- tracted and/or salaried providers), emphasis on
mation and discharge planning. To coordinate primary care and non-institutional services, use
care across the sectors, case management can be of micromanagement techniques to ensure appro-
implemented. With full integration the respon- priate quality of care and to control costs (e.g.,
sibility for the integrated organisation lies in one utilization review, disease management) and mul-
place. This applies to all services, resources and tidisciplinary teams working across the network
funding. This may be incorporated in one man- with joint clinical responsibility for outcomes
aged structure or through contractual agreements [23, p. 73]. A key element of Kaiser Permanentes
between different organisations. Furthermore, approach to chronic care is the categorisation of
this classification of integration degree can be their chronically ill patients into three groups
expanded by introducing a second dimension, based on their degree of need [23, p. 73].
i.e., the user needs. User need should be defined
by criteria, like stability and severity of condi-
tion, duration of illness (chronic condition), ser- Strategies for Successful
vice needed and capacity for self-direction Implementation
(autonomy). Accordingly, a low level of need will
not require a fully integrated system, then [10, If one asks staff members of care institutions
24]. about integrated care and potential hurdles they
Kaiser Permanente is a good example of what often mention competing organisational objec-
has been described as a fully integrated system. tives and different employment terms that consti-
According to Goodwin et al. [26] a set of key ele- tute major barriers to effective care, and as
High Multidisciplinary teams manage all care

in all key settings
Common records used as part of joint
practice/management
Fund pooling for purchasing from both
sides/new service
User need
Identify population at risk

Moderate Discharge planning
Routine, bidirectional reporting
Case managers/linkage staff
Defined payment arrangements
Identify emergent need

Low Refer and follow-up
Provide information on request
Understand who pay for what
Linkage Coordination Integration
Level of integration
Fig. 1.1 Levels of integration and user needs as defined by Nolte (Used with permission from Nolte and McKee [23])
Britnell states, point to incoherent policy require- Table 1.1 Continuum of integrated care strategies
ments (such as the requirements of collaboration Funding:
and competition), and leaves the impression that Pooling of funds (at various levels)
some policy initiatives focus on the means and Prepaid capitation (at various levels)
not the end [27]. This unbalanced attitude towards Administrative:
collaboration and competition was also con- Consolidation/decentralisation of responsibilities/
firmed in a seminar of The Nuffield Trust, where functions
participants complained that policy-makers had Intersectional planning
given more attention to the development of com- Needs assessment/allocation chain
petition in the NHS than the promotion of col- Joint purchasing or commissioning
laboration and integration [28, p. 2]. This looks Organisational:
Collocation of services
very much like negative coordination and its
Discharge and transfer agreements
related consequences. It would be interesting to
Inter-agency planning and/or budgeting
further analyse the underlying politics.
Service affiliation or contracting
With a view to international experiences and
Jointly managed programmes or services
research one can identify about ten core ingredi-
Strategic alliances or care networks
ents which decide on successful integration
Consolidation, common ownership or merger
efforts [27]: a defined and registered popula-
Service delivery:
tion; aligned incentives; shared and joint
Joint training
accountability; seamless IT and information sys-
Centralised information, referral and intake
tems; shared clinical protocols and pathways; Case/care management
collaboration between clinicians and managers; Multidisciplinary/interdisciplinary teamwork
authentic patient involvement; relentless focus on Around-the-clock (on-call) coverage
quality improvement systems; collaborative Integrated information systems
organisational cultures; and inspired leadership Clinical:
that endures and is not continually reorganised Standard diagnostic criteria (e.g., DSM IV)
[27, p. 2]. Shortell et al. [29] discuss various Uniform, comprehensive assessment procedures
organisational models of integrated care delivery. Joint care planning
Independent of which particular model is in the Shared clinical record(s)
focus, information systems are in a central posi- Continuous patient monitoring
tion, as they provide data across the whole care Common decision support tools (i.e., practice
system. guidelines and protocols)
Kodner et al. [30] identified five areas to work Regular patient/family contact and ongoing support
on when planning and implementing integrated From Open Access Source: Kodner and Spreeuwenberg
care projects: funding, administrative, organisa- [30]. Originally adapted from Kodner and Kyriacou [42]
and Leutz [24]
tional, service delivery, and clinical [31]. Table
1.1 gives an overview. Without consideration of
these interdependent areas a successful imple- care. The structure, segregation of funds and
mentation might fail due to the barriers and bot- the flow of money.
tlenecks which often occur at various levels. The Administrative: The manner in which govern-
key issues are: ment regulatory and administrative functions
are structured and devolved can help elimi-
Funding: More often than not, form follows nate program complexities, streamline eligi-
financing [sic: not function]. This means that bility and access, and better manage system
the division, structure and flow of funds for resources.
health and social care and related services Organisational: Networking, both vertically
can affect virtually all aspects of integrated and horizontally and through formal or infor-
mal means, is a major method to improve how
organisations work together. Collaboration is the remaining barriers to integrated care are not
another important strategy. Joint working technical, they are political. Last but not least,
relationships within and between agencies in staff members hope to improve their job
the health and social care sectors can opti- satisfaction.
mise resources, facilitate overall efficiency, There is some evidence of a positive impact of
and enhance the capacity for seamless care, integrated care programmes on the quality of
that is, the smooth and uninterrupted provi- patient care [10, 34]. There is also a cautious
sion of necessary care. appraisal that warns that Even in well-
Service delivery: The mode of service delivery performing care groups, it is likely to take years
and management how staff are trained, per- before cost savings become visible [35 p. 431].
form their responsibilities and tasks, work Based on a literature review from 1996 to 2004
together, and relate to patients and family car- Ouwens et al. [11] found out that integrated care
ers and their needs have a major impact on programmes seemed to have positive effects on
a number of critical variables in integrated the quality of care. The most common concepts
care. Such variables include service access, of integrated care were disease management pro-
availability and flexibility, continuity and co- grammes involving:
ordination of care, consumer satisfaction, and
quality and cost outcomes; and, self-management support
Clinical: Shared understanding of patient patient education programmes
needs, common professional language and Combined with one or more of:
criteria, the use of specific, agreed-upon prac- structured clinical follow-up
tices and standards throughout the lifecycle of case management
a particular disease or condition, and the application of multidisciplinary patient care
maintenance of ongoing patient-provider teams
communication and feedback are essential multidisciplinary clinical pathways
quality ingredients in integrated care [32, Patient feedback, patient reminders
p. 4]. education for healthcare professionals.
However, disease management programmes

Objectives, Expectations, focused on approaches that improved the finan-
and Evidence cial balance of the organisation, potentially to the
detriment of the patient. In the Netherlands the
Patients expect seamless care without gaps, and a market power of the care groups, antitrust con-
high quality of care, which is described by access cerns and possibly limited choices for patients
to services, interactions with physician/health- are discussed, since care groups work with pre-
care professional, and interactions with other ferred providers [36].
health-care providers, personal self-responsibility, Nevertheless, because of the variation in defi-
and communication. Providers expect provider nitions of integrated care programmes and the
connectedness. Health insurance managers want components used cover a broad spectrum, the
expenditure savings due to reduction of redun- results should be interpreted with caution. They
dancies in care, and governments and politicians also emphasise the relevance of clear and consis-
see an appropriate way of closing the gap between tent definitions as well as well described inter-
rising expenditures and decreasing financial ventions in order make reliable comparisons of
options. Berwick et al. [33] describe the political programmes and to show the efficiency of the
position as pursuing simultaneously the triple integrated care approaches.
aim of improving the experience of care, improv- Economics and financing issues are important
ing the health of populations, and reducing per drivers for integrated care approaches [10].
capita costs of health care. But they also state that Nevertheless, sound economic evaluations of
integrated care approaches are missing. In their nerves (causing neuropathy), and other structures
systematic review of the effectiveness of inte- such as skin (causing cheiroarthropathy) and the
grated care Ouwens et al. [11] could report on lens (causing cataracts). Different degrees of
only seven (about 54 %) reviews which had macrovascular, neuropathic and cutaneous com-
included an economic analysis. Four of them plications lead to the diabetic foot. A propor-
showed financial advantages. In their study tion of patients, particularly with type 2 diabetes
Powell Davies et al. [34] found that less than have metabolic syndrome including central adi-
20 % of studies that measured economic out- posity, dyslipidaemia, hypertension and non
comes found a significant positive result. alcoholic fatty liver disease. Glucose manage-
Similarly, de Bruin et al. [37] evaluated the ment can have severe side effects, particularly
impact of disease management programmes on hypoglycaemia and weight gain. Under-treatment
health-care expenditures for patients with diabe- is not only associated with long term complica-
tes, depression, heart failure or chronic obstructions but infections, vascular events and increased
tive pulmonary disease (COPD). Thirteen studies hospitalisation. Absence of treatment in type 1
of 21 showed cost savings, but the results were diabetes can rapidly lead to diabetic keto-acidosis
not statistically significant, or not actually tested and death. Indeed, a common feature between
for significance. However, one must bear in mind type 1 and type 2 diabetes is substantial prema-
that important variation was found between the ture mortality, significantly reduced quality of
studies with respect to study design, number and life [40] and increased co-morbid depression.
combination of components of disease manage- Diabetes doubles the risk for depression, and on
ment programmes, interventions within compo- the other hand, depression may increase the risk
nents, and characteristics of economic evaluations for hyperglycaemia and finally for complications
[37]. of diabetes [41]. Essentially, diabetes affects
Central criticism comes from Evers [38] who every part of the body once complications set in,
criticises the methods of the available studies. and the crux of diabetes management is to nor-
Instead of measuring the relative efficiency malise (as much as possible) the blood glucose
(using cost-effectiveness, cost-utility) he wants and manage any associated risk factors, thereby
to see the evidence in terms of absolute efficiency preventing complications and maintaining the
(cost-benefit analyses). Only in the case of abso- highest quality of life. On the whole, metabolic
lute efficiency can outcomes be interpreted as if syndrome responds to oral medications (with
in a business investment calculation. So, well- lifestyle approaches as a very important compo-
designed economic evaluation studies of inte- nent of management). However, glucose manage-
grated care approaches are needed, in particular ment requires minute by minute, day by day
in order to support decision-making on the long- management addressing the complexity of diabe-
term financing of these programmes [30, 39]. tes, including clinical and behavioural issues.
Savings from integrated care are only a hope as While other conditions also have the patient as
long as there is no carefully designed economic therapist, diabetes requires a fully empowered
analysis with a kind of full-cost accounting. patient with all of the skills, knowledge and moti-
vation every hour of the waking day. A patient
that is fully engaged in self-management, and has
The Objectives of Integrated support systems, is empowered to manage their
Diabetes Care diabetes and will likely experience better out-
comes compared with those who do not have
Diabetes is a condition in which longstanding access to this support. Given the complexity of
hyperglycaemia damages arteries (causing mac- diabetes treatment and self-management, inte-
rovascular, e.g., ischaemic heart, peripheral and grated systems should include a means for diabe-
cerebrovascular disease, and microvascular dis- tes self-management education and support over
ease, e.g., retinopathy, nephropathy), peripheral the life-course of diabetes. Additionally, attention
to the psychosocial aspects of diabetes is criti- 4. National Collaboration for Integrated Care and
Support. Integrated care and support: our shared com-
cally important to consider. Unfortunately, as dis-
mitment. London: Department of Health; 2013.
cussed above, modern health systems do not 5. N.N. Negative Koordination. n.d. Available from:
work in this way for a variety of reasons. As inte- http://www.olev.de/k/koordination.htm
grated systems become more widespread, con- 6. Scharpf FW. Coordination in hierarchies and net-
works. In: Scharpf F, editor. Games in hierarchies and
sideration of patient-centred care from a variety
networks. Frankfurt a.M.: Campus-Verl.; 1993:
of sources should be included. Furthermore, in 12565.
diabetes, the boundaries between primary care 7. Holten Mller NL, Dourish P. Coordination by avoid-
and secondary care are blurred. Diabetes special- ance: bringing things together and keeping them apart
across hospital departments. Proceedings of the 16th
ist services, although secondary care, can provide
ACM international conference on supporting group
primary care, and there are GPs, diabetes educa- work. 2010:9.
tors, and other ancillary providers who can pro- 8. Zingerli C, Bisang K, Zimmermann W. Towards pol-
vide a level of specialist care. In view of this, icy integration: experiences with intersectoral coordi-
nation in international and national forest policy.
another more practical definition of horizontal
Berlin Conference 2004 on the human dimension of
integration can be articulation of activities under global environmental change Greening of Policies
the same or separate organisations working with Interlinkages and Policy Integration. Berlin; 2004.
different policies/structures and of vertical inte- 9. Wenzel H. N 2.10 The global financial crisis and
health. South East Eur J Pub Health Spec. vol 2015. A
gration articulation of activities under the same
global public health curriculum 1st version.
or separate organisations working within the 2015:929.
same policies/structures. 10. Nolte E, Pitchforth E. What is the evidence on the
This book describes different approaches to economic impacts of integrated care? Copenhagen
World Health Org. 2014;11:47888.
integrating diabetes care and their outcomes from
11. Ouwens M, Wollersheim H, Hermens R, Hulscher M,
a range of perspectives including that of a patient Grol R. Integrated care programmes for chronically ill
organisation (Diabetes UK). The chapters cover patients: a review of systematic reviews. Int J Qual
different health systems from taxpayer funded Health Care. 2005;17(2):1416.
12. Vogeli C, Shields AE, Lee TA, Gibson TB, Marder
and free at the point of care (the United Kingdom
WD, Weiss KB, et al. Multiple chronic conditions:
National Health Service: three different prevalence, health consequences, and implications for
approaches), Sweden, and United States Veterans, quality, care management, and costs. J Gen Intern
to a range of insurance based systems in the Med. 2007;22 Suppl 3:3915.
13. Committee on Quality of HealthCare in America
United States (University and Health Management
IoM. Crossing the quality chasm: a new health system
Organisation), Germany, South Africa, and the for the 21st century. Summary. Washington, DC:
Netherlands. The final chapter pulls together the National Academy Press; 2001.
common and differing themes to paint a picture 14. Laplante A. Integrated systems improve medical care,
control costs, according to Enthoven. Stanford report.
of how perhaps, the ideal diabetes integrated care
26 Oct 2005.
system should be organised. 15. New South Wales Government. Info summary inte-
grated care. n.d.
16. Haggerty JL, Reid RJ, Freeman GK, Starfield BH,
Adair CE, McKendry R. Continuity of care: a multi-
References disciplinary review. BMJ. 2003;327(7425):121921.
17. Juhnke C, Mhlbacher AC. Patient-centredness in
1. Bunker JP. Medicine matters after all. J R Coll integrated healthcare delivery systems needs, expec-
Physicians Lond. 1995;29(2):10512. tations and priorities for organised healthcare sys-
2. Glasby J, Dickinson H, Peck E. Guest editorial: part- tems. Int J Integr Care. 2013;13(28):e051.
nership working in health and social care. Health Soc 18. Nair KM, Dolovich LR, Ciliska DK, Lee HN. The
Care Community. 2006;14(5):3734. perception of continuity of care from the perspective
3. Simmons D, Wenzel H. Diabetes inpatients: a case of of patients with diabetes. Fam Med.
lose, lose, lose. Is it time to use a diabetes-attributable 2005;37(2):11824.
hospitalization cost to assess the impact of diabetes? 19. Price M, Lau FY. Provider connectedness and com-
Diabet Med. 2011;28(9):112330. munication patterns: extending continuity of care in
the context of the circle of care. BMC Health Serv 31. Leichsenring K. Developing integrated health and
Res. 2013;13:309. social care services for older persons in Europe. Int
20. Gulliford M, Naithani S, Morgan M. What is continuity J Integr Care. 2004;4:e10.
of care? J Health Serv Res Policy. 2006;11(4):24850. 32. Kodner D. All together now: a conceptual exploration
21. Schrijvers G, Goodwin N. Looking back whilst mov- of integrated care. Healthc Q. 2009;13(sp):615.
ing forward: observations on the science and applica- 33. Berwick DM, Nolan TW, Whittington J. The triple
tion of integrated care over the past 10 years and aim: care, health, and cost. Health Aff (Millwood).
predictions for what the next 10 years may hold. Int 2008;27(3):75969.
J Integr Care. 2010;10:e057. 34. Powell Davies G, Harris M, Perkins D, Roland M,
22. Grne O, Garcia-Barbero M. Trends in integrated care Williams A, Larsen K, et al. Coordination of care
reflections on conceptual issues Copenhagen: World within primary health care and within other sectors: a
Health Organization, Regional Office for Europe. 2002. systematic review. Research Centre for Primary
23. Nolte E, McKee M. Integration and chronic care: a Health Care and Equity, School of Public Health and
review. In: Nolte E, McKee M, editors. Caring for Community Medicine, UNSW, 2006.
people with chronic conditions: a health system per- 35. de Bakker DH, Struijs JN, Baan CB, Raams J, Wildt
spective. Maidenhead: Open University Press; 2008. J-E, Vrijhoef HJM, et al. Early results from adoption
p. 1. online resource (xxi, 259). of bundled payment for diabetes care in the
24. Leutz WN. Five laws for integrating medical and social Netherlands show improvement in care coordination.
services: lessons from the United States and the United Health Aff (Project Hope). 2012;31(2):42633.
Kingdom. Milbank Q. 1999;77(1):77110. ivv. 36. Struijs JN, Baan CA. Integrating care through bun-
25. Ahgren B, Axelsson R. Evaluating integrated health dled payments lessons from The Netherlands. N
care: a model for measurement. Int J Integr Care. Engl J Med. 2011;364(11):9901.
2005;5:e01. discussion e3, e9. 37. de Bruin SR, Heijink R, Lemmens LC, Struijs JN,
26. Goodwin N, Peck E, Freeman T, Posaner R. Managing Baan CA. Impact of disease management programs
across diverse networks of care: lessons from other on healthcare expenditures for patients with diabetes,
sectors London: report to the National Co-ordinating depression, heart failure or chronic obstructive pul-
Centre for NHS Service Delivery and Organisation R monary disease: a systematic review of the literature.
& D (NCCSDO). 2004. Health Policy. 2011;101(2):10521.
27. Britnell M. Patients need to be the focus of integrated care. 38. Evers SM. Value of integrated care: revival of the
Healthcare network [internet]. 2013. Available from: monetary valuation of health care benefits. Int J Integr
http://www.theguardian.com/healthcare-network/2013/ Care. 2010;10:e045.
jun/04/patients-need-to-be-focus-of-integrated-care 39. Vondeling H. Economic evaluation of integrated care:
28. Ham C, Smith J. Removing the policy barriers to inte- an introduction. Int J Integr Care. 2004;4:e20.
grated care in England the Nutfield trust for research 40. International Diabetes Federation. IDF diabetes
and policy studies in health services briefing. The atlas. 6th ed. Brussels: International Diabetes
Nuffield Trust for Research and Policy Studies in Federation; 2013. Available from: www.idf.org/
Health Services. 2010. diabetesatlas.
29. Shortell SM, Gillies RR, Anderson DA. The new 41. Roy T, Lloyd CE. Epidemiology of depression and
world of managed care: creating organized delivery diabetes: a systematic review. J Affect Disord.
systems. Health Aff. 1994;13(5):4664. 2012;142:S821.
30. Kodner DL, Spreeuwenberg C. Integrated care: mean- 42. Kodner DL, Kyriacou CK. Fully integrated care for
ing, logic, applications, and implications a discus- frail elderly: two American models. Int J Integr Care.
sion paper. Int J Integr Care. 2002;2:e12. 2000;1:e08.
Integrating Outpatient Care
the Toyota Way: An Individualized 2
Multidisciplinary Team-Care Model
for Diabetes Care Delivery
R. Harsha Rao and Peter Perreiah
Introduction pleting all the mandated process measures for

comprehensive care requires far more time than
Diabetes mellitus poses unique challenges for is traditionally available in a single patient visit;
both providers and patients challenges that are and (b) most providers do not themselves possess
arguably more problematic to overcome than skills in all the ancillary disciplines essential for
those posed by other chronic diseases. effective care, such as Diabetes Self-Management
The unique challenge to providers is to satisfy Education (DSME) or Medical Nutrition Therapy
two specic demands in diabetes care. The rst is (MNT).
to anticipate and recognize the onset of compli- Diabetes presents patients with similarly
cations through comprehensive diabetes care, unique dual challenges in mastering diabetes
which demands meticulous attention to a large self-management with self-awareness, self-
number of process-of-care measures at each visit. empowerment and self-condence.
The second, arguably greater challenge for pro- Comprehensive Diabetes Self-Management
viders is to forestall the development of compli- demands the acquisition of a variety of skills in
cations through effective diabetes care, which order to full a multitude of tasks in many differ-
demands mastery over many different skills in a ent areas of daily life. Effective Diabetes Self-
variety of distinct elds in order to achieve per- Management, on the other hand, demands
formance goals covering multiple facets of man- constant vigilance, consistent discipline and per-
agement. Individually and collectively, these dual sistent attention over a lifetime, without respite,
challenges constitute a virtually unsustainable to nutritional self-discipline, monitoring blood
burden for providers. That is because (a) com- glucose levels, and adherence to antidiabetic
medication use. Together, they constitute a bur-
den that most patients nd difcult to sustain
even with expert assistance, and all-but-
impossible without it.
Not surprisingly, achieving successful and
sustained self-management remains just as elu-
R.H. Rao (*) sive for patients as delivering comprehensive and
VAPHS-UD, University Drive, Room7W-109, effective care is for many providers. National
Pittsburgh, PA 15240 USA Health and Nutrition Examination Surveys
e-mail: [email protected]
(NHANES) show that approximately half of dia-
P. Perreiah betic patients in the U.S. fail to reach goals in
Sapience Technologies, Alpharetta, GA, USA
e-mail: [email protected] each of the three major performance (outcome)

12 R.H. Rao and P. Perreiah
measures in diabetes care (A1c <7 %, BP just as culpable as clinical inertia in the further-
<130/80, and LDL <100 mg/dl) [1, 2]. Those sta- ance of therapeutic futility, specically with
tistics are disappointing in themselves, given that regard to how that paradigm drives traditional
the development of diabetic complications is clinical models for diabetes care delivery.
closely linked to a failure to attain and maintain
each of those three goals. It is even more trou-
bling that just ~19 % of patients are successful in Traditional Clinical Models
achieving all three goals, which is the hallmark for Diabetes Care in the US
of effective care (i.e., care that forestalls
complications). There are two models currently in use for diabe-
The inherent complexity of delivering com- tes care delivery in the US:
prehensive and effective diabetes care is not in
doubt, but the fact that effective diabetes care 1. The Single Provider-Patient dyad: This
remains an exercise in futility in ~80 % of patients model, which is the most widely used method
suggests that factors other than complexity may for diabetes care delivery, is predicated on the
be at work. One major contributor, according to principle that one provider can cover all
Phillips et al, is Clinical Inertia, which they aspects of diabetes care and management for a
dene as recognition of the problem, but a fail- patient with diabetes. The undeniable advan-
ure to act [3]. Although the term might appear tage of this model lies in the intimacy that
self-explanatory, the authors make it a point to characterizes one-on-one interactions. Such
restrict its application to conditions like diabetes, intimacy becomes the foundation of personal-
hypertension and hyperlipidemia, for which ized care that makes it possible to individual-
goals for management are well dened, effec- ize goals selectively and calibrate intensity,
tive therapies are widely available, and practice depending on patient need. These advantages
guidelines for each of these diseases have been are offset, however, by the constraints of time-
disseminated extensively. These criteria explic- delimited patient visits, which are mandated
itly exclude a failure to act because the cause or to meet productivity targets or necessitated by
signicance of an identied symptom or abnor- the individual practitioner looking to the bot-
mality is unknown or unclear [3]. tom line. Such time constraints make it impos-
A failure to intensify therapy despite clear sible for one provider to cover multiple tasks
indication of benet the essence of clinical in a comprehensive manner at any visit, forc-
inertia has been ascribed to a widespread ten- ing compromises in task selection at any visit.
dency of providers to either justify inaction with Inevitably patients and providers nd them-
soft reasons (essentially excuses) like improv- selves prioritizing tasks depending on per-
ing control or target almost reached [3], or ceived immediacy and need. These constraints
overestimate the care they provide [4]. According prevent consistent fullment of all the process-
to Philips et al, the root cause is a failure of medi- of-care measures required to detect and pre-
cal education and training programmes to empha- vent complications.
size the importance of focusing on the Another major drawback of the Single
achievement of therapeutic goals, or teach prac- Provider-Patient dyad is that most clinical
tice organization to achieve therapeutic goals [3]. providers cannot full patient needs for inte-
While there is no denying the critical importance grated care, simply because the skills required
of such provider-driven factors, attributing a for DSME and MNT are outside the domain
failure to attain therapeutic goals in diabetes to of most clinical care providers. Even when the
clinical inertia alone runs the risk of oversimpli- importance of these interventions for effective
fying a complex problem that may have more diabetes care is recognized, they require ad
than just one layer. The current paradigm of reim- hoc referrals to other providers who possess
bursement for chronic care in the U.S. may be the requisite skill sets. The necessity for such
2 Integrating Outpatient Care the Toyota Way: An Individualized Multidisciplinary 13
referrals burdens patients with multiple visits, global parameters, such as the percentage of
so that their success is subject to the patient all patients above or below some threshold
determining whether their perceived impor- A1c (e.g., <7 % or >9), not from individual
tance is worth the inconvenience of additional patient outcomes. Put another way, at the indi-
visits. With no assurance of follow-through, vidual level, the system provides a greater
comprehensive management becomes hostage nancial reward for treating complications
to patient discretion. after they occur (downstream revenue genera-
It is clear, therefore, that the economic tion), rather than preventing them (upstream
costs of additional visits for the patient, pro- cost reduction).
vider time constraints and a lack of provider The barriers to integrated care delivery in
skills in MNT and DSME combine to contrib- the traditional single patient-provider dyad
ute as much as clinical inertia to the failure of have led to the development of alternative
the Single Provider-Patient dyad to deliver models for chronic care (including for diabe-
comprehensive and effective diabetes care. tes) based on the concept of a Patient-Centered
Care fragmentation with this model is exacer- Medical Home (PCMH) [8]. At its most fun-
bated further by arcane rules of fee-for-service damental level, the goal of PCMH is to maxi-
reimbursement in the US, which disallow mize health outcomes by providing
reimbursement for some services rendered by comprehensive and continuous medical care
more than one provider for the same principal led by a healthcare provider through team-
diagnosis (Diabetes, in this case) on the same based healthcare delivery. The PCMH concept
day (e.g., for Clinical Care and DSME), with of integrated multidisciplinary care delivery is
the exception of some types of MNT [5, 6]. at the core of the Group Visit model for
Since integrated multidisciplinary care, by diabetes.
denition, calls for contemporaneous and syn- 2. The Group Visit model: The inherent inabil-
chronized care by more than one provider, ity of the Single Provider-Patient dyad to
each with a different skill set, such care deliver comprehensive disease management
becomes nancially unsustainable if only one for patients with diabetes has led to the intro-
or two providers (out of three or four) are duction of the Group Visit model to address
reimbursed. This is one reason fee-for-service and overcome the inefciencies and inade-
reimbursement can be a prohibitive disincen- quacies noted above [9]. The Group Visit
tive to the integration of multidisciplinary model is founded on the premise that many
care in diabetes. facets of diabetes care are repetitive for indi-
Another reason is that fee-for-service, the vidual patients and replicative with rela-
most widespread nancial model in U.S. tively small variation across patients. In
healthcare, adds a layer of particular complex- this model a group of patients receives serial
ity to chronic disease care. For the most part, input from multiple providers covering dif-
fee-for-service reimbursement couples pay- ferent prespecied areas in one session. This
ment to the volume of services provided, not assures comprehensive coverage of multiple
the overall cost or outcomes. Thus, providers facets of diabetes care (breadth of care) with
are rewarded for increasing volume, which the added advantage of achieving higher
does not necessarily translate into greater patient throughput (efciency/volume). The
value [7]. This model may work for acute Group model enables multiple providers with
care, where treatment is the goal, but not for different specialized skills to deliver all
chronic care, where prevention takes prece- aspects of diabetes care (MNT and DSME in
dence over treatment. Even though reimburse- particular) to a group of patients in a single
ment for chronic care is being increasingly session. Thus, Group Visits are designed to
linked to provider performance, diabetes- full at least in theory the current deni-
specic performance is usually measured by tion of Chronic Disease Management (CDM)
as a group of coherent interventions, day is required for such calibration and risk
designed to prevent or manage one or more stratication.
chronic conditions using a.systematic and An additional criticism of the Group Visit
structured multidisciplinary approach poten- model is that achieving the aforementioned
tially employing multiple treatment modali- economies of scale requires large patient num-
ties. The goal of chronic disease prevention bers and a signicant increase in resource
and management is to identify persons with allocation, including-infrastructure changes
one or more chronic conditions, to promote and manpower commitments. The need for
self-management by patients and to address such resources is a stumbling block to the
the illness or conditions according to disease widespread acceptance of this model outside
severity and patient needs and based on the of large organizations like Accountable Care
best available evidence, maximizing clinical Organizations (ACOs). Recent changes in
effectiveness and efciency regardless of coding and billing do incentivize ACOs to
treatment setting(s) or typical reimbursement adopt Group visits for diabetes care. However,
patterns. Routine process and outcome such factors provide little incentive for indi-
measurements should allow feedback to all vidual practitioners without access to the
those involved, as well as to adapt the pro- infrastructure and resources necessary for
gramme [10]. Group visits. For these reasons the adoption of
The increasing adoption of the Group Visit the Group Visit model remains limited primar-
model in larger healthcare programmes has led ily to ACOs.
to changes in reimbursement rules for Group
visits and new billing codes for such visits [6].
This allows for economies of scale that can A Brief Overview of Healthcare
overcome the fact that per-patient reimburse- Delivery in the VA System
ments for group visits are individually too low
to be protable. Unfortunately, studies show The Veterans Health Administration is in many
that while the model reliably delivers compre- ways unique (for the U.S.). Run by the Veterans
hensive care reected in process-of-care mea- Affairs Department of the Federal Government, it
sures (i.e., documentation in identied diabetes is the largest integrated healthcare system in the
care domains), it does not consistently deliver U.S., serving 8.76 million Veterans each year
effective care (i.e., achieving BP, lipid or gly- through more than 1700 sites of care, including
caemic goals) [9]. A recent meta-analysis of hospitals, community clinics and community liv-
randomized control trials is more encouraging, ing centres, domiciliary units, Vet Centres, and
with reductions in A1c ~0.5 %, but not blood various other facilities [12].
pressure or cholesterol [11]. A brief summary of VA healthcare benets
The reason why Group Visits fail to consis- follows for the benet of readers unfamiliar with
tently achieve performance targets is not clear, the VAs mission and mandate. Even though this
but one is left to wonder whether the absence summary is excerpted (almost) verbatim from the
of personalized care might play a role. A key source document, it must, of necessity, be incom-
component of CDM, as dened above, is cali- plete, in the interests of brevity. The authors
bration according to disease severity and risk explicitly deny any claim that what follows is a
stratication based on patient need. Group comprehensive or accurate description of the full
visits, by their very nature, are incapable of panoply of federal benets available to qualify-
delivering individualized care calibrated to ing Veterans. Readers are strongly advised to
patient needs and risk stratication. access the source document from which this
Consequently, a face-to-face visit in a Single summary is excerpted to verify/correct any
Provider-Patient dyad visit, either after the details that may be vague, incorrect, missing or
Group Visit, or in a separate visit on another misleading [12]. The key summary features are:
Basic eligibility: VA healthcare benets are cinations, or screening for hypertension,

available to any person who served 24 contin- hepatitis B, tobacco, alcohol, hyperlipidemia,
uous months or the full period for which he/ breast cancer, cervical cancer, Human papil-
she was called to active duty in the active lomavirus (HPV), colorectal cancer by faecal
military, naval, or air service and who was occult blood testing, education about the risks
discharged or released under conditions other and benets of prostate cancer screening, HIV
than dishonourable. Reservists and National testing and prevention counselling (including
Guard members may also qualify for VA the distribution of condoms), and weight
healthcare benets if they are called to active reduction or smoking cessation counselling
duty (other than for training only) and com- (individual and group). Laboratory, at plain
plete the full period for which they were called lm radiology, electrocardiograms, and hos-
or ordered to active duty by Federal order. pice care and in-home video telehealth are
This minimum duty requirement may be also exempt from copays.
waived for veterans discharged for hardship, Medication Copayments: While many
early out or a disability incurred or aggravated Veterans are exempt for medication copays,
in the line of duty. nonservice-connected Veterans in Priority
Service connection: The VA prioritizes health- Groups 7 and 8 are charged $9 for each 30-day
care enrolment based on degree of service supply of medication, provided on an outpa-
connected disability, ranging from highest pri- tient basis for treatment of a nonservice-
ority (>50 % service connection, Priority connected condition. Veterans enrolled in
Group 1) to lowest (no service connection), Priority Groups 2 through 6 are charged $8 for
and applies geographic mean income thresh- each 30-day or less supply of medication; the
old tests to further stratify priority in those maximum copay for medications that will be
without service connection (Groups 78). charged in calendar year 2013 is $960 for
Inpatient care: Copayments for inpatient stays nonservice-connected medications. Copays
range from zero for the highest priority groups apply to prescription and over-the-counter
to a maximum of $1216 for inpatient stays up medications, such as aspirin, cough syrup or
to 90 days for those above the income thresh- vitamins, dispensed by a VA pharmacy.
old in the lowest priority group. Copays are not charged for medical supplies
Copayments for Outpatient Care: Many such as syringes or alcohol wipes.
Veterans qualify for free healthcare services The preceding paragraphs have been tran-
based on a VA compensable service-connected scribed from the source document (with only
condition or other qualifying factor, but most minor edits). We add two items to the above
are asked to provide a nancial assessment to that are essential to complete the picture of an
determine if they qualify for free services. integrated healthcare system that has more in
Veterans whose income exceeds the estab- common with government-run healthcare sys-
lished VA Income Thresholds as well as those tems in other countries (akin to the U.Ks
who choose not to complete the nancial NHS), than the indemnity insurance-based
assessment must agree to pay required copays healthcare system that predominates in the
to become eligible for VA healthcare services U.S. The rst is that VA employees, including
(Primary Care Services: $15; Specialty Care physicians, are either salaried employees of
Services: $50). The copay amount is limited to the U.S. government (for the most part) or fee-
a single charge per visit regardless of the num- based contractors compensated by time or
ber of healthcare providers seen in a single patient volume (either way, care decisions are
day, and is based on the highest level of clini- not linked to nancial incentives or disincen-
cal service received. Copays do not apply to tives). The other is the way in which care is
outpatient visits solely for preventive screen- delivered and coordinated within the VA sys-
ing and/or inuenza and pneumococcal vac- tem, with specic relevance to diabetes care
delivery at VAPHS, which we describe briey plications requires input from many different
in our own words. specialists, whereas hospitalizations for acute
Primary, Secondary and Tertiary Care emergencies often fall to hospitalists and criti-
Delivery at VAPHS: The VA system is orga- cal care specialists. Thus, diabetes care is
nized into regional collaboratives called fraught with the potential for sometimes con-
Veterans Integrated Service Networks (or icting, even contradictory management strat-
VISNs), usually comprised of one or two ter- egies, making care coordination mandatory
tiary care Hub hospitals (the Pittsburgh and for success. This is an area in which the VA
Philadelphia VA hospitals are, respectively, system excels, with its integrated network,
the Western and Eastern hubs in VISN4), sev- common electronic record, and shared respon-
eral feeder Spoke hospitals for each hub, sibility for care.
which provide both secondary and primary
care, and a number of Community Based
Outpatient Clinics (CBOCs) devoted to pri- The Need for a New Model
mary care, clustered at varying distances to Deliver Outpatient Diabetes Care
around each spoke and hub hospital, based on
geographic location. All patients must have a A realization that the traditional Single Provider-
primary care provider (PCP) who directs and Patient dyad used at the Pittsburgh VA was
coordinates care, including referrals for spe- incompatible with delivering both comprehen-
cialty care, following the concept of a Patient- sive and effective diabetes care encouraged us to
Centered Medical Home that emphasizes explore other avenues for diabetes care delivery.
care coordination and communication to We understood, furthermore, that the alternative
transform primary care into what patients of Group visits would require major changes to
want it to be [8]. All documentation is elec- infrastructure that were not practical or nan-
tronic (paperless), through the VAs unique cially feasible at our institution. Third, we were
Computerized Patient Record System (CPRS), emboldened to develop a third way by the fact
which allows nationwide access to patient that there would be no nancial disincentives to
records, regardless of location. Care coordina- multi-provider visits in an integrated healthcare
tion, with the PCP acting as the gatekeeper, is system like the VA, unlike a fee-for-service sys-
an integral component of care across the VA, tem. Lastly, the VA system has the unique ability
but policies governing how that coordination to integrate and coordinate care across multiple
is achieved are set at the local level, and thus disciplines.
vary by location. At the Pittsburgh VA These were the reasons why we explored the
(VAPHS), all specialty care providers are feasibility of constructing a chronic disease care
required to send Inter-facility model centred on an integrated multidisciplinary
Communications via CPRS to the PCP after team that would deliver diabetes care that was
any specialty consultation, documenting both comprehensive and effective, yet retained
assessment and management plans (diagnostic the intimacy of the traditional Single Patient-
and therapeutic). Provider dyad. Critical to the success of that
Care coordination achieves critical impor- effort was funding through a Physician Champion
tance for diabetes, in particular, because of the Award from the Jewish Healthcare Foundation
need for management at many different levels [13], as well as direct advisory guidance during
and locations. At the most basic level, the development and implementation from the
symptomatic management of acute hypo- and Pittsburgh Regional Health Initiative (PRHI), one
hyperglycaemia often devolves to the PCP, of the nations rst regional collaboratives of
even when a specialist oversees more advanced medical, business and civic leaders organized to
strategies for glycaemic management. At address healthcare safety and quality improve-
another level, the wide variety of chronic comments [14].
We started with the fundamental premise that healthcare reform, to hold true for healthcare
the model had to satisfy the needs of both com- delivery [16]. Perfecting Patient CareSM (or PPC)
prehensive and effective care without compro- is PRHIs agship healthcare process improve-
mising either the personalized one-on-one care of ment methodology based on the principles of the
the single patient-provider dyad or the coordi- Toyota Production System. See Fig. 2.2.
nated care of the Group visit model. In other
words, the goal was to preserve the advantages of
both existing models while eliminating their dis- Redesigning the Diabetes Clinic
advantages. In order to achieve such a seemingly at the VA Using Toyota Principles
impossible goal, we turned to industry, speci-
cally the principles of the Toyota Production Our initial purpose in redesigning diabetes man-
Systems [15], to develop a model of multidisci- agement was to simply combine four distinct
plinary outpatient diabetes care that is both com- clinical disciplines in diabetes care (DSME,
prehensive and effective. In order to understand MNT, Blood Pressure/Lipid Management, and
how concepts developed for industrial manufac- Glycaemic Management) into a single, clinic
turing can be applied to bedside medicine, a brief visit. From such crude and unpolished begin-
introduction to the Toyota Way is warranted. nings off-handedly referred to in an initial team
meeting as one-stop shopping, our purpose was
transformed, thanks to direct engagement by
An Introduction to Lean Systems PRHI and funding support from JHF, into a
Design sophisticated application, which we call the
Individualized Multidisciplinary Team-Care
In his book The Toyota Way, Jeffery Liker lays Model. The model, as implemented, has a far
out four Core Tenets for achieving efciency and more ambitious purpose that goes beyond just
improving quality based on Toyota's unique man- patient convenience to the delivery of integrated,
agement system [15]. These Core Tenets, shown multidisciplinary care of high quality that not
in Fig. 2.1, are (i) a Long-term Philosophy, (ii) only meets patient needs but achieves better out-
the Right Process, (iii) People as Partners, and comes. (Parenthetically, it may be noted here that
(iv) Continuous Reection to Solve Problems. our model differs fundamentally from efforts to
Even though these tenets are principally associ- integrate diabetes care in Health Disparities
ated with manufacturing processes, they have Collaboratives (HDC) in the US or the Diabetes
been shown by PRHI, a leader in the eld of
Fig. 2.2 Likers 4P Tenets adapted to an Individualized

Fig. 2.1 Likers 4P model (Adapted with permission Multidisciplinary Team-Care Model for integrated diabe-
from Liker [15]) tes care delivery
Integrated Care Initiative (DICI) in the U.K) [17, Rule 4: Improvements are made using scien-
18]. The Individualized Multidisciplinary Team- tic method, with guidance from a teacher, as
Care Model integrates multidisciplinary collab- close as possible to the work, aiming towards
orative outpatient specialist care (DSME, MNT the ideal.
and clinical) for diabetes in a tertiary care set-
ting, whereas HDC and DICI focused on inte- We operationalized these principles in the pro-
grating patient education and lifestyle cess of implementing our redesign by, rst,
modications (DSME and MNT) into primary
care for diabetes in a community setting [17, 18]. (A) Outlining the actual work required of rede-
Our redesign of diabetes care delivery has a sign (in six stages), then,
direct analogy in manufacturing, where a product (B) Constructing and implementing the model
manufactured in a traditional job shop moves and, nally,
from one functional grouping of machines to (C) Re-evaluating constantly to improve model
another (e.g., stamping, drilling, assembly, paint- efciency and performance (kaizen)
ing, etc). Process redesign in manufacturing is
often done by regrouping machines around the
needs of a product group into a manufacturing (A) Outline the Actual Work of Redesign for
cell. Individual product components enter the Integrated, Multidisciplinary Care
cell in a specied order and are rapidly trans- This was achieved in six stages, as
formed at the cells stations into a nished prod- follows:
uct. In industry, transforming traditional 1. Define the Explicit Purpose of Redesign in
production into cellular production often yields Relation to Care Delivery: After extensive
dramatic improvements in quality, inventory discussions, team members reached con-
reduction and efciency. Distilled to its essence, sensus that any new model for integrating
our redesign of diabetes care delivery is analo- multidisciplinary care in diabetic patients
gous to a cellular manufacturing process, in that must focus on delivering continuing
it involves the regrouping of specied tasks into care, rather than initial care. The rea-
stations responsible for each care discipline, sons for that restriction will become read-
with the patient moving from one station to the ily apparent when we describe the elements
next, accumulating care that is both comprehen- of the model in greater detail, but they can
sive and integrated in the aggregate. be summarized briey as follows:
We were guided in our redesign by four prin- (a) A focused, time-delimited and structured
ciples derived from Spear and Bowens Rules in clinic visit is ideal for implementing and
Use for business, which form the core of PPCSM adjusting an established plan of continu-
[19]. Grunden terms these principles Rules of ing care but ill-suited to the elastic and
Work Design that Guide Process Improvement sometimes drawn-out process of eval-
[16], and describes them as follows: uating, discussing, and getting patient
buy-in for an initial plan of care and
Rule 1: Activities (work) must be highly spec- therapeutic strategy, which can vary
ied as to content, sequence, timing, location greatly in both length and complexity,
and expected outcome. depending on individual patient need.
Rule 2: Connections between customers and (b) An essential precondition, therefore, is
suppliers must be highly specied, direct, to establish an initial plan of care in a
with a clear yes-or-no way to send requests traditional Single Provider-Patient dyad
and receive responses. visit prior to enrolment in the multidis-
Rule 3: The pathway for every product and ciplinary clinic for continuing care,
service must be predened, highly specied, (c) The only other precondition for enrol-
simple, and direct no loops or forks. ment is the patient must have the abil-
ity and motivation to engage in a (b) Decide what, if any, remaining tasks
comprehensive diabetes management can be eliminated or automated.
strategy, and must possess a basic (c) Allocate those tasks to team members
understanding of DSME and MNT. exclusive to their particular skill set.
2. Define the Objectives of Care Delivery in (d) Arrange and assign each team member
Relation to Patient Needs: The redesign to individual stations of care work-
was based on fullling specic patient ing in sequence during each visit
needs, as follows: (e) Assess the cycle times for each mem-
(a) Set individualized clinical goals based ber of the team to complete their cur-
on patient need and risk stratication rent list of tasks at each station.
(b) Meeting 100 % of all process-of-care (f) Allocate any tasks that overlap between
measures (HbA1c, LDL, blood pres- two or more team members, depending
sure, creatinine and urinary microalbu- on skill set, with the goal of balancing
min levels, annual foot and eye exams, the work among all stations.
and aspirin and statin use/contraindi- (g) Continue rearranging station sequenc-
cations/alternatives). ing and/or task lists until all station
(c) Ordering all necessary lab tests to full task lists have about the same cycle
process-of-care measures time and cycle time variability.
(d) Ensuring timely completion (annual at (h) Set up materials, equipment, informa-
least) of periodic Foot and Eye Exams tion systems and back up assistance to
(e) Providing DSME and MNT contem- allow providers to accomplish their
poraneously with clinical care work without interruption.
(f) Enabling process efciency to utilize 5. Run the redesigned process with actual
all resources available to care for the patients:
assigned patient population. (a) Intensively observe whether tasks
3. Document the Current Process for Diabetes assigned to each station can be accom-
Care Delivery, to identify areas of de- plished with high quality and within
ciency/improvement, including: the targeted cycle times.
(a) A complete description of tasks cur- (b) Note any instances where task comple-
rently performed by each provider dur- tion or quality breaks down, and exam-
ing various patient contacts (i.e., for ine individual events for evidence of
clinical care, DSME, and MNT) root causes.
(b) The timing and sequence of all pro- (c) Measure both quality and efciency
vider tasks outcomes, based on delivering high
(c) The actual time for completing provider quality care that is both comprehen-
tasks (cycle times) and their variability sive (i.e., achieves all process-of-care
(d) Any shared tasks requiring joint pro- measures) and effective (i.e., meets
vider participation performance goals for A1c, BP and
(e) Any potentially duplicative tasks by Lipids) in reducing long term
different providers (i.e., task sharing). complications.
(f) The current performance relative to 6. Continuously redesign the process to meet
patient need and efciency. patient, provider and business needs:
4. Sort the tasks as follows: (a) Assess whether patient, provider, and
(a) Identify essential tasks that must be business needs are all met.
accomplished in each continuing care (b) Look to reduce the cycle times of indi-
visit and which belong in other patient vidual tasks.
contacts. (c) Rebalance work between stations.
(d) As the process becomes more stable 2. Define Work Content across the 4 stations .
and efcient, decide by consensus how The rst step was to set Takt time1 to
gains in improvement can be leveraged accommodate a <15 min cycle time at each
to enhance care, reduce provider work- station (total visit length = 60 min), and
load, or service more patients. assure unambiguous work flow (PPC Rule
(e) Call for help outside the team, if addi- #2) along a highly specified path (PPC
tional resources or other enablers are Rule #3), in the following sequence:
needed to support the process in meet- (i) Station 1 (DSME [Cycle Time = 13,-
ing objectives. 2,+4]): The CDE-RN does the follow-
(B) Construct and Implement a Model of ing tasks:
Integrated, Multidisciplinary Care (a) Collect the home blood glucose
The practical aspects of implementing log or download from metre or
our model of integrated multidisciplinary insulin pump
care for diabetes can now be outlined, keep- (b) Measure blood pressure;
ing in mind that the purpose of the redesign (c) Take a nger-stick blood sample to
is explicitly restricted to continuing care. measure HbA1c and Lipid levels in
The model is organized into stations of the clinic (using point-of-care
care, each assigned to a single discipline [POC] laboratory equipment);
and staffed by a provider with particular (d) Provide diabetes education in one
skills in that discipline. These stations are of four predetermined patient
setup in a specied sequence, like a manu- knowledge/skill areas, in a
facturing cell, with individual patients mov- repeating cycle over four visits. It
ing through each station and service elements is vital that the patient be familiar
of diabetes care delivered serially to provide with the basics because the pur-
multidisciplinary care in the aggregate. pose is to review and reinforce
Based on this, a model for diabetes care familiar information, not intro-
delivery was constructed as follows: duce new information. Thus, the
1. Assemble the essential components of dia- patient must participate in a pre-
betes care into a comprehensive patient liminary DSME session prior to
visit involving a team of diabetes care pro- enrolment.
viders assigned to specic stations of The four assigned tasks differ,
care, each responsible for explicitly depending on whether the patient
defined work content related to their exper- needs reinforcement of basic skills
tise (PPC Rule #1), covering all aspects of or more advanced skills, and are
multidisciplinary diabetes care, as calibrated to patient needs. The
follows: four basic skills reviewed are:
(i) A Certied Diabetes Nurse Educator Metre technique
(CDE-RN) Injection technique
(ii) A Diabetologist/Endocrinologist Sick-day and hypoglycaemia
(Team leader, who oversees/problem management, including instruc-
solves at all stations) tion on glucagon administration
(iii) A Nutritionist with CDE certication by spouse/home caregiver
(CDE-RD) Foot care
(iv) A Clinical Pharmacist (Pharm D)
(v) A Nurse Practitioner with CDE certi- 1
Takt time is the maximum amount of time in which a
cation and diabetes management
product needs to be produced. Adjustable time unit used
experience (CDE-NP) in lean production to synchronize the rate of production
with the rate of demand.
More advanced skills for patients In patients on a Multiple Daily Insulin

on an insulin pump include (MDI) regimen or using an Insulin Pump,
priming and relling the insu- the focus of MNT is on more advanced
lin pump skills, including:
infusion set insertion Carbohydrate counting,
technique including verication by food
ability to change pump basal logs, if necessary
rates, and Effect of dietary fat and pro-
familiarity with the pumps tein on carbohydrate
bolus administration tool (e.g., absorption
Carb Smart or Bolus Wizard) Dual, extended and square-
(e) Work content is designed speci- wave bolus strategies, and
cally to assure that the nurse com- Hypoglycaemia prevention
pletes tasks a to c (above) plus strategies, e.g., the proactive
one of the DSME skill areas in d. use of carbohydrate intake
within a cycle time of 13 min on before exercise
average, although that can be as (c) Work content at this station is
short as 11 min, or as long as designed specically to assure
17 min when unexpected delays that the nutritionist weighs the
occur in accessing pump and patient and provides one of the
metre software. MNT skill areas in (b) within a
(ii) Station 2 (MNT [Cycle Time = 11 min, cycle time of 11 min on average,
1,+3]): The CDE Nutritionists tasks although that can be as short as 10
include the following: min, or as long as 14 min
(a) Weigh the patient, discuss impli- (iii) Station 3 (BP-Lipids [Cycle Time
cations of weight gain, or need for 10 min, 4, +1]): The initial con-
weight loss figuration of the model had this sta-
(b) Review dietary principles in one tion manned by a clinical pharmacist
of four predetermined patient who performs the following tasks
knowledge areas over four visits (this configuration changed subse-
in turn in a repeating cycle. Just as quently, for reasons we will outline
for DSME, the intent is to review later):
and reinforce familiar information, (a) Rechecks BP in those not at goal
not introduce new information, at initial measurement (Station 1)
which is why it is essential for the (b) Orders labs as needed for annual
patient to participate in a prelimi- surveillance
nary nutrition education session (c) Performs medication
prior to enrolment. reconciliation
The four assigned tasks differ, (d) Interprets POC Lipid results and
depending on whether the patient reconciles with previous lab results
needs reinforcement of basic (e) Adjusts/intensies/rells BP,
skills or advanced skills. lipid, and aspirin therapy, accord-
The four basic nutritional ing to patient need, to achieve
skills reviewed patient-specic targets (BP
food groups <140/90 in all patients, and
food choice <130/80; LDL <100 mg/dl or <70
hypoglycemia, and mg/dl, depending on risk
portion control. stratication).
(f) Work content at the BP/Lipid sta- nized, in order to provide input for
tion varies more than at any other managing both glycaemia and the
Station, depending on whether or unanticipated problem within the
not the patient is at goals for BP allotted Takt of 15 min. The MD
and Lipid therapy. Thus, cycle then exits allowing the NP to con-
time can be as short as 7 min in centrate on providing extended
patients at goal for both BP and task completion for such patients,
Lipids (which applies to the great while the MD takes the next
majority of patients currently patient in line for Glycaemic
seen in the clinic) up to a maxi- Management, so that there are no
mum of 12 min in the rare patient hold-ups in patient throughput.
needing intensication of both (v) Floating Station (Supervising
BP and Lipid therapy. This Diabetologist): Work content at this
assures task completion with a station consists of the following tasks:
cycle time well within the 15 min (a) See all patients at Station 4 to dis-
Takt time, so that the model is cuss/endorse decisions on glycae-
able to accommodate delays (i.e., mic management
make-up for lost time) at one of (b) Sign off on all changes in therapy
the earlier stations. at Stations 3 and 4
(iv) Station 4 (Glycaemia): A CDE- (c) Provide continuous oversight of
Nurse Practitioner performs the fol- work ow across the four stations
lowing tasks: (d) Act as an on-site problem solver
(a) A diabetes-focused exam (e.g., for interruptions in work ow
injection sites, feet) (e) Function as an extra outlet to
(b) Reviews and records results from maintain work ow when hold-
Diabetes Retinopathy ups occur at any station because
Surveillance Reports of unanticipated complexity (as
(c) Reviews the home blood glucose discussed above).
(or insulin pump) printout (f) Perform medication
(d) Adjusts therapy as needed to meet reconciliation
patient-specic glycaemic targets (g) Document and send Inter-facility
(A1c), calibrated to patient need, Communication to PCP
based on individual risk (h) Seek specialist consultation for
stratication. newly recognized or existing
(e) Ensures compliance with annual problems (e.g., Cardiology,
retinopathy surveillance (referral Nephrology, Podiatry, Vascular
to ophthalmology) Surgery, and Psychiatry etc.)
(f) Work content at this station is pre- 3. Ensure Task Completion through
dictable for the most part (~14 Documentation: Template-based electronic
min) and, while stable, is variable documentation in modular form for each
enough that cycle time can extend station assures completion of all assigned
to as much as 25 min when unan- tasks. Documentation modules for each
ticipated problems or complica- station were developed by individual team
tions are recognized, such as an members and only nalized after extensive
infected abrasion or ulcer on the dialogue among team members to ensure
foot. In such patients, the appropriateness and brevity, and to elimi-
Supervising MD enters Station 4 nate duplication. Previously documented
as soon as the problem is recog- information in CPRS is imported into a
templated note that mandates completion that such hold-ups were of little conse-
of all identied tasks in specic elds at quence at the start of the visit prompted
each Station, while also allowing for inclu- a redistribution of the task of initial BP
sion of free text. Thus, work content and measurement to the DSME Station,
documentation requirement for each visit achieving better work balance and eve-
and station is explicitly defined (PPC Rule ning out cycle times across stations
#1). At the end of the visit, the unique (heijunka).
capability of CPRS allows the four mod- (ii) The reordering of station sequence
ules, each individually signed by the over time: This constitutes a second
assigned provider at each Station, to be example of how evidence from ongoing
combined to appear as a single cohesive monitoring was used to make adjust-
and comprehensive note in the electronic ments in work ow (Fig. 2.3ac).
record, rather than as four separate notes. BP-Lipids was initially thought to be
(C) Re-evaluate constantly to Improve Model ideally positioned as Station 2 (Fig.
Efficiency and Performance (Kaizen) 2.3a. First Iteration), but monitoring
Team meetings are held regularly to con- showed signicant hold-ups in work-
stantly evaluate performance through ow occurring even after it was divested
problem-solving (PPC Rule #4). The purpose of the task of initial BP measurement.
is to engage in team dialogue focused on Continued monitoring revealed that the
making sure the model is working for each hold-ups occurred because it often took
team member, without nger-pointing or >15 min for the POC-lipids test to
blame (the essence of kaizen). We cite three result, which meant the pharmacist did
specic examples of how kaizen was utilized not receive those within the 15 min
to make changes in work content, work ow, takt, with further delays added on
and model design. whenever treatment changes were
(i) The reassignment of the task of BP called for. The BP/Lipids Station was
measurement from the BP/Lipids therefore moved to what was then
Station to its current placement in though to be its ideal position at
Station 1, DSME: This represents an Station #3 in the visit sequence,
early example of how constructive dia- exchanging places with MNT (Fig.
logue based on evidence was used to 2.3b Second Iteration). This allowed
reassign work content in order to for an additional 15 min to elapse while
improve workow. Initially, the team the patient received MNT at the newly
assumed that the natural placement of congured Station #2, before the
the task of BP measurement would be patient was seen for BP/Lipid manage-
in the BP/Lipids Station. However, it ment at Station #3, by which time the
became clear early in implementation POC Lipid result was available for any
that hold-ups at that station were an adjustments in therapy.
intermittent but recurring problem. (iii) Changing the configuration of the model
Evidence from time measurements from its original conception based on
revealed a periodic imbalance in work- changing circumstances. We have been
load because the pharmacist was some- forced into yet another reconguration
times compelled to wait as much as of the model, which further demonstrates
10 min for the patient to reach a resting the exibility of the model. This was
state for accurate BP measurements, prompted by administrative reallocation
particularly when repeat measurements of manpower resources, which termi-
were called for in patients not at goal nated the Clinical Pharmacists partici-
on the rst measurement. A realization pation in the clinic. Consequently, the
Fig. 2.3 Changes made to

Station sequence over time
(kaizen in practice). (a) First
iteration: ve stations in their
original sequence. (b) Second
iteration: sequence reversal of
BP/Lipids and MNT
stations, prompted by hold-ups
traced to POC Lipid results
taking >15 min to become
available. (c) Third iteration
(current), showing BP/Lipids
last in sequence as an
Optional Station.
Supervising MD provides
one-on-one BP/Lipid
management at the end of the
visit in patients not meeting
goals, and oversees glycaemic
management (See text for
details). *Intake restricted to
Continuing/Established Care,
not Initial Care. Key: CDE
Certied Diabetes Educator,
NP Nurse Practitioner, Pharm
D Doctor of Pharmacy, POC
Point-of-care, RD Registered
Dietitian, RN Registered Nurse
tasks assigned to this station were reas- one-on-one intervention for BP/Lipids
signed, of necessity, to the Supervising management to achieve and maintain
Diabetologist, the only team member goal for both measures.
free to engage in completing those The current conguration (Fig.
tasks. As part of the reconguration of 2.3c, Third Iteration) makes use of this
task assignment, it was necessary to fact by effectively combining the last
move BP-Lipids to the last Station in two stations in 90 % of patients meet-
line (Station #4), exchanging places with ing BP and Lipid goals, so that the
Glycaemia, which became Station #3. patient visit ends after three Stations.
The reconguration required the team to The downtime afforded by this com-
accept that the Supervising MD would bination of stations allows the super-
be, of necessity, unavailable to engage in vising MD to complete documentation
glycaemic management on the spot. In tasks for the BP/Lipids Station, includ-
anticipation of this, it was decided to ing medication reconciliation, and
reserve a 30 min time slot at the end of ordering labs in anticipation of the
clinic for specic interactions between next patients needs, during the rst
the NP and MD regarding glycaemic 5 min of the cycle time at Station 3,
management. In the event that changes in while the NP completes a preparatory
recommendations became necessary, glycaemic evaluation. The supervising
these would be subsequently communi- MD then enters Station #3 during the
cated to the patient by the NP, and docu- latter half of cycle time, combining
mented by the MD in the Supervising endorsement of success in reaching
Diabetologist component of the com- BP/Lipid goals with supervisory func-
posite visit note. tions at the Glycaemia station (now
Our expectation of insoluble prob- Station #3). In the minority of patients
lems resulting from the potentially crip- who need specic interventions
pling loss of what was originally because BP-Lipid goals are not met,
considered a critical component of the the Supervising MD can render those
model has turned out to be completely at an Optional Station #4 during a
unfounded! The keys to such a stress- truncated visit (~78 min) after the
free turnaround were vigorous team completion of the Glycaemia visit,
dialogue and evidence-based task which still leaves enough time for the
monitoring, as soon as it became clear MD to full a glycaemic supervisory
that the loss of the Pharm Ds participa- role for the next patient at Station 3.
tion was irrevocable. The critical impor- The above examples demonstrate
tance of kaizen a combination of the inherent plasticity of the model, to
dialogue and evidence is shown in our the extent that we were able to accom-
discovery that cycle time at the BP/ modate a loss of manpower with little
Lipids Station could be as low as 4 min or no disruption in work ow. That
in patients at goal for both parameters. experience further validates the adapt-
(Parenthetically, we must note here to ability of the Toyota Way to care deliv-
be revisited later that the model has ery in a multitude of chronic disease
been successful in achieving BP/Lipid states. It must be reiterated, however,
goals in ~90 % of patients after the sec- that the ability to make the BP/Lipid
ond visit, so that visit complexity is Station optional in the current congu-
drastically curtailed in 90 % of patients ration is critically dependent on the
receiving ongoing care for BP/Lipid fact that BP/Lipid goals are met in
management.) As a result, most patients 90 % of patients. This would not be
need only one session at most, two of possible in a population in whom these
goals are not met in a signicant num- SBP) in 57 patients who were seen at least three
ber of patients; in that case, the con- times in the traditional single provider clinic
guration shown in Fig. 2.3b, Second prior to redesign and followed for at least three
Iteration, would be mandatory. visits after redesign. Signicant improvements
were achieved in all three measures compared to
prior performance in the same patients who had
Performance and Results been attending the traditional single provider-
patient clinic prior to redesign.
Process-of-Care Measures (Table 2.1) Figure 2.4 shows that mean HbA1c declined by
0.6 % after redesign (7.4 % compared to 8 % for
As part of annual performance reviews at VAPHS the same patients before redesign) and that a
over the past 8 years, we are required to show greater proportion of patients achieved an HbA1c
compliance with standards of care in a random of <8.0 % (a modied care goal driven by the fact
sample of ~20 patients each year. These reviews that most of our patients are of advanced age and
show 100 % documentation in all ADA specied have multiple co-morbidities). Similarly, Fig. 2.5
domains of diabetes care (HbA1c, LDL, blood shows that mean LDL fell by 20 mg/dl (0.5
pressure, creatinine and urinary microalbumin mmol/l), with a goal LDL of <100 mg/dl (<2.6
levels, annual foot and eye exams, and aspirin mmol/l) being achieved in 90 % of patients, com-
and statin use/contraindications/alternatives). No pared to 75 % in the prior clinic, with no patient
published diabetic care model approaches, let having an LDL >130 mg/dl (3.4 mmol/l). Finally,
alone equals, this level of performance. as shown in Fig. 2.6, SBP levels fell by 11 mmHg,
and SBP <130 mmHg was achieved in almost
twice as many patients as before (63 % vs 35 %),
Performance Measures with 100 % of patients maintaining goal SBP
(Figs. 2.4, 2.5, and 2.6) <140 mmHg. Most importantly, in every instance
in which SBP was >130 mmHg, or LDL >100,
To evaluate performance, we secured IRB per- there was documentation of action taken to inten-
mission to track surrogate measures associated sify therapy, or the reason for a decision not to
with better long-term outcomes (A1c, LDL and intervene.
Table 2.1 Fullment of 12 process-of-care measures in the Individualized Multidisciplinary Team-Care Model for
Process measure Documentationb Assessmentc Interventiond
Blood pressure 100 % 100 %
A1c (POC testing) 100 % 100 %
LDL (POC testing) 100 % 100 %
Annual foot exam 100 % 100 %
Annual dilated eye exama 100 % 100 %
Annual urinary ACR 100 % 100 %
Annual creatinine 100 % 100 %
Medication reconciliation 100 %
ASA/contraindications 100 % 100 %
Lipid Rx/contraindications 100 % 100 % 100 %
HTN Rx/contraindications 100 % 100 % 100 %
Glycaemia Rx/contraindications 100 % 100 % 100 %
ACR Albumin Creatinine Ratio, ASA aspirin, HTN hypertension, POC Point of Care, Rx Treatment
a
Retinopathy (absent/present and type/severity) documented from Annual Surveillance exams
b
Documentation that each measure was either performed/resulted or due/ordered
c
Assessment of each Measure documented as normal/at goal or abnormal/not at goal
d
Intervention (therapy intensication/contraindication) documented in all patients not at goal
Fig. 2.4 (a, b) Change in A1c in 57 patients seen for 3 Multidisciplinary Team-Care Model for Delivering
visits before and after changing from a traditional Single Integrated Diabetes Care (* p < 0.05)
Patient-Provider Model to an Individualized
Fig. 2.5 (a, b) Change in LDL Cholesterol in 57 patients Multidisciplinary Team-Care Model for Delivering
seen for 3 visits before and after changing from a tradi- Integrated Diabetes Care (* p < 0.01)
tional Single Patient-Provider Model to an Individualized
Provider Patient Interactions a net of eight patient appointments of 30 min

in the Individualized each with a Nurse Practitioner in a 4 h clinic ses-
Multidisciplinary Team Care Model sion (which included direct supervisory input
(Table 2.2) from a Diabetologist), for a total of 240 min of
face-to-face patient contact. In the redesigned
In the traditional single provider clinic that clinic, 14 visits are scheduled, with three over-
existed prior to redesign, there were ten sched- books, for a net of 11 patient visits, on average,
uled appointments, including two overbooks, for totaling 60 min each (15 min with four provid-
Fig. 2.6 (a, b) Change in Systolic Blood Pressure in 57 Individualized Multidisciplinary Team-Care Model for
patients seen for 3 visits before and after changing from Delivering Integrated Diabetes Care (* p < 0.01)
a traditional single patient: provider model to an
Table 2.2 Patient-provider interactions before and after implementation of the Individualized Multidisciplinary Team-
Care Model
%
Parameter Before redesign After redesign Change
Number of providers 2 4 100 %
Daily appointment slots 10 14 40 %
Average # of patients seen/day 8 11 38 %
Scheduled clinic duration (min) 240 240
Scheduled visit duration (min) 30 60 100 %
Mean time Check-in to Depart (min) 56 63 12 %
Mean (max) wait time (min) 23 (58) 8 (19) 65 %
Average face-to-face time (min) 33 55 68 %
Integrated delivery of DSME calibrated to No Yes
need
Integrated delivery of MNT calibrated to need No Yes
Fragmented/uncoordinated ancillary care Yes No
ers). This translates to 660 min of face-to-face abilities and goals. Finally, an unexpected ben-
patient contact, which represents a 175 % increase et from time-constrained visits in the rede-
in available time for care delivery in the 4 h signed clinic is a dramatic improvement in
session. punctuality. Average patient-wait time is now 8
The inclusion of MNT and DSME in an inte- min, with a maximum of 19 min, so that 90 %
grated visit, in particular, represents a major of patients are seen within 5 min of their sched-
improvement in care that cannot be quantied. uled appointment time, compared to an average
In addition, one-on-one interactions at every wait time of 23 min previously, when only 30 %
station ensure patient-centred (individualized) were seen within 15 min of their scheduled
care delivery calibrated to each patients needs, appointment time.
Conclusions lary care providers to set and attain care

goals, based on individual patient needs;
Krumholz et al. identify eight domains of care (iv) It places equal emphasis on patient self-
that must be covered in any CDM programme management (DSME and MNT) and thera-
[20, 21]. The component interventions encom- peutic management (BP/Lipids, and
passing those domains comprise a precise yard- Glycaemia) for attaining care goals;
stick for measuring the effectiveness of a CDM (v) Care at each station is calibrated to match
programme, as follows: interventions to individual patient need,
based on proactive risk stratication;
(i) an identied population with specic (vi) It meets all process and performance
health and disease conditions; measures;
(ii) the application of evidence-based practice (vii) It incorporates feedback loops through
guidelines to treat those patients; open communication between all care pro-
(iii) collaborative practice models that include viders to not only set, achieve and maintain
physician and support-service providers; individualized care goals but also to
(iv) patient self-management education (may improve care delivery through alterations
include primary prevention, behaviour in the practice model;
modication programmes, and compli- (viii) It uses information technology to create a
ance/surveillance); templated note that mandates documenta-
(v) risk stratication to match interventions tion of all process measures at each station,
with need; and to compile notes at each station into a
(vi) process and outcomes measurement, eval- single cohesive visit note.
uation, and management (including pri-
mary prevention, behavior modication In addition, the model has proven to be
programs, and compliance/surveillance); remarkably successful in fullling all process-of-
(vii) routine reporting and feedback loops that care and performance measures. By providing
include communication with the patient, comprehensive and effective diabetes care with-
physician, health plan, and ancillary pro- out compromising individualized attention the
viders; and hallmark of patient-centred care our
(viii) appropriate use of information technology Individualized Multidisciplinary Team-care
(including use of specialized software, Model has achieved a level of success exceeding
data registries, automated decision support that in published studies of other models, where
tools, and callback systems). documentation in each of the nine ADA-identied
domains ranges from 12 % to 70 % individually
The Individualized Multidisciplinary Team- (and only 10 % for all nine domains collectively),
care Model of Diabetes Care at VA Pittsburgh, and goal for any one outcome measure (A1c, LDL
which was designed according to PPCSM or SBP) is reached in just 3560 % of patients
Principles, derived from the Toyota Production and all three in just 19 %.
System, has achieved an exceptional level of suc- One source of ongoing disappointment must,
cess in fullling all of the above criteria, as however, be mentioned before closing. It is our
follows: failure to imbue others with our enthusiasm for
changing diabetes care delivery, which means
(i) The model is designed for a specic, at- that our success has not been replicated else-
risk population (veterans with diabetes); where in the VA system. That, however, may
(ii) Goals of care are set according to evidence- reect the inertia that resists any change to a deep
based practice guidelines; rooted tradition. That is what we encountered
(iii) It delivers collaborative care through ongo- when we rst set out to redesign diabetes care
ing dialogue between physician and ancil- delivery, and our experience shows that the iner-
tia becomes particularly obdurate when faced 8. American College of Physicians: Guidelines for
Patient-Centered Medical Home (PCMH) recognition
with a paradigm-shifting change that seeks to
and accreditation programs. https://www.acponline.
replace long-held practices with those based on org/running_practice/delivery_and_payment_mod-
concepts borrowed from industry! Our experi- els/pcmh/understanding/guidelines_pcmh.pdf .
ence shows that overcoming the resistance Accessed 3 Dec 2015.
9. Davis AM, Sawyer DR, Vinci LM. The potential of
requires unshakeable belief, sustained commit-
group visits in diabetes care. Clin Diabetes.
ment, and enthusiastic buy-in from all presump- 2008;28:5862.
tive stakeholders, including (most importantly) 10. Peytremann-Bridevaux I, Burnand B. Disease man-
decision-makers responsible for allocating agement: a proposal for a new denition. Int J Integr
Care. 2009;9:e16.
resources. If all those prerequisites are mar-
11. Housden L, Wong ST, Dawes M. Effectiveness of
shalled, then it is possible to (a) improve surro- group medical visits for improving diabetes care: a
gate measures associated with improved systematic review and meta-analysis. CMAJ.
outcomes; (b) achieve 100 % performance on all 2013;185:E63544.
12. U.S. Department of Veterans Affairs. Federal benets
ADA-identied process-of-care measures; and
for veterans, dependents and survivors. Chapter 1:
(c) improve punctuality and timeliness in provid- health care benets. http://www.va.gov/opa/publica-
ing patient-centred care for diabetes. tions/benets_book/benets_chap01.asp. Accessed 2
Dec 2015.
13. Fulton EL. Physician champions awarded grant mon-
ies for research. The Bulletin of the Allegheny County
References Medical Society p 195197. April 2006. http://eliza-
bethfulton-pr.com/wp-content/uploads/2009/06/fea-
1. Centers for Disease Control and Prevention. National ture_champions.pdf. Accessed 21 Sep 2015.
diabetes statistics report: estimates of diabetes and its 14. Pittsburgh Regional Health Initiative-Perfecting
burden in the United States. Atlanta: US Department Patient Care. www.prhi.org. Accessed 21 Sep 2015.
of Health and Human Services; 2014. 15. Liker JK. The Toyota way: 14 management principles
2. Casagrande SS, Fradkin JE, Saydah SH, Rust KF, from the worlds greatest manufacturer. New York:
Cowie CC. The prevalence of meeting A1C, blood McGraw-Hill; 2004.
pressure, and LDL goals among people with diabetes, 16. Grunden N. The Pittsburgh way to efcient health-
19882010. Diabetes Care. 2013;36:22719. care. Improving patient care using Toyota based
3. Phillips LS, Branch WT, Cook CB, Doyle JP, El-Kebbi methods. New York: Productivity Press; 2008.
IM, Gallina DL, et al. Clinical inertia. Ann Intern 17. Landon BE, Hicks LS, OMalley AJ, Lieu TA, Keegan
Med. 2001;135:82534. T, McNeil BJ, et al. Improving the management of
4. Drass J, Kell S, Osborn M, Bausell B, Corcoran Jr J, chronic disease at community health centers. N Engl
Moskowitz A, et al. Diabetes care for Medicare ben- J Med. 2007;356:92134.
eciaries. Attitudes and behaviors of primary care 18. Simmons D, Yu D, Wentzel H. Changes in hospital
physicians. Diabetes Care. 1998;21:12827. admissions and inpatient tariff associated with a dia-
5. American Association of Diabetes Educators. betes integrated care initiative: preliminary ndings.
Diabetes education services reimbursement tips for J Diabetes. 2014;6:819.
primary care practice. Revised May 2009. www.dia- 19. Spear S, Bowen HK. Decoding the DNA of the Toyota
beteseducator.org/docs/default-source/practice/ production system. Harv Bus Rev. 1999;77:96106.
Entrepreneurial-DE/reimbursement-tips-2009. 20. Krumholz HM, Currie PM, Riegel B, Phillips CO,
pdf?sfvrsn=0. Accessed 21 Sep 2015. Peterson ED, Smith R, et al. A taxonomy for disease
6. Hodorowicz MA. Reimbursement for shared medical management. A scientic statement from the
appointments incorporating diabetes self-management American Heart Association Disease Management
education/training or diabetes medical nutrition ther- Taxonomy Writing Group. Circulation.
apy. Diabetes Spectr. 2012;25:8490. 2006;114:143245.
7. Porter ME, Lee TH. The strategy that will x health 21. Goetzel RZ, Ozminkowski RJ, Villagra VG. Return
care. Harv Bus Rev. https://hbr.org/2013/10/the- on investment in disease management: a review.
strategy-that-will-x-health-care. Accessed 3 Dec 2015. Health Care Financ Rev. 2005;26:119.
Approaches to Integrated
Diabetes Care: United States: San 3
Francisco
David H. Thom and Thomas Bodenheimer
Introduction our approach was developed within the larger

national movements and models for integrated
Despite the availability of effective treatment for care.
diabetes and the publication of clear guidelines In this chapter, we will briefly review the his-
for diabetes management, 41 % of persons with tory of integrated care in the US including the
diabetes in the United States do not have their major movements that, in the past two decades,
blood sugar under optimal control as defined by a have contributed to current models of integrated
glycosylated haemoglobin (HbA1c) less than 7 % care. The chapter will then review the evidence
[1]. This fact suggests that the traditional lone- for several approaches to integrated care for
doctor model for diabetes care busy doctors patients with diabetes, including emerging areas
seeing patients in rushed 15 min visits without of telemedicine/digital technology. We will con-
the help of a team is not working. Over the past clude with a description of approaches to inte-
20 years, a new approach integrated care has grated care for patients with diabetes at our
been taking hold in the United States. Non- institution and a note on the issue of paying for
physicians nurse practitioners, physician assis- integrated care within the US health-care
tants, registered nurses, pharmacists, health system.
educators, nutritionists, medical assistants, health It is worth noting that integrated care has
coaches, and community health workers have generally been used in the US to mean integra-
become engaged in the care of patients with dia- tion of physical and mental health care while the
betes. Our group has contributed to several inno- terms coordinated care and team care more
vations for providing better integrated care to closely approximate the meaning of integrated
patients with diabetes and other chronic condi- care used in this book. We will use the term inte-
tions. To understand our approach to integrated grated care in its broader sense: bringing together
diabetes care it is important to know about how multiple sources and components of care, includ-
ing primary and secondary care, services pro-
vided different professionals and non-licensed
health workers, and care delivered in the commu-
D.H. Thom (*) T. Bodenheimer nity as well as through the medical care system.
Department of Family and Community Medicine, For clarification, the term clinician as used in
University of California, San Francisco School of this chapter refers to any health professional who
Medicine and San Francisco General Hospital,
San Francisco, CA, USA has the authority to order diagnostic studies and
e-mail: [email protected]; prescribe medications and who receives re-
[email protected] imbursement for their services. In the US this

32 D.H. Thom and T. Bodenheimer
group typically includes physicians, nurse practi- tion, also provided a more supportive environ-
tioners and physician assistants. ment for the incorporation of dieticians, nurse
In addition to definition of integrated care, educators, pharmacists and mental health spe-
there is the question of how integrated care is cialists into patient care.
evaluated. The most common measure of success
of integrated care, or virtually any intervention
for patients with diabetes, is the level of glycosyl- The Chronic Care Model
ated haemoglobin (HbA1c). However, there are
additional recognized diabetes-related process In 1998 a leader in the care of patients with
and outcome measures, including screening for chronic illnesses, Dr. Ed Wagner, then at Group
complications of diabetes (retinopathy, renal Health Cooperative of Puget Sound, published an
insufficiency and foot ulcers) and controlling integrated model of patient-centred care known
diabetes-associated risk factors (typically hyper- as the Chronic Care Model (CCM) (Fig. 3.1) [6,
tension and hyperlipidemia). Moreover, patient- 7]. The CCM recognizes three spheres of care for
centred measures include general quality of life patients with chronic conditions: the health-care
[2], the diabetes distress scale [3], the symptoms system as a whole, primary care and specialty
of depression [4], and patient satisfaction [5]. practices, and the broader community in which
Additional measures may include the economic those practices are located. Fundamental to the
impact of integrated care and the extent to which Chronic Care Model is the necessity to engage
a model of integrated care is adopted and main- patients to be informed, activated partners in the
tained. Most of the studies reported in the current management of their own conditions. Numerous
chapter use HbA1c as their primary, or only, out- studies have shown the CCM to be effective in
come measure; additional measures are included improving outcomes [8], including for patients
when available. with diabetes [917]. The CCM has been the
basis for many if not most interventions to
improve integrated care.
Integrated Care for Chronic Disease
in the United States
Care Managers
Integrated care for diabetes in the US is best
understood within the broader context of efforts Multiple studies over the past 20 years have
to improve the coordination of care for patients found that care managers (generally nurses or
with chronic disease in general. The movement pharmacists) can significantly improve HbA1c
to provide integrated care is multifaceted and his- levels for patients with diabetes, particularly
torically complex, reflecting the heterogeneity when given authority to adjust treatment plans
and changing landscape of health-care delivery in using approved algorithms [18, 19]. The impor-
the US. One of the earliest models of publicly tance of care managers being able to adjust medi-
available integrated care in the US is that devel- cations is illustrated by comparing outcomes
oped by health maintenance organizations. In the from two studies at the same institution [20]. In
1940s Kaiser Permanente and Group Health the study where changes in medication required
Cooperative of Puget Sound pioneered many approval by the primary care physician, there was
aspects of integrated care including comprehen- no change in HbA1c levels over 18 months [21]
sive care for patients, shared medical records, while in the second study, where medication
partnership between specialists and GPs who changes were allowed based on an approved pro-
were salaried employees, which stood in contrast tocol, the HbA1c declined 2.1 % (from 10.1 % to
to the dominant model at the time of solo or small 8.0 %) in the intervention group compared to
independent, fee for service practices. Prospective 0.9 % (from 10.2 % to 9.3 %) in the control group,
funding (capitation) with an emphasis on preven- a significant effect [22] Care management using a
3 Approaches to Integrated Diabetes Care: United States: San Francisco 33
Fig. 3.1 The chronic care model (Used with permission from Wagner [7])
therapeutic algorithm has also been shown to be vention that included a site coordinator who
effective in improving blood pressure control for facilitated pre-visit planning with patients and a
patients with diabetes [23]. monthly review of performance with a local phy-
sician champion at each clinic [25]. At 12 months,
intervention practices had significantly greater
Multiple Interventions improvement in achieving recommended clinical
values for systolic blood pressure (SBP), HbA1c,
Interventions to improve integrated care for and low density lipoprotein (LDL) cholesterol
patients with diabetes may combine multiple than control clinics. Diabetes process measures
components. For example care management can also improved significantly more in the interven-
be combined with team care, group visits and/or tion group, including increases in the percent of
a patient activation intervention. In a randomized patients receiving annual foot examinations
controlled trial of group visits with care manage- (35 %), eye examinations (26 %); and renal test-
ment by team members including a diabetes ing (29 %). A cluster RCT of 11 primary care
nurse educator, a psychologist, a nutritionist, and practices in Pittsburgh Pennsylvania comparing a
a pharmacist, HbA1c levels declined by 1.3 % CCM-based multifaceted intervention, a provider
(from 9.5 % to 8.2 %) in the intervention subjects education intervention and usual care found a
versus 0.2 % (from 9.5 % to 9.3 %), in the control significant improvement in HbA1c in the CCM
subjects a highly significant difference [24]. The group (dropping from 7.6 % to 7.0 %) but no
intervention group also showed significant improvement in the other groups [11]. A similar
improvements in self-care practices and patients pattern was seen for improvement in lipids. The
experienced significantly lower utilization of proportion of patients self-monitoring glucose
both hospital and outpatient services. In increased from 78 % to 100 % in the CCM group,
Minnesota, a randomized controlled trial (RCT) compared to 8491 % in the education group
compared usual care to a multicomponent inter- with no change in the usual care group. Another
RCT of a Diabetes Outpatient Intensive Treatment is that care is coordinated and/or integrated
(DOIT) programme, a multiday group education across all elements of the complex health-care
and skills training experience combined with system (e.g., subspecialty care, hospitals, home
daily medical management followed by care health agencies, nursing homes) and the patients
management over 6 months, found a significant community (e.g., family, public and private
benefit on patients HbA1c [26]. However, the community-based services) [31]. To accomplish
intervention was resource intensive and depended this, the PCMH encourages the use of clinic
on multiple contacts with a highly trained nurse based care teams and care coordination [32, 33].
care manager. Teams, led by the primary care physician, are
tasked with insuring that care is coordinated
across specialties and providers with attention to
Organizational Systems medical, psychological and social needs [34].
The PCMH model includes multiple features
Better health-care delivery organization allows aimed at better integrated care, including care
for implementation and maintenance of multiple managers/coordinators, patient registries,
components to improve care, including use of improved access, regular reporting of quality
diabetes registries, care management, electronic measures, patient access to electronic health
health records, patient education, and payment information, and payment restructuring to sup-
incentives. A study of patients receiving Medicare port these activities. The model is being widely
in 90 managed care organizations (MCOs) in implemented across the US with support from
2001 found strong relationships between better payers, policy makers, patient advocacy groups,
organizational systems and all 6 measures of the and professional organizations including the
quality of care for patients with diabetes (HbA1c American Academy of Family Medicine and the
and LDL testing, screening for nephropathy and American College of Physicians. Standards for
retinal disease, and control of glucose and lipid three levels of PCMH status has been developed
levels) [27]. For example, 50 % of patients in the by the National Center for Quality Assurance
bottom quartile of MCO organizational perfor- [35]. Recognition by the NCQA can, in turn, be
mance had HbA1c levels >9.5 %, compared to tied to higher payments for services.
20 % the top quartile. A 2005 Minnesota study of The description of the PCMH in shown in
41 medical groups, ranging in size from less than Table 3.1 is consistent with the most common
10 to more than 200 physicians, created a scoring versions of the model [2, 31, 36, 37], but is not
system based on the presence or absence of 53 meant to be complete or definitive as key charac-
systems and process items within five domains: teristics and components of the PCMH have been
general health system, delivery system redesign, described in slightly different ways at different
clinician formation system, decision support for time by different groups. Some key characteris-
clinicians and self-management support [28]. tics overlap or reinforce each other: for example,
Significant correlations ranging from 0.39 to 0.46 some level of coordinated care is necessary for
were found between the sum of domain scores care to be comprehensive and accessibility is
and testing of HbA1c, LDL and blood pressure, important for patient safety. The PCMH model
and for control of diabetes (HbA1c 8 %) and also includes each patient having a personal phy-
cholesterol (LDL <130 mg/dL (3.4 mmol/l)). sician or other clinician who is responsible for
their care. Use of health information technology,
including electronic health records and health
The Patient-Centered Medical Home information exchanges, is considered necessary
to reach the full potential of the medical home.
The CCM has been a major impetus to the The success of the medical home model depends
Patient-Centered Medical Home (PCMH) move- in large part on realignment of monetary incen-
ment [29, 30]. (A primary principle of the PCMH tives away from the number of patient visits with
Table 3.1 Key characteristics of the Patient-Centered Medical Home

Key characteristic Description Components/examples
Comprehensive Provides for all the patients Team care
health-care needs or taking Practice panels
responsibility for appropriately Primary care
arranging care with other qualified
professionals
Patient-centred Recognizes and respects patients Shared decision-making
individual needs, values, preferences Patient and family engagement
and culture and partners with Patient self-management support
patients and their families
Continuous relationship with a personal
physician or other clinician
Coordinated/integrated Care is coordinated and/or integrated Health information technology, including
across all elements of the health-care health information exchanges
system, including subspecialty care, Care coordinators
hospitals, nursing homes, home Referral tracking
health agencies, and community-
based services
Accessible Convenient access to care by in Enhanced clinic hours including evenings
person visits and by other modes of and weekends
communication 24 h telephone access to medical advice
from care team
E-mail communication
Patient portals
High quality and safety Provides care that meets nationally Evidence-based medicine and clinical
recognized measures of quality; decision support tools
engages in continuous quality Health information technology support
improvement; has protocols in place Measurement of health status and quality
protect patients from medical errors indicators
Electronic prescribing
Patients participate in quality improvement
activities at the practice level
clinicians, to targeted processes and outcomes of to hospital or nursing home), linking patients
the five key medical home characteristics. with community resources to facilitate referrals
It has been recognized that the PCMH exists and respond to social service needs, integrating
within a medical neighbourhood which refers mental health and specialty care into care deliv-
to health-care entities beyond the primary care ery through collocation or referral agreements,
clinic (e.g., area hospitals, specialists, home care tracking patients receiving outside services and
agencies) as well as community resources [29]. following up on patients soon after emergency
See Fig. 3.2. PCMHs are expected to assume room visit or hospital discharge [33]. Performance
accountability for coordinating their patients measures of the coordinated care function of the
care by establishing relationship and connectivity PCMH include the Care Coordination Process
with members of the medical neighbourhood Measures [38], which is used for PCMH recogni-
such as specialists, hospitals and community tion, the PCMH-A [39], a practice self-assessment
agencies. This role of the PCMH has been further measure, and the Key Activities for Care
elucidated and expanded in the Care Coordination which measures specific actions of
Coordination Model (Fig. 3.3) which includes key importance for care coordination, such as
information sharing, care coordination for refer- developing information sharing agreements or
rals and transitions in care (e.g., from community tracking systems [33].
Fig. 3.2 The medical neighbourhood: care coordination Institute for Healthcare Innovation. Supported by The
model (From Reducing Care Fragmentation: A Toolkit for Commonwealth Fund. April 2011)
Coordinating Care. Prepared by Group Healths MacColl
Fig. 3.3 Ten building blocks of high-performing primary care

Because integration of care is fundamental to Implementation of the team model requires shar-
the PCMH, the PCMH has been touted as a model ing of tasks and responsibilities that have tradi-
to provide high-quality integrated care for diabetes tionally been the purview of the physician. The
[34, 40]. A review of PCMH demonstration proj- term team care has traditionally been used to
ect sites, ranging in size from a single clinic to indicate a group of health-care professionals such
1200 practices, which have provided outcomes for as physicians, nurses, pharmacists, or social
patients with diabetes, reported improvement in workers, who work together in caring for a group
diabetes-related care processes and outcomes at of patients. In a 2006 systematic review of 66 tri-
five of the sites and improvement in quality of care als testing 11 strategies for improving glycaemic
from chronic conditions (which included diabetes) control for patients with diabetes, only team care
in the other three sites [34]. Care coordination was and case management showed a significant
a key feature for all sites. In addition, three sites impact on reducing HbA1c levels [18]. Four
had performance-based payments or incentives types of team care were defined (1) the patient
[4144], two had capitated care [44, 45], one pro- seeing a health professional in addition to their
vided support for a care coordinator [46] and two primary care clinician; (2) care from a multidisci-
had no change in payment structure [47, 48]. plinary team of professionals in addition to the
A recently reported study of the Pennsylvania primary clinician; (3) expansion of an existing
Chronic Care Initiative examined the impact of a professional role (e.g., nurse or pharmacist play-
3 year multifaceted intervention that included ing a more active role in patient monitoring or
practice coaching, learning collaboratives, dis- adjusting medication regimens) or (4) shared
ease registries, payment for care managers and care between specialists and primary care clini-
incentives for practice transformation to meet cians. Based on 26 randomized controlled trials
National Committee for Quality Assurance (the majority conducted in the US), team care
(NCQA) standards as medical homes [42]. Data interventions were associated with a mean
from the 27 small primary care practices partici- decline in HbA1c of 0.33 % independent of any
pating in the intervention were compared to 29 co-interventions. Of the four types of team care
similar practices in the area. At 3 years, partici- intervention, multidisciplinary care, expansion of
pating practices significantly outperformed existing professional roles and shared care all
comparison practices on all four screening mea- had nearly identical effect sizes, while simply
sures related to diabetes: HbA1c testing, LDL having the patient see an additional health profes-
testing, nephropathy monitoring and annual eye sional (which arguably does not constitute team
examinations. However, only the proportion of care) had no impact. Studies of patients with
patients receiving nephropathy monitoring actu- baseline HbA1c levels of 8.0 % or higher reported
ally increased in the participating practices, mov- greater effect sizes than studies with patients
ing from 78 % to 86 %. The percent of patients whose baseline HbA1c levels were less than
receiving the other three screening measures 8.0 %.
actually decreased slightly in the participating Perhaps the most widely studied addition to
clinics, but decreased markedly in the compari- the traditional physician care model of patients
son clinics, creating a significant difference in with diabetes is a pharmacist. In randomized con-
favour of the participating clinics. trolled trials conducted in the US, patients ran-
domized to receive diabetes education,
medication counselling and adjustment, and
The Patient Care Team instructions by a pharmacist have experienced
significantly greater reductions in their HbA1c
The use of patient care teams has been promoted then control patients, as well as improved process
as a means to implement integrated care both in measures and better control of other cardiovascu-
parallel with and as part of the PCMH [32, 49]. lar disease risk factors [22, 5052].
Expansion of Care Teams to Include workers have also been shown to reduce HbA1c
Non-licensed Health Workers compared to nurse care managers [60]. Project
Dulce, a programme targeting Latino immigrants
Part of the team care model promoted by the found that peer-led education by community
PCMH is that all members of the team work at health workers (promotoras) significantly low-
the top of their licence or certification through ered HbA1c, diastolic blood pressure, and lipids
additional instruction or training and wider use of compared to usual care [61, 62]. In 2013, diabe-
algorithms. For example, medical assistants may tes was identified as a top priority area for com-
assume such responsibilities as checking patient munity health workers [63].
medicines (medication reconciliation), perform-
ing diabetic foot exams, screening for depres-
sion, providing education, giving navigation Incorporating Mental Health
assistance for access to resources within the into the Patient Care Team
health system, and supporting patient activation,
engagement and self-management [34, 53]. The association between depression and diabetes
Another model using non-licensed health is well established, as are the personal and social
workers is the health coaching model. Health stresses for coping with diabetes as a serious
coaching provides in-depth self-management chronic disease. However, few studies have
support by assisting patients to gain the knowl- examined the impact of a mental health expert
edge, skills, and confidence to become informed, into the patient care team. In a clustered RCT of
active participants in their care. Health coaching 387 patients with a PHQ-9 score 10 and either
has several central components: setting agendas diabetes or heart disease or both, a multi-method
to ensure that patients concerns are discussed in intervention that included up to eight therapy ses-
medical visits; ensuring that patients know your sions and education of the primary care practitio-
numbers, e.g., know their HbA1c level and goal; ners found a significant reduction in symptoms of
closing the loop which means checking depression [64]. HbA1c was not assessed. Team-
patients understanding of the care plan decided based interventions can improve quality of life
on in the medical visit by having the patient and reduce depression scores for patients with
repeat the care plan themselves; and action plans diabetes even without including mental health
which are behaviour-change agreements includ- expertise [65]. One of the barriers to incorporating
ing medication adherence made between health mental health care into the patient care team is
coaches and patients [54, 55]. Action plans have the difficulty of collocating mental health provid-
been found in a RCT to improve HbA1c levels ers with clinicians. In fact, one qualitative study
more than traditional patient education [56]. found that while patients appreciate coordination
In randomized controlled trial, medical assis- and communication between their clinicians and
tant health coaching significantly reduced patient their mental health providers, they actually prefer
HbA1c levels from 9.8 % to 8.6 % compared to a to avoid collocation of the two services [66].
reduction of 9.99.4 % for usual care (net differ-
ence of 0.6 %) [57]. Training peer supporters can
also be effective. In a randomized controlled trial Engaging the Patient as Part
study by Heisler et al., reciprocal peer support of the Integrated Care Team
was more effective than being assigned a nurse
care manager in lowering HbA1c levels for male Fundamental to the Chronic Care Model is the
veterans with diabetes [58]. In another RCT, peer need for patients to be informed, activated part-
support significantly reduced mean HbA1c by ners in the management of their own conditions.
1.1 % (from 10.1 % to 9.0 %) in the coached In recent years, the concept of integrated care has
group compared to 0.3 % (from 9.9 % to 9.6 %) in been expanded to include the patient as a key
the usual care group [59]. Community health member of the care team [67, 68]. While patient
engagement is most commonly thought of as building blocks (Fig. 3.3) are characteristics
occurring at the level of direct patient care, found in primary care practices with excellent
engagement can also occur at organizational and patient access, continuity of care, use of data to
policy [69]. drive improvement, well-functioning teams, and
Two areas closely related to patient engage- an orientation towards population-wide care.
ment are patient activation and shared decision- Blocks in the first row are practice characteristics
making. Patient activation, while sometimes used that support the higher blocks. Block 1, engaged
synonymously with patient engagement, refers to leadership, refers to leadership at all levels of the
increasing patients knowledge, confidence, and/ organization, including clinicians, nursing, medi-
or skills for disease self-management, while cal staff and in some clinics patients as well.
engagement generally refers to actual patient Data-driven improvement (block 2) requires sys-
behaviours [70]. None the less, patient engage- tems that can track and feedback clinical (e.g.,
ment is most often assessed using the widely cancer screening and diabetes management),
validated Patient Activation Measure [71]. operational (continuity of care and access), and
Patient activation has been associated with higher patients experience metrics. Empanelment
quality of care for chronic conditions generally (block 3) means linking each patient to a care
and to have a modest association with lower costs team and a primary care clinician. Many exem-
[72]. A recent meta-analysis of interventions to plar practices have created teams (block 4) with
increase the activation of patients with type 2 dia- well-trained non-clinicians who add primary care
betes found modest but significant net improve- capacity by sharing the care [82, 83]. The
ment in HbA1c (0.4 %), SBP (2.2 mmHg), LDL patient-team partnership (block 5) recognizes the
and weight (2.3 lb) [73]. expertise that patients bring to the medical
Shared decision-making has various defini- encounter so that patients are not told what to do
tions, but at a minimum seems to require that the but are engaged in shared decision-making that
patient and physician share information and pref- respects their personal goals. Block 6, population
erences and engage in a process of coming to a management, includes proactively addressing
mutually agreeable decision [74]. Ethically, gaps in care, such as screening LDL and foot
shared decision-making respects a patients examinations for patients with diabetes. Health
autonomy and right to make an informed choice; coaching and complex care management are also
pragmatically, shared decision-making allows the included in this block. Blocks 7, 8 and 9 are
patient to contribute information and preferences self-explanatory and are considered key charac-
that will presumably result in a better decision teristics of primary care. The last block, template
and greater patient participation in implementing of the future, refers to a practice which supple-
the decision [75]. Multiple measures of shared ments one-on-one clinician visits with group vis-
decision-making have been developed [76, 77] its, telephone visits, patient-portal e-visits, and
and interventions to increase shared decision- visits to non-clinician team members, operating
making have been tested [78, 79]. Shared deci- in a payment system that supports this model,
sion-making can also include decision-making rather than simply paying for in-person clinician
between the patient and the care team [80]. visits. For patients with diabetes, such high-per-
forming practices closely follow a number of
diabetes-related processes of care and outcome
Practice Transformation and the Ten metrics drilled down to the level of individual cli-
Building Blocks nicians and teams, and organize diabetes care as
a team responsibility rather than the sole prov-
A PCMH-related model that may be more trans- ince of the clinician. Patients with diabetes may
formative of diabetes care is the Building Blocks have access to RN or pharmacist care managers,
of High-Performing Primary Care [81, 82]. The group diabetes visits, and health coaches.
Telemedicine and Digital Health Another promising use of digital technology

Tools: The Future of Integrated is to enhance communication between primary
Care for Patients with Diabetes? care clinicians and specialist colleagues to
improve patient care for a variety of conditions,
The rapidly increasing availability of telecom- including diabetes. One such programme,
munication options has recently opened up new e-referral (electronic referral), was developed
opportunities for providing more integrated care. for use at San Francisco General Hospital to
Telemedicine, defined as medical activity using facilitate communication between primary care
some form of telecommunication, includes tele- clinicians and specialists in the context of making
phone, internet and smartphone-based pro- an electronic referral or consultation [90, 91].
grammes [84]. A recent systematic review and E-referral documents the pertinent clinical his-
meta-analysis of 13 RCTs of telemedicine for tory and reason for the consultation. A designated
patients with diabetes reported that telemedicine- specialist reviewer then responds, perhaps
based interventions resulted in a statistically sig- requesting further information, suggesting addi-
nificant decline in mean HbA1c (difference = tional tests or management strategies, redirecting
0.44 %) and LDL (difference = 6.6 mg/dL (0.17 the referral if indicated, or simply making an
mmol/l)) levels compared to controls, with no appointment for the patient. E-referral been
impact on blood pressure [85]. Specific examples shown to increase access to care and increase the
of the application of telemedicine to the manage- appropriateness of referrals. It can also improve
ment of diabetes are described below. the primary care clinicians capacity for complex
Automated telephone self-management sup- decision-making and promote a dialogue between
port (ATSM) uses interactive telephone technol- primary care clinicians and specialists [92]. The
ogy to provide surveillance and patient education effectiveness of e-referral has been best docu-
which can be combined with nurse care manage- mented for rheumatology [93] and gastroenterol-
ment [86]. In one study, English-, Spanish-, and ogy [94] and is currently being used for referrals
Cantonese-speaking patients with diabetes seen to endocrinologists and diabetes nurse educators
in community (safety-net) clinics in San at San Francisco General Hospital. Similar sys-
Francisco were randomized to receive usual care, tems are being implemented in Los Angeles,
interactive weekly automated telephone self- Boston, and other sites.
management support with nurse follow-up A variety of internet-based programmes have
(ATSM), or monthly group medical visits with been used to improve patient self-care for diabe-
physician and health educator facilitation (GMV). tes [95] including increasing blood glucose mon-
No significant difference in HbA1c levels was itoring frequency [96], reducing HbA1c levels
found [87, 88]. Only the ATSM group showed [97104] and increasing self-reported physical
improvements in reported interpersonal pro- activity [99, 102]. An early example from Seattle,
cesses of care. Both intervention arms showed Washington provided patients with type 2 diabe-
significant improvements in self-management tes access to their electronic health records, a
behaviours versus the usual care arm with gains secure e-mail with providers, diabetes education,
being greater for the ATSM group than for the feedback on blood glucose readings, and an inter-
GMV group. The ATSM group also had fewer active tool to track exercise, diet and medication;
bed days per month than the usual care and GMV intervention patients demonstrated a greater
groups and less interference with daily activities decline in HbA1c (0.7 %) compared to patients
than the usual care group. A follow-up study randomized to receive usual care alone [103].
using waitlist controls found that, compared with Internet-based programmes may also be used to
waitlisted patients, immediate intervention par- augment care management [104] and to provide
ticipants had significantly greater 6-month behavioural support [105]. Similar programmes
improvement in overall diabetes self-care behav- can include a mechanism to upload blood glu-
iours [89]. cose values measured by the patient, of particular
importance in type 1 diabetes [106]. Internet-based able [110]. Essentially all allow short messaging
patient engagement may be effective by itself, from a care provider and entry and retrieval of
even without a care manager [98, 107]. data such as blood glucose or blood pressure.
A burgeoning number of mobile phone appli- Many allow two-way communication between
cations (apps) have been created to improve the patient and patient care team, can provide tar-
patient self-management of diabetes and commu- geted educational materials, allow access to per-
nication between the patient and the health-care sonal health information, and can be used to
team. The simplest technology allows for one adjust medications. Examples include
way messaging, e.g., for appointment reminders CarePlanManager (CircleLink Health,
or to take medications. In 2011 Liang et al. Stamford, CT, USA) (https://www.careplanman-
reported a meta-analysis of 22 trials investigating ager.com/), MedAdherence (Norwalk, CT, USA)
mobile phone-based interventions for diabetes (www.medadherence.com), CareMessage (San
[108]. Half the trials were RCTs; approximately Francisco, CA, USA) (http://caremessage.org/),
half were of patients with type 1, and half with Twine (Twine Health, Cambridge, MA, USA)
type 2, diabetes. All apps used short messages to (http://www.twinehealth.com/), WellDoc
support self-monitoring of blood glucose, educa- DiabetesManager (Baltimore, MD, USA) (http://
tion, diet, exercise and medication adjustment. www.welldoc.com) and CareSmarts (Larkspur,
Eight studies also included a component of inter- CA, USA) (https://www.mhealth-solutions.com)
net support and 14 studies featured downloading to name a few. WellDoc DiabetesManager pro-
or entering daily blood glucose values. The meta- vides three components (1) real-time educational
analysis found significant reductions in HbA1c and behavioural messaging to patients based on
of 0.8 % for patients with type 2 diabetes and blood glucose values, medications and lifestyle
0.3 % for patients with type 1 diabetes compared behaviours plus summary data; (2) a portal where
to usual care controls. Studies with both mobile patients and their physicians can access the data;
phone and Internet interventions showed greater and (3) a data analysis and evidence-based
reduction in HbA1c than the studies with only a treatment recommendations for physicians
mobile phone intervention and studies with daily [110113]. A cluster RCT of 26 practices (total of
intervention reported greater reduction in HbA1c 163 patients) compared usual care to three
than those with only weekly intervention; how- versions of the intervention: component 1 only,
ever, these differences did not reach statistical and components 1 and 2, to the full version with
significance. all three components [113]. At 12 months, the
A more recent study enrolled 65 people with mean HbA1c had declined 0.7 % (from 9.2 % to
HbA1c >8.0 % who were established (>6 months) 8.5 %) in usual care group, compared to 1.6 %
patients in the endocrinology clinics of the Walter (from in 9.3 to 7.7) in the component 1 group, by
Reed Health Care System. Participants were ran- 1.2 % (from 9.0 to 7.9) in the component and two
domized to receive usual care or self-care group, and by 1.9 % (from 9.9 % to 7.9 %) in the
video messages from their diabetes nurse practi- full version, The decline in HbA1c was signifi-
tioner [109]. Video messages were sent daily to cantly greater in the full version group compared
cell phones of study participants. Participants to usual care (difference of 1.2 % P < 0.001) but
who received the messages had small but signifi- not significantly different compared to the other
cant improvement in their HbA1c levels com- two intervention groups. There were no appre-
pared to those who received usual care (0.2 % ciable differences between groups for patient-
difference over 12 months). Haemoglobin A1c reported diabetes distress, depression, diabetes
decline was greatest among participants who symptoms, or blood pressure and lipid levels.
received video messages and viewed more than Another mobile phone-based application,
ten messages a month (0.6 % difference). CareSmarts, developed at University of Chicago
More sophisticated mobile phone applications Medicine, provides automated text messages
for helping patients are now commercially avail- to provide patient education, support patient
self-management and facilitate communication screening for early renal disease) and intermedi-
between the patients and members of their health- ate outcomes including HbA1c and blood pres-
care team [110, 114]. Patient responses to self- sure [115].
assessment questions are used to monitor patient In contrast, several studies using integrated
care and, if outside an established range, can trig- care models for patients with diabetes living in
ger a member of the care team to contact the rural areas have shown improvement in out-
patient. In a prospective study, patients with dia- comes. In North Carolina, a team-based interven-
betes in the University of Chicagos employee tion used a nurse, pharmacist and dietitian to
health plan were offered the programme; 74 provide point-of-care education, coaching and
enrolled in the programme and 274 did not [110]. medication intensification for rural African
Those enrolled had a higher baseline HbA1c (7.9 Americans with diabetes [116]. Patients receiv-
vs 7.4 but were similar in demographic character- ing the intervention had a significantly greater
istics to those not enrolled. Enrolled patients drop in HbA1c compared to usual care patients
experienced significant improvements in HbA1c (0.5 % vs 0.1 %). A significantly higher propor-
(-0.4 %, care utilization, and self-reported healthy tion of intervention patients achieved an HbA1c
eating, glucose monitoring and diabetes medica- level <7.5 % (68 % vs 59 %) and a systolic blood
tion adherence at 6 months, but outcomes for pressure <140 mmHg (69 % vs 57 %). A recently
patients not enrolled were not reported in suffi- published multidisciplinary team care manage-
cient detail to provide a meaningful comparison. ment intervention for 3373 American Indians/
The website www.diabetesmine.com provides an Alaska Natives found modest but significant
updated list and evaluations of mobile apps for decreases in HbA1c (0.2 %) after 1 year as well
type 1 and type 2 diabetes. Wider application of as a significant reduction in LDL cholesterol
mobile or smart phone applications in the US [117].
will require compliance with requirements for In the IDEATel study, 1665 Medicare recipi-
handling personal medical information. It ents with diabetes, residing in federally desig-
remains to be seen what proportion of the popula- nated medically underserved areas of rural
tion will adopt such applications and how these New York State, were randomized to receive a
applications will be integrated into their medical home telemedicine unit with nurse care manage-
care. ment versus usual care [118]. Patients in the tele-
medicine group received a web camera that
allowed video conferencing with nurse care
Integrated Care for the Urban managers, a home glucose metre and blood pres-
and Rural Underserved sure cuff that connected to the internet, access to
their own clinical data and to an educational web-
While most studies have found a positive associa- page. Telemedicine achieved modest though sta-
tion between characteristics of a Patient-Centered tistically significant net overall reductions over 5
Medical Home and the quality of care for patients years of follow-up for the primary endpoints of
with diabetes, it is not clear that this relationship HbA1c (0.3 %), LDL (3.8 mg/dl), systolic and
holds among community (safety net) clinics diastolic blood pressure (4.3 and 2.6 mmHg,
already meeting basic standards of a medical respectively). In an observational study of tele-
home. A study of 40 community health clinics in medicine using touchscreen internet technology
Los Angeles, all of whom met the NCQA stan- and home monitoring of blood glucose and blood
dards to qualify as a medical home, found that pressure in 109 rural patients, significant
scoring higher on a widely used medical home improvements for baseline were seen for HbA1c
assessment instrument, the NCQA Physician which decreased from a mean of 9.7 % to 7.8 %,
Practice ConnectionsPatient-Centered Medical for systolic blood pressure which dropped from
Home tool, was not associated with better diabe- 131 to 123 mmHg, and for LDL which moved
tes performance on processes of care (e.g., from 103 to 93 mg/dl [119].
Project Dulce is a programme, originating in ors. Many of our patients have current or past
San Diego, which has provided integrated care to problems with drug abuse, alcoholism, and men-
low-income people with diabetes since 1997. tal health conditions. As in many under-resourced
Project Dulce is a mixed community and primary safety net systems in the US, improvement in
care-based model, combining RN care manage- integrated care for our patients with diabetes and
ment and peer-led education classes. People with other chronic diseases has taken place in a some-
diabetes are referred by primary care providers what piece-meal and opportunistic fashion, in
(PCPs) from many community health centres. contrast with the more comprehensive forms of
Patients are linked to a RN care manager in regu- primary care transformation promoted by the
lar contact with the PCP and are encouraged to PCMH movement. Leaders of San Franciscos
join a group self-management support class public safety-net system, tasked with providing
taught by trained community health workers care for low-income and uninsured patients, work
(CHWs) known as promotoras. The classes, in closely with researchers and clinical leaders in
English, Spanish, Vietnamese, and other lan- Family and Community Medicine, General
guages cover the basic concepts of diabetes, Internal Medicine, and other disciplines in the
healthy eating, exercise, and medications. The schools of Medicine, Nursing and Pharmacy at
promotoras often have diabetes themselves and the University of California, San Francisco as
must complete a 4-month training and mentoring well as with health-care clinicians, staff and
programme. The peer-led classes take place in patients. The result has been the development or
the community, while visits with the registered adoption of several models for integrated care
nurses (RNs) are performed in the patients clinic described earlier in this chapter, including auto-
so that the RN can interact with the PCP. RNs mated telephone support [87, 88], electronic
travel from one clinic to another on a regular referrals [9092], and patient care teams [53] that
schedule. RNs order and review laboratory stud- include pharmacists, nurses, dieticians and men-
ies, do foot exams, refer for eye exams, and man- tal health providers. The system has implemented
age medications in consultation with the PCP. A registries of patients with diabetes [120], health
study conducted by Project Dulce found that par- coaching [54, 55, 57, 59] and practice transfor-
ticipants in the programme had significant mation using the 10 Building Blocks of High-
improvements in HbA1c, LDL cholesterol, and Performing Primary Care model [82, 83]. Several
diastolic blood pressure compared with controls clinics have mental health professionals on-site
[62]. A further study looked at patients who in the form of social workers and psychologists,
attended the peer-led classes but did not engage allowing for warm handoffs for patients with dia-
in RN care management and found that they also betes and depression or other mental health
had a significant improvement in HbA1c com- issues. Processes have been worked out to allow
pared with controls [61]. health coaches access to patients electronic
health records (with patients permission) and the
ability to communicate information to clinicians
Our Approach to Integrated using the electronic medical record. A patient
Diabetes Care portal which will initially allow patients access to
laboratory results and appointments is being
The authors primary research and practice site is rolled out, with plans to expand allowing secure
the system of community clinics and hospital electronic communication between patients and
outpatient clinics operated by the City and their care team. We are also looking at ways to
County of San Francisco. This safety net sys- include patients in electronic communications
tem serves a highly diverse population of low- between primary care clinicians and subspecial-
income patients, many of whom face challenges ists and to increase patient involvement at the
such as being non-English speaking and of low organizational level of care. The potential of
literacy, in addition to economic and social stress- many of these approaches has not been fully
realized and work is continuing towards the Another approach to estimating cost-
goal of providing better integrated care for all of effectiveness is to use models to estimate future
our patients. cost savings from reduction in HbA1c levels and
better management of associated cardiovascular
disease risk factors such as hypertension and
Paying for Integrated Diabetes Care hyperlipidemia [128]. Adopting such a model of
cost-effectiveness assumes that gains in the short
Integrated care for patients for diabetes has run can be maintained over time, and that the
become increasingly common in the US due to party incurring the short term costs will also ben-
consolidation of health care into larger organiza- efit from any long-term savings.
tional units, recognition of the importance of An important question in assessing the cost of
integrated care, and efforts to organize, support integrated care is whether it needs to be cost-
and reward clinical care that is more integrated. saving or cost-neutral to be adopted, or is it
These developments have been spearheaded by enough to increase quality-adjusted life years
the PCMH movement and further supported by (QALYs) at a reasonable cost (usually pegged
provisions of the Affordable Care Act of 2010. at between $30,000 and $60,000 per QALY
The PCMH has promoted integrated care by saved). Most integrated care programmes for
advocating for changes in payment structure to patients with diabetes that have been evaluated
move beyond paying only for face-to-face clini- for cost-effectiveness would meet this more lib-
cian visits to include payment for care coordina- eral criterion [124, 126].
tion services, improved access and In practice, integrated care programmes for
communication, and adoption of health informa- patients with diabetes are often part of general-
tion technology [37]. Financial facilitators of ized programmes of care for patients with other
integrated care include support for development chronic medical conditions, making the alloca-
and adoption of health information technology, tion of costs and savings with respect to inte-
most notably as part of the 2010 Affordable Care grated care for diabetes difficult to estimate. At
Act, and tying higher payment to NCQA certifi- this point, integrated care for patients with diabe-
cation as a PCMH. In 2015, the Centers for tes appears to be a widely accepted goal. The
Medicare and Medicaid Services, which admin- question becomes: which model of integrated
isters the federal programmes for health care to care is most effective at reasonable cost?
elderly and low-income Americans, introduced Answering this question depends both on what
the first nonvisit-based payment for chronic costs are included and what outcomes are mea-
care management [121, 122]. sured; the answers may vary among different
The cost-effectiveness of integrated care for patient populations and different care systems.
patients with diabetes depends on the model of
integrated care used, the system in which it is
used, and the time-horizon chosen [123]. Models Summary and Conclusions
of cost benefit for using health coaching interven-
tions for patients with poorly controlled diabetes An unacceptably large proportion of patients
have generally found a benefit in reducing HbA1c with diabetes in the US do not meet national
levels, but at the cost of paying for the added cost guidelines for control of their diabetes and related
of health coaching which is not offset in the short conditions. The importance of integrated, patient-
term by savings from emergency department vis- centred care has long been recognized and was
its and hospitalizations [124126]. Costs for the basis for the development of the Chronic Care
medication and numbers of primary care visits Model in the mid-1990s. The CCM in turn has
may increase as health coaches improve patient provided a basis for development and testing of
adherence to medications and re-engage patients new models of care, including care management,
in the health-care system [123, 127]. care teams, and the Patient-Centered Medical
Home (PCMH). The PCMH has been seen as a 2. El Achhab Y, Nejjari C, Chikri M, Lyoussi
B. Disease-specific health-related quality of life
strong model for integration in general and for
instruments among adults diabetic: a systematic
the care of patients with diabetes in particular. In review. Diabetes Res Clin Pract.
concert with the PCMH movement, a substantial 2008;80(2):17184.
amount of work has taken place to make inte- 3. Polonsky WH, Fisher L, Earles J, Dudl RJ, Lees J,
Mullan J, Jackson RA. Assessing psychosocial dis-
grated care more patient-centred, through patient
tress in diabetes. Diabetes Care.
activation, patient education and engagement, 2005;28(3):62631.
shared decision-making, and self-management 4. Kroenke K, Spitzer RL, Williams JBW. The PHQ-9.
support including health coaching. Research in J Gen Intern Med. 2001;16(9):60613.
5. Narayan KM, Gregg EW, Fagot-Campagna A, Gary
these areas has generally supported the view that
TL, Saaddine JB, Parker C, Imperatore G, Valdez R,
including patients with diabetes in their care can Beckles G, Engelgau MM. Relationship between
improve patient care processes and outcomes, quality of diabetes care and patient satisfaction.
including better control of blood glucose, blood J Natl Med Assoc. 2003;95(1):6470.
6. Wagner EH, Austin BT, Von Korff M. Organizing
pressure and LDL cholesterol. Such interven-
care for patients with chronic illness. Milbank Q.
tions do not appear to reduce short-term costs, 1996;74(4):51144.
but most studies have found them to be cost- 7. Wagner EH. Chronic disease management: what will
effective in terms of quality-adjusted life years it take to improve care for chronic illness? Eff Clin
Pract. 1998;1:24.
saved, particularly if the benefits are projected
8. Coleman K, Austin BT, Brach C, Wagner
over time to include late complications of poorly EH. Evidence on the chronic care model in the new
controlled diabetes. millennium. Health Aff (Millwood).
Many PCMH-based models now incorporate 2009;28(1):7585.
9. Parchman ML, et al. Risk of coronary artery disease
multiple intervention components with ongoing
in type 2 diabetes and the delivery of care consistent
evaluation. Realizing the potential of these with the chronic care model in primary care settings:
approaches to integrated care is particularly chal- a STARNet study. Med Care. 2007;45:112934.
lenging for providing care to patients with diabe- 10. DiPiero A, Dorr DA, Kelso C, Bowen JL. Integrating
systematic chronic care for diabetes into an aca-
tes in urban and rural underserved populations.
demic general internal medicine resident-faculty
Our work in San Franciscos safety-net system practice. J Gen Intern Med. 2008;23(11):174956.
illustrates those challenges and has led to new 11. Piatt GA, Orchard TJ, Emerson S, Simmons D,
approaches for providing integrated care. The Songer TJ, Brooks MM, Korytkowski M, Siminerio
LM, Ahmad U, Zgibor JC. Translating the chronic
role of telehealth in providing integrated care will
care model into the community: results from a
continue to grow. How digital programmes will randomized controlled trial of a multifaceted diabe-
be incorporated into current models of integrated tes care intervention. Diabetes Care.
care, and the balance between digital programmes 2006;29:8117.
12. McCulloch DK, Price MJ, Hindmarsh M, Wagner
and personal relationships, will likely be an
EH. A population-based approach to diabetes man-
important area for future research. While the agement in a primary care setting: early results and
growing support for integrated care has included lessons learned. Eff Clin Pract. 1998;1(1):1222.
financial incentives from both public and private 13. Smith SA, Shah ND, Bryant SC, Christianson TJ,
Bjornsen SS, Giesler PD, et al. Chronic care model
payers, the optimal models of integrated care and
and shared care in diabetes: randomized trial of an
their cost-effectiveness in both the short- and electronic decision support system. Mayo Clin Proc.
long-term remains an area of active investigation. 2008;83(7):74757.
14. Stock S, Pitcavage JM, Simic D, Altin S, Graf C,
Feng W, Graf TR. Chronic care model strategies in
the United States and Germany deliver patient-
References centered, high-quality diabetes care. Health Aff
(Millwood). 2014;33(9):15408.
1. Selvin E, Parrinello CM, Sacks DB, Coresh J. Trends 15. Stellefson M, Dipnarine K, Stopka C. The chronic
in prevalence and control of diabetes in the United care model and diabetes management in US primary
States, 19881994 and 19992010. Ann Intern Med. care settings: a systematic review. Prev Chronic Dis.
2014;160(8):51725. 2013;10:E26.
16. Sunaert P, Bastiaens H, Nobels F, Feyen L, Verbeke 29. Nutting PA, Crabtree BF, Miller WL, Stange KC,
G, Vermeire E, De Maeseneer J, Willems S, De Stewart E, Jan C. Transforming physician practices
Sutter A. Effectiveness of the introduction of a to patient-centered medical homes: lessons from the
chronic care model-based program for type 2 diabe- national demonstration project. Health Aff
tes in Belgium. BMC Health Serv Res. 2010;10:207. (Millwood). 2011;30:43945.
17. Nutting PA, Dickinson WP, Dickinson LM, Nelson 30. American College of Physicians. The advanced
CC, King DK, Crabtree BF, Glasgow RE. Use of medical home: a patient-centered, physician-guided
chronic care model elements is associated with model of health care. Philadelphia: American
higher-quality care for diabetes. Ann Fam Med. College of Physicians; 2005. Position Paper. [cited
2007;5(1):1420. 2015 Jun 16]. Available from: https://www.acpon-
18. Shojania KG, Ranji SR, McDonald KM, et al. line.org/advocacy/current_policy_papers/assets/
Effects of quality improvement strategies for type 2 adv_med.pdf.
diabetes on glycemic control: a meta-regression 31. American Academy of Family Physicians (AAFP),
analysis. JAMA. 2006;296:42740. American Academy of Pediatrics (AAP), American
19. Pimouguet C, Le Goff M, Thibaut R, Dartigues JF, College of Physicians (ACP), American Osteopathic
Helmer C. Effectiveness of disease-management Association (AOA). Joint principles of the patient-
programs for improving diabetes care: a meta- centered medical home. 2007. [Cited 2015 Jun 16].
analysis. CMAJ. 2011;183(2):E11527. Available from: http://www.aafp.org/dam/AAFP/
20. Davidson MB. How our current medical care system documents/practice_management/pcmh/initiatives/
fails people with diabetes. Diabetes Care. PCMHJoint.pdf
2009;32:3702. 32. Wagner EH. The role of patient care teams in chronic
21. Krein SL, Klamerus ML, Vijan S, et al. Case man- disease management. BMJ. 2000;320:56972.
agement for patients with poorly controlled diabetes: 33. Wagner EH, Sandhu N, Coleman K, Phillips KE,
a randomized controlled trial. Am J Med. Sugarman JR. Improving care coordination in pri-
2004;116:7329. mary care. Med Care. 2014;52:S338.
22. Choe HM, Mitrovich S, Dubay D, Hayward RA, 34. Bojadzievski T, Gabbay RA. Patient-centered medi-
Krein SL, Vijan S. Proactive case management of cal home and diabetes. Diabetes Care.
high risk patients with type 2 diabetes mellitus by a 2011;34(4):104753.
clinical pharmacist: a randomized controlled trial. 35. National Committee on Quality Assurance. Patient-
Am J Manag Care. 2005;11:25360. Centered Medical Home. [Cited 2015 Jun 16].
23. Ishani A, Greer N, Taylor BC, Kubes L, Cole P, Available from: http://www.ncqa.org/Programs/
Atwood M, Clothier B, Ercan-Fang N. Effect of R e c o g n i t i o n / P r a c t i c e s /
nurse case management compared with usual care PatientCenteredMedicalHomePCMH/
on controlling cardiovascular risk factors in patients NewVersionofPCMH.aspx
with diabetes: a randomized controlled trial. 36. Agency for Healthcare Research and Quality
Diabetes Care. 2011;34(8):168994. (AHRQ). Defining the PCMH. [cited 2015 Jun 22].
24. Sadur CN, Moline N, Costa M, et al. Diabetes man- Available from: https://pcmh.ahrq.gov/page/
agement in a health maintenance organization: effi- defining-pcmh
cacy of care management using cluster visits. 37. Stange KC, Nutting PA, Miller WL, Jan CR,
Diabetes Care. 1999;22:20117. Crabtree BF, Flocke SA, Gill JM. Defining and mea-
25. Peterson KA, Radosevich DM, OConnor PJ, suring the patient-centered medical home. J Gen
Nyman JA, Prineas RJ, Smith SA, Arneson TJ, Intern Med. 2010;25:60112.
Corbett VA, Weinhandl JC, Lange CJ, Hannan 38. National Committee for Quality Assurance (NCQA)
PJ. Improving diabetes care in practice: findings Care Coordination Standards. McColl Institute
from the TRANSLATE trial. Diabetes Care. 2011. [Cited 2015 Jun 16]. Available from: http://
2008;31(12):223843. www.improvingchroniccare.org/downloads/1_
26. Polonsky WH, Earles J, Smith S, et al. Integrating ncqa_care_coordination_process_measures.pdf
medical management with diabetes self-management 39. Daniel DM, Wagner EH, Coleman K, et al. Assessing
training: a randomized control trial of the diabetes progress toward becoming a patient-centered medi-
outpatient intensive treatment program. Diabetes cal home: an assessment tool for practice transfor-
Care. 2003;26:304853. mation. Health Serv Res. 2013;48:187997.
27. Fleming B, Silver A, Ocepek-Welikson K, Keller 40. Jortberg BT, Miller BF, Gabbay RA, Sparling K,
D. The relationship between organizational systems Dickinson WP. Patient-centered medical home: how
and clinical quality in diabetes care. Am J Manag it affects psychosocial outcomes for diabetes. Curr
Care. 2004;10(12):93444. Diabet Rep. 2012;12(6):7218.
28. Solberg LI, Asche SE, Pawlson LG, Scholle SH, 41. Paulus RA, Davis K, Steele GD. Continuous innova-
Shih SC. Practice systems are associated with high- tion in health care: implications of the Geisinger
quality care for diabetes. Am J Manag Care. experience. Health Aff (Millwood).
2008;14(2):8592. 2008;27:123545.
42. Friedberg MW, Rosenthal MB, Werner RM, Volpp does what in primary care? JAMA Intern Med.
KG, Schneider EC. Effects of a medical home and 2014;174:10256.
shared savings intervention on quality and utilization 54. Bennett HD, Coleman EA, Parry C, Bodenheimer T,
of care. JAMA Intern Med. 2015;175(8):13628. Chen EH. Health coaching for patients with chronic
43. McCarthy D, Mueller K, Tillmann I. HealthPartners: illness. Fam Pract Manag. 2010;17:249.
consumer-focused mission and collaborative 55. Ghorob A. Health coaching: teaching patients how
approach support ambitious performance improve- to fish. Fam Pract Manag. 2013;20(3):402.
ment agenda. New York: The Commonwealth Fund 56. Naik AD, Palmer N, Petersen NJ, Street RL, Rao R,
Publication; 2009. p. 1250. [Cited 2015 Jun 16]. Suarez-Almazor M, Haider P. Comparative effec-
Available from: http://www.commonwealthfund. tiveness of goal setting in diabetes mellitus group
org/~/media/files/publications/case-study/2009/ clinics. Arch Intern Med. 2011;171(5):4539.
jun/1250_mccarthy_healthpartners_case_study_61_ 57. Willard-Grace R, Thom DH, Hessler D, Bodenheimer
for-pf.pdf. T, Chen EH. Health coaching to improve control of
44. Steiner BD, Denham AC, Ashkin E, Newton WP, diabetes, hypertension and hyperlipidemia for low-
Wroth T, Dobson Jr LA. Community care of North income patients: a randomized controlled trial. Ann
Carolina: improving care through community health Fam Med. 2015;28(1):3845.
networks. Ann Fam Med. 2008;6:3617. 58. Heisler M, Vijan S, Makki F, Piette JD. Diabetes
45. Colorado PCMH Multi-Payer. Multi-state pilot: a control with reciprocal peer support versus nurse
year in review. [cited 2015 Jun 16]. Available from: care management: a randomized trial. Ann Intern
h t t p s : / / w w w. p c p c c . o rg / w e b i n a r / Med. 2010;153(8):50715.
cmd-colorado-pcmh-multi-payer-multi-state-pilot- 59. Thom DH, Ghorob A, Hessler D, De Vore D, Chen
year-review EH, Bodenheimer T. Peer health coaching improves
46. Alidina S, Schneider EC, Singer SJ, Rosenthal glycemic in low-income patients with diabetes:
MB. Structural capabilities in small and medium- results of a randomized controlled trial. Ann Fam
sized patient-centered medical homes. Am J Manag Med. 2013;11:13744.
Care. 2014;20(7):e26577. 60. Spencer MS, Rosland AM, Kieffer EC, Sinco BR,
47. Reid RJ, Coleman K, Johnson EA, Fishman PA, Hsu Valerio M, Palmisano G, et al. Effectiveness of a
C, Soman MP, et al. The group health medical home community health worker intervention among
at year two: cost savings, higher patient satisfaction, African- American and Latino adults with type 2
and less burnout for providers. Health Aff diabetes. Am J Public Health. 2011;101:225360.
(Millwood). 2010;29(5):83543. 61. Philis-Tsimikas A, Fortmann A, Lleva-Ocana L,
48. Jan CR, Ferrer RL, Miller WL, Palmer RF, Wood R, Walker C, Gallo LC. Peer-led diabetes education
Davila M, et al. Patient outcomes at 26 months in the programs in high-risk Mexican Americans improve
patient-centered medical home national demonstra- glycemic control compared with standard
tion project. Ann Fam Med. 2010;8(1 Suppl):S5767. approaches: a Project Dulce promotora randomized
S92. trial. Diabetes Care. 2011;34(9):192631.
49. Miller BF, Kessler R, Peek CJ, Kallenberg GA. A 62. Philis-Tsimikas A, Walker C, Rivard L, et al.
national research agenda for research in collabora- Improvement in diabetes care of underinsured
tive care: papers from the collaborative care research patients enrolled in Project Dulce. Diabetes Care.
network research development conference. AHRQ 2004;27:1105.
Publication No. 11-0067. 2011. [Cited 2015 Jun 16]. 63. California Health Workforce Alliance (CHWA),
Available from: http://www.ahrq.gov/research/ Taking Innovation to scale: community health work-
collaborativecare/ ers, promoters, and the triple aim. A statewide
50. Jaber LA, Halapy H, Fernet M, Tummalapalli S, assessment of the roles and contributions of
Diwakaran H. Evaluation of a pharmaceutical care Californias community health workers. Preliminary
model on diabetes management. Ann Pharmacother. findings, observations, and recommendations. Public
1996;30:23843. Health Institute/California Health Workforce
51. Johnson KA, Chen S, Cheng IN, Lou M, Gregerson Alliance 2013. [cited 2015 Jun 16]. Available from:
P, Blieden C, et al. The impact of clinical pharmacy www.calhealthworkforce.org
services integrated into medical homes on diabetes- 64. Coventry P, Lovell K, Dickens C, Bower P, Chew-
related clinical outcomes. Ann Pharmacother. Graham C, McElvenny D, Hann M, Cherrington A,
2010;44:187786. Garrett C, Gibbons CJ, Baguley C, Roughley K,
52. Choe HM, Bernstein SJ, Mueller BA, Walker PC, Adeyemi I, Reeves D, Waheed W, Gask L. Integrated
Stevenson JG, Standiford CJ. Pharmacist leads pri- primary care for patients with mental and physical
mary care team to improve diabetes care. Am multimorbidity: cluster randomised controlled trial
J Health Syst Pharm. 2009;66(7):6224. of collaborative care for patients with depression
53. Bodenheimer T, Willard-Grace R, Ghorob comorbid with diabetes or cardiovascular disease.
A. Expanding the roles of medical assistants: who BMJ. 2015;350:h638.
65. Katon W, Russo J, Lin EH, Schmittdiel J, Ciechanowski 78. Loh A, Simon D, Wills CE, Kriston L, Niebling W,
P, Ludman E, Peterson D, Young B, Von Korff Harter M. The effects of a shared decision-making
M. Cost-effectiveness of a multicondition collabora- intervention in primary care of depression: a cluster-
tive care intervention: a randomized controlled trial. randomized controlled trial. Pat Educ Counsel.
Arch Gen Psychiatry. 2012;69(5):50614. 2007;67:32432.
66. Knowles SE, Chew-Graham C, Adeyemi I, Coupe 79. Elwyn G, Edwards A, Hood K, Robling M, Atwell
N, Coventry PA. Managing depression in people C, Russell I, Wensing M, Grol R, Study Steering
with multimorbidity: a qualitative evaluation of an Group. Achieving involvement: process outcomes
integrated collaborative care model. BMC Fam from a cluster randomized trial of shared decision
Pract. 2015;16(1):32. making skill development and use of risk communi-
67. Barello S, Graffigna G, Vegni E. Patient engagement cation aids in general practice. Fam Pract.
as an emerging challenge for healthcare services: 2004;21(4):33746.
mapping the literature. Nurs Res Pract. 80. Legare F, Stacey D, Pouliot S. Interprofessionalism
2012;2012:905934. and shared decision-making in primary care: a step-
68. Graffigna G, Barello S, Libreri C, Bosio CA. How to wise approach towards a new model. J Interprofess
engage type-2 diabetic patients in their own health Care. 2011;25:1825.
management: implications for clinical practice. 81. Bodenheimer T, Ghorob A, Willard-Grace R,
BMC Public Health. 2014;14:648. Grumbach K. The 10 building blocks of high-
69. Carman KL, Dardess P, Mauer M, et al. Patient and performing primary care. Ann Fam Med.
family engagement: a framework for understanding 2014;12:16671.
the elements and developing interventions and poli- 82. Willard R, Bodenheimer T. The building blocks of
cies. Health Aff. 2013;32:22331. high-performing primary care: lessons from the
70. Hibbard JH, Greene J. What the evidence shows field. California HealthCare Foundation April 2012.
about patient activation: better health outcomes and [Cited 2015 Jun 16]. Available from: http://www.
care experiences; fewer data on costs. Health Aff chcf.org/publications/2012/04/building-blocks-
(Millwood). 2013;32(2):20714. primary-care
71. Hibbard JH, Stockard J, Mahoney ER, Tusler 83. Ghorob A, Bodenheimer T. Sharing the care to
M. Development of the patient activation measure improve access to primary care. N Engl J Med.
(PAM): conceptualizing and measuring activation in 2012;366(21):19557.
patients and consumers. Health Serv Res. 2004;39(4 84. Balkhi AM, Reid AM, Westen SC, Olsen B, Janicke
Pt 1):100526. DM, Geffken GR. Telehealth interventions to reduce
72. Greene J, Hibbard JH, Sacks R, Overton V, Parrotta management complications in type 1 diabetes. World
CD. When patient activation levels change, health J Diabetes. 2015;6:3719.
outcomes and costs change, too. Health Aff 85. Marcolino MS, Maia JX, Alkmim MB, Boersma E,
(Millwood). 2015;34(3):4317. Ribeiro AL. Telemedicine application in the care of
73. Bolen SD, Chandar A, Falck-Ytter C, Tyler C, diabetes patients: systematic review and meta-
Perzynski AT, Gertz AM, Sage P, Lewis S, Cobabe analysis. PLoS ONE. 2013;8(11):e79246.
M, Ye Y, Menegay M, Windish DM. Effectiveness 86. Piette JD, Weinberger M, McPhee SJ, Mah CA,
and safety of patient activation interventions for Kraemer FB, Crapo LM. Do automated calls with
adults with type 2 diabetes: systematic review, meta- nurse follow-up improve self-care and glycemic
analysis, and meta-regression. J Gen Intern Med. control among vulnerable patients with diabetes?
2014;29(8):116676. Am J Med. 2000;108(1):207.
74. Makoul G, Clayman ML. An integrative model of 87. Schillinger D, Hammer H, Wang F, Palacios J,
shared decision making in medical encounters. McLean I, Tang A, Youmans S, Handley M. Seeing
Patient Educ Couns. 2006;60:30112. in 3-D: examining the reach of diabetes self-
75. Elwyn G, Frosch D, Thomson R, Joseph-Williams management support strategies in a public health
N, Lloyd A, Kinnersley P, Cording E, Tomson D, care system. Health Educ Behav.
Dodd C, Rollnick S, Edwards A, Barry M. Shared 2008;35(5):66482.
decision making: a model for clinical practice. J Gen 88. Schillinger D, Handley M, Wang F, Hammer
Intern Med. 2012;27:13617. H. Effects of self-management support on structure,
76. Elwyn G, Hutchings H, Edwards A, Rapport F, process, and outcomes among vulnerable patients
Wensing M, Cheung W, et al. The OPTION scale: with diabetes: a three-arm practical clinical trial.
measuring the extent that clinicians involve patients Diabetes Care. 2009;32(4):55966.
in decision-making tasks. Health Expect. 89. Ratanawongsa N, Handley MA, Sarkar U, Quan J,
2005;8:3440. Pfeifer K, Soria C, Schillinger D. Diabetes health
77. Dy SM. Instruments for evaluating shared medical information technology innovation to improve qual-
decision making: a structured literature review. Med ity of life for health plan members in urban safety
Care Res Rev. 2007;64:62349. net. J Ambul Care Manag. 2014;37(2):12737.
90. Chen AH, Kushel MB, Grumbach K, Yee Jr 103. Ralston JD, Hirsch IB, Hoath J, Mullen M, Cheadle
HF. Practice profile. A safety-net system gains effi- A, Goldberg HI. Web-based collaborative care for
ciencies through eReferrals to specialists. Health type 2 diabetes: a pilot randomized trial. Diabetes
Aff (Millwood). 2010;29:96971. Care. 2009;32(2):2349.
91. Chen AH, Murphy EJ, Yee Jr HF. eReferral a new 104. McMahon GT, Gomes HE, Hohne SH, Hu TM-J,
model for integrated care. New Engl J Med. Levine BA, Conlon PR. Web-based care manage-
2013;368:24503. ment in patients with poorly controlled diabetes mel-
92. Straus SG, Chen AH, Yee Jr H, Kushel MB, Bell litus. Diabetes Care. 2005;28(7):16249.
DS. Implementation of an electronic referral system 105. Bond GE, Burr RL, Wolf FM, Feldt K. The effects of
for outpatient specialty care. AMIA Annu Symp a web-based intervention on psychosocial well-
Proc. 2011;2011:133746. being among adults aged 60 and older with diabetes:
93. Scheibe MM, Imboden JB, Schmajuk G, Margaretten a randomized trial. Diabetes Educ.
M, Graf JD, Chen AH, Yelin EH, Yazdany 2010;36(3):4465.
J. Efficiency gains for rheumatology consultation 106. Azar M, Gabbay R. Web-based management of dia-
using a novel electronic referral system in a safety betes through glucose uploads: has the time come for
net health setting. Arthritis Care Res (Hoboken). telemedicine? Diabetes Res Clin Pract.
2015;67(8):115863. 2009;83(1):917.
94. Sewell JL, Day LW, Tuot DS, Alvarez R, Yu A, Chen 107. McMahon GT, Fonda SJ, Gomes HE, Alexis G,
AH. A brief, low-cost intervention improves the Conlin PR. A randomized comparison of online- and
quality of ambulatory gastroenterology consultation telephone-based care management with internet
notes. Am J Med. 2013;126(8):7328. training alone in adult patients with poorly con-
95. Pereira K, Phillips B, Johnson C, Vorderstrasse trolled type 2 diabetes. Diabetes Technol Ther.
A. Internet delivered diabetes self-management edu- 2012;14(11):10607.
cation: a review. Diabetes Technol Ther. 108. Liang X, Wang Q, Yang X, Cao J, Chen J, Mo X,
2015;17(1):5563. Huang J, Wang L, Gu D. Effect of mobile phone
96. McIlhenny CV, Guzic BL, Knee DR, Wendekier intervention for diabetes on glycaemic control: a
CM, Demuth BR, Roberts JB. Using technology to meta-analysis. Diabet Med. 2011;28(4):45563.
deliver healthcare education to rural patients. Rural 109. Bell AM, Fonda SJ, Walker MS, Schmidt V, Vigersky
Remote Health. 2011;11:1798. RA. Mobile phone-based video messages for diabe-
97. Avdal EU, Kizilci S, Demirel N. The effects of web- tes self-care support. J Diabetes Sci Technol.
based diabetes education on diabetes care results: a 2012;6(2):3109.
randomized control study. Comput Inform Nurs. 110. Salber P, Niksch A. Digital health tools for diabetes.
2011;29(2 Suppl):Tc2934. J Ambul Care Manag. 2015;38(2):1969.
98. Lorig K, Ritter PL, Laurent DD, Plant K, Green M, 111. Quinn CC, Clough SS, Minor JM, Lender D, Okafor
Jernigan VB, Case S. Online diabetes self- MC, Gruber-Baldini. A WellDoc mobile diabetes
management program: a randomized study. Diabetes management randomized controlled trial: change in
Care. 2010;33:127581. clinical and behavioral outcomes and patient and
99. Kim CJ, Kang DH. Utility of a web-based interven- physician satisfaction. Diabetes Technol Ther.
tion for individuals with type 2 diabetes: the impact 2008;10(3):1608.
on physical activity levels and glycemic control. 112. Quinn CC, Gruber-Baldini AL, Shardell M, Weed K,
Comput Inform Nurs. 2006;24:33745. Clough SS, Peeples M, Terrin M, Bronich-all L, Barr
100. Noh JH, Cho YJ, Nam HW, Kim JH, Kim DJ, Yoo E, Lender D. Mobile diabetes intervention study:
HS, Kwon YW, Woo MH, Cho JW, Hong MH, Yoo testing a personalized treatment/behavioral commu-
JH, Gu MJ, Kim SA, An KE, Jang SM, Kim EK, nication intervention for blood glucose control.
Yoo HJ. Web-based comprehensive information sys- Contemp Clin Trials. 2009;30(4):3344.
tem for self-management of diabetes mellitus. 113. Quinn CC, Shardell MD, Terrin ML, Barr EA,
Diabetes Technol Ther. 2010;12:3337. Ballew SH, Gruber-Baldini AL. Cluster-randomized
101. Song M, Choe MA, Kim KS, Yi MS, Lee I, Kim J, trial of a mobile phone personalized behavioral
Lee M, Cho YM, Shim YS. An evaluation of web- intervention for blood glucose control. Diabetes
based education as an alternative to group lectures Care. 2011;34(9):193442.
for diabetes self-management. Nurs Health Sci. 114. Nundy S, Dick JJ, Chou CH, Nocon RS, Chin MH,
2009;11:27784. Peek ME. Mobile phone diabetes project led to
102. Glasgow RE, Kurz D, King D, Dickman JM, Faber improved glycemic control and net savings for
AJ, Halterman E, Woolley T, Toobert DJ, Strycker Chicago plan participants. Health Aff (Millwood).
LA, Estabrooks PA, Osuna D, Ritzwoller D. Twelve- 2014;33(2):26572.
month outcomes of an Internet-based diabetes self- 115. Clarke RM, Tseng CH, Brook RH, Brown AF. Tool
management support program. Patient Educ Couns. used to assess how well community health centers
2012;87:8192. function as medical homes may be flawed. Health
Aff (Millwood). 2012;31(3):62735.
116. Bray P, Cummings DM, Morrissey S, Thompson D, 121. Edwards ST, Landon BE. Medicares chronic care
Holbert D, Wilson K, Lukosius E, Tanenberg management payment payment reform for primary
R. Improved outcomes in diabetes care for rural care. N Engl J Med. 2014;371:204951.
African Americans. Ann Fam Med. 122. Bindman AB. Nudging medical practice change one
2013;11(2):14550. regulation at a time. JAMA. 2015;313:11978.
117. Moore K, Jiang L, Manson SM, Beals J, Henderson 123. Gilmer TP, Philis-Tsimikas A, Walker C. Outcomes
W, Pratte K, Acton KJ, Roubideaux Y. Case manage- of Project Dulce: a culturally specific diabetes man-
ment to reduce cardiovascular disease risk in agement program. Ann Pharmacother.
American Indians and Alaska Natives with diabetes: 2005;39(5):81722.
results from the Special Diabetes Program for 124. Gilmer TP, Roze S, Valentine WJ, et al.
Indians Healthy Heart Demonstration Project. Am Cost-effectiveness of diabetes case management for
J Public Health. 2014;104(11):e15864. low-income populations. Health Serv Res.
118. Shea S, Weinstock RS, Starren J, Teresi J, Palmas W, 2007;42(5):194359.
Field L, et al. A randomized trial comparing tele- 125. Brown 3rd HS, Wilson KJ, Pagan JA, Arcari CM,
medicine case management with usual care in older, Martinez M, Smith K, Reininger B. Cost-
ethnicity diverse, medically underserved patients effectiveness analysis of a community health worker
with diabetes mellitus. J Am Med Inform Assoc. intervention for low-income Hispanic adults with
2006;13:4051. diabetes. Prev Chronic Dis. 2012;9:E140.
119. Shane-McWhorter L, Lenert L, Petersen M, Woolsey 126. Brownson CA, Hoerger TJ, Fisher EB, Kilpatrick
S, McAdam-Marx C, Coursey JM, Whittaker TC, KE. Cost-effectiveness of diabetes self-management
Hyer C, LaMarche D, Carroll P, Chuy L. The Utah programs in community primary care settings.
remote monitoring project: improving health care Diabetes Educ. 2009;35(5):7619.
one patient at a time. Diabetes Technol Ther. 127. Piette JD, Heisler M, Wagner TH. Cost-related med-
2014;16(10):65360. ication underuse among chronically ill adults: the
120. Rogers EA, Hessler D, Dube K, Willard-Grace R, treatments people forgo, how often, and who is at
Gupta R, Bodenheimer T, Grumbach K. risk. Am J Public Health. 2004;94(10):17827.
The panel management questionnaire: a tool to 128. Hoerger TJ. Using costs in cost-effectiveness models
measure panel management capability. Perm for chronic diseases: lessons from diabetes. Med
J. 2015;19(2):49. Care. 2009;47(7 Suppl 1):S217.
A Primary Health-Care System
Approach to Improving Quality 4
of Care and Outcomes in People
with Diabetes: The University
of Pittsburgh Medical Center
Experience
Janice C. Zgibor, Francis X. Solano Jr.,

and Linda Siminerio
Introduction cant improvement in diabetes outcomes on a

population basis [8]. However, relying on a sys-
Although substantial evidence exists demonstrat- tem that places the burden solely on primary care
ing the efcacy of interventions for controlling providers (PCPs) to reach evidence-based targets
risk factors for diabetes complications [16], dis- [9] for risk factor control may not be ideal.
semination into community practice is not wide- Traditional approaches used to enhance the qual-
spread, for example, only 18 % of people with ity of diabetes care including continuing medical
diabetes in the USA are achieving goals for education, self-study through medical literature
HbA1c, blood pressure, and cholesterol (the or online courses, may improve knowledge short-
ABCs) [7]. Since the majority of ambulatory care term but do little to actually change quality of
visits for diabetes take place in primary care, this patient care [10, 11]. Primary care providers need
environment provides an opportunity for signi- innovative strategies to support the current
demands for high quality care with limited
resources. To achieve this, it is important to
J.C. Zgibor (*) broaden our perspective to include models of
Department of Epidemiology and Biostatistics, care that may improve quality using more nontra-
College of Public Health, University of South ditional approaches.
Florida, Tampa, FL, USA
e-mail: [email protected] As early as 2000, the University of Pittsburgh
Medical Center (UPMC) began adopting the ele-
F.X. Solano Jr.
Community Medicine Inc, University of Pittsburgh ments of the Chronic Care model (CCM) [1215]
Medical Center, Pittsburgh, PA, USA to enhance the quality of diabetes care using risk
Physician Services Division, University of Pittsburgh stratication, provider incentives, integration of
Medical Center, Pittsburgh, PA, USA care managers, travelling diabetes educators,
Division of General Internal Medicine, School of population management, and collaboration with
Medicine, University of Pittsburgh, a health insurer. Many initiatives were undertaken
Pittsburgh, PA, USA and successful outcomes were achieved. This
e-mail: [email protected] chapter highlights these initiatives and describes
L. Siminerio the detail that led to large-scale change in patient
Division of Endocrinology, School of Medicine, outcomes, diabetes care delivery, and population
University of Pittsburgh, Pittsburgh, PA, USA
e-mail: [email protected] management.

52 J.C. Zgibor et al.
While several models of care are reported in Provider-Based Strategies

the literature, the CCM was chosen to provide the for Quality Improvement
framework to enhance diabetes care at the
UPMC. The CCM posits that quality care is In 2000, UPMC embarked on a system-wide
delivered using a comprehensive patient-centred quality improvement initiative in the CMI prac-
approach that rewards outcomes not just pro- tices. The philosophy of the quality improvement
cesses of care. Previous research conducted on (QI) programme was to improve quality of care
the effectiveness of the CCM demonstrated while keeping it simple. This initiative was
improvement in these parameters in a variety of called Intervention Now!! Presently it includes
primary health-care settings [16]. The elements up to 360 primary care physicians and 45,000 of
of the CCM are reviewed in detail in Chap. 3. their patients with diabetes. The leadership of
UPMC chose to implement the CCM to move CMI (the President of CMI and Medical Director
towards strategies for the redesign of primary and his administrative team) used the CCM
care, to improve processes and outcomes in peo- framework for the initiative.
ple with diabetes. All elements of the CCM To launch the initiative, CMI leadership estab-
(health system, community, decision support, lished evidence-based clinical goals. To dissemi-
clinical information systems, self-management, nate these goals, the Medical Director of CMI
and delivery system design) [1215] were imple- met with primary care practices and presented
mented in UPMC primary care practices over the results from recent studies, treatment goals, and
15-year timeframe described here. strategies for improvement. There was no system-
wide electronic health record (EHR) at the time,
therefore data were entered onto ow sheets at
Population Served the patient and practice levels. Physicians
received information about local Diabetes Self-
The University of Pittsburgh Medical Center Management Education (DSME) programmes
(UPMC), established in 1990, is a global non- and were encouraged to refer patients to these
prot organization with more than 62,000 programmes for education. Given the geographic
employees, 5500 afliated physicians and 20 distance of some of the CMI practices from the
hospitals. There are more than 500 clinical loca- UPMC hub in Pittsburgh, PA, the Medical
tions that include doctors ofces in the Physicians Director delivered the programme via teleconfer-
Service Division and Community Medicine, Inc., ence. Initial targets for improvement were to
(CMI) which has 150 primary care locations and intensify treatment in those with an HbA1c >9 %,
330 primary care providers. UPMC primarily statin use in those over the age of 40 at high risk
serves a diverse population in western for a cardiovascular event regardless of LDLc
Pennsylvania [17] and provides care for more level, blood pressure control in those with a blood
than 55,000 people with diabetes. There are pressure >140/90 mmHg, ACE or ARB use for
approximately 600 primary care providers in renal protection and aspirin use in those at high
UPMC. Primary care providers at UMPC and for risk for a coronary heart disease event. The pilot
this chapter include general practice, family prac- phase of the initiative occurred in 2000 and
tice and internal medicine physicians. included 148 volunteer providers. Average
The University of Pittsburgh Medical Center HbA1c was 7.6 % while 34 % of patients with
also has an insurance division with more than diabetes had an HbA1c <7 %. By 2003, the initia-
three million members. UPMC Health Plan is the tive was adopted by 198 providers and approxi-
second largest health insurer in western mately 25,000 of their patients. A reduction in
Pennsylvania (PA) and serves Medicare, average HbA1c to 6.9 % across all patients was
Medicaid, special needs populations, and work- reported with 61 % of patients achieving an
ers compensation. The insurer also serves parts HbA1c <7 %. Late adopters usually had worse
of Ohio, West Virginia and Maryland [18]. glycaemic control, so this was unlikely to be due
4 A Primary Health-Care System Approach to Improving Quality of Care and Outcomes 53
to dilution by the inux of patients with better Patient-Centred Strategies-Diabetes

glucose control. Eighty-three percent of practices Self-Management Education
lowered their patients average HbA1c during the
year of the intervention. There were practical The person living with diabetes is expected to
limitations to this tracking method including make a multitude of daily self-management deci-
accuracy, incomplete capture of patients and their sions and perform complex care activities.
HbA1c, and the potential for reporting bias. In DSME/DSMS provides the foundation to help
order to facilitate more efcient data collection people with diabetes navigate these decisions and
for HbA1c, the primary laboratory in the local activities [19, 20]. DSME is the process of facili-
area connected with the UMPC data network so tating the knowledge, skill and ability necessary
that periodic data dumps could be extracted. This for diabetes self-care. DSMS refers to the support
facilitated a more comprehensive data capture that is required for implementing and sustaining
that was less prone to bias. The laboratory data behaviours needed for self-management over the
were sent to physicians in an anonymized report long term [21]. While different members of the
on a quarterly basis. The Medical Director was health-care team and community can contribute
then able to share anonymized data with the pro- to this process, initial DSME should be provided
viders, which provided a means for them to by a health professional while ongoing support
examine their own data and compare their prac- can be provided by personnel within a practice
tice results to their peers. Physicians were ranked and a variety of community-based resources.
on their progress towards established goals. DSME/DSMS interventions are designed to: be
Improvements continued throughout CMI in all culturally relevant, address patients health
198 primary care practices and a report from beliefs, current knowledge, limitations, emo-
2008 showed that 75 % of HbA1c tests were tional concerns, family support, nancial status,
<7 %, an absolute improvement of 41 % from the medical history, health literacy, and other factors
beginning of the initiative. Since the proportion that inuence each persons ability to meet the
of patients with an HbA1c below goal had challenges of self-management [21, 22].
dra matically improved overtime, initiatives DSME/DSMS are now considered to be a cor-
to control LDLc and blood pressure were started nerstone of care, and have repeatedly been shown
as the next phase in the quality improvement to improve diabetes-related outcomes [2327]. In
initiatives. addition to signicant improvements in HbA1c
[20, 2831], DSME is reported to have a positive
impact on other clinical, behavioural and psycho-
The Legacy of Implementation social aspects of diabetes [32, 33] and reduce the
onset and/or advancement of diabetes complica-
Following the initial pilot work, several demon- tions [26] improve quality of life [34, 35],
stration projects were simultaneously initiated to improve lifestyle behaviours [36], enhance self-
examine the most effective and globally accept- efcacy and empowerment [37], increase healthy
able means for implementation and adoption of coping [38], and decrease the presence of
the CCM. The demonstration projects imple- diabetes-related distress [39] and depression
mented all elements of the CCM to varying [40]. DSME/DSMS has also been shown to be
degrees. Self-management is considered to be a cost-effective by reducing hospital admissions
core tenet of the CCM, thus at the foundation of and readmissions [41, 42] as well as estimated
these interventions was attention to diabetes self- lifetime health-care costs related to a lower risk
management education and support (DSME/ for complications [43].
DSMS). This next section describes the role of While these improvements clearly reafrm the
DSME/DSMS within the context of quality importance of DSME, participation in DSME is
improvement strategies and implementation of low and educator services are underutilized [43
the CCM. 45]. It is the position of the American Diabetes
Association (ADA), Institute of Medicine and Demonstration Projects

National Committee for Quality Assurance that
all individuals with diabetes receive DSME/ Phase 1: Exploring Referral
DSMS at diagnosis and as needed thereafter [9, and Participation Practices
22, 46]. Despite these recommendations, recently In 2000, our research team began exploring
released data show that only 6.8 % insured, newly DSME services when UPMC physicians reported
diagnosed US adults with diabetes [45], and only a lack of access to diabetes education as a major
4 % of Medicare participants participated in barrier to quality care despite the fact that UPMC
DSME [43]. It has been suggested that the tradi- had supported hospital-based DSME pro-
tional way in which DSME is prescribed and grammes. The results of a needs assessment
delivered may be problematic [44]. Currently in revealed that only three of the 19 DSME pro-
the US, a physician must refer patients to grammes were recognized by the ADA. This rec-
DSME. In a national survey examining access, ognition ensures quality DSME services are
physicians reported that they want patients to delivered and enables programmes to bill for their
receive DSME but nd referral procedures dif- services. UPMC applied for and received recogni-
cult. Other challenges reported include a poor tion for all programmes in November 2000 [47].
understanding of the need for and effectiveness
of DSME. In addition, DSME/DSMS has histori- Phase 2: Addressing Access to DSME
cally been provided through didactic, formal pro- (20012003)
grammes despite efforts underway to present We explored improving access to DSME by
diabetes content in a more dynamic and patient- implementing a care model shown to improve
focused approach. Survey authors concluded that processes of care within primary care practices
efforts are needed to increase referral by physi- and patient outcomes. Again, the CCM provided
cians, follow-up by patients, and make DSME the organizational approach with particular atten-
available in forms that make it attractive to tion paid to team care and self-management [12
patients and physicians [44]. 15]. A cluster randomized trial was implemented
to examine the implementation of the CCM in an
underserved, urban community [16]. Eleven pri-
Stepped Approach to Integrating mary care practices, along with their patients
DSME in Practice with either type 1 or type 2 diabetes (although the
majority had type 2), were randomized to three
We chose DSME/DSMS as the foundation of our groups: (1) CCM intervention where a diabetes
implementation of the CCM for several reasons. educator provided DSME in the practice, (2) pro-
First, the patient is responsible for the majority of vider education where a series of problem-based
their self-care outside of the physician ofce. cases were presented to physicians, and (3) usual
DSME provides the skills necessary to do this self- care. On average, the CCM group showed a sig-
management using a patient-centred approach. nicant decline in HbA1c (0.6 %, p= 0.008),
Second, primary care providers do not have the while the other groups did not. The CCM group
time to facilitate behaviour change strategies and also showed improvement in patient knowledge
ongoing support. Evidence demonstrates that this (p = 0.07) and empowerment (p = 0.02).
is a strength of DSME/DSMS. Third, diabetes Access to diabetes education in the rural areas
educators are skilled clinicians with the expertise was particularly problematic. To begin to explore
necessary to make therapeutic recommendations these challenges, a study was conducted using the
for treatment intensication. Finally, their role in CCM was as a framework to test access to DSME
US primary care is cost-effective. In this next sec- in a rural area. A review of the practice population
tion we describe our demonstration projects and revealed that 95 % of patients with type 2 diabetes
the evolution of DSME/DSMS in primary care as had signicant cardiovascular disease risk factors;
it stands today. however, only 7 % had received relevant services
like Medical Nutrition Therapy and none had vider and diabetes educators to address this
received DSME. After integrating diabetes educa- signicant challenge. A point of service education
tor service into the practice, providers adherence model was examined in four UPMC practices to
to the ADA Standards of Care signicantly address these issues [52]. A nurse diabetes educa-
improved, diabetes educator utilization increased, tor offered DSME in the ofce on designated dia-
and patients who received DSME had signicant betes days. The diabetes educator made therapy
improvements in knowledge [48] and mean HbA1c recommendations to the primary care provider
from baseline to study end (7.2 % vs. 6.5 %, based on patient assessments, trained staff on new
p = 0.007). Results suggest that implementing a therapies and served as a practice resource. The
model that establishes a diabetes educator within providers and the diabetes educator reported many
the primary care setting is effective in improving advantages with the intervention that included
process, clinical and behavioural outcomes [28]. increased communication on management plans
In another study in a rural area, 295 patients and diabetes educators involvement in medication
with type 2 diabetes were identied within pri- initiation and adjustments. Patients reported more
mary care practices, 162 (65 %) reported they had condence in provider communication on treat-
never received DSME. Despite efforts by diabe- ment plans and satisfaction with ease of diabetes
tes educators to improve awareness of local edu- educators access for inquiries [52]. These results
cation services by primary care providers, 123 again afrmed the role of diabetes educators deliv-
(76 %) of the 162 patients had never received a ering services in the primary care practices.
referral. Those patients who received a referral Our ndings were encouraging; however, admin-
had a higher number of risk factors and comorbid istrators challenged us to demonstrate how this could
conditions than those who did not. Eighty-three be supported and the benets of DSME sustained?
percent of the patients, who received a referral Attempts to answer these questions were made in a
from the primary care provider, however, had systematic way again using the CCM as a frame-
already participated in DSME. Studies suggest work. DSME visits were tracked by reimbursement
that physician recommendations are central to a G codes in the UPMC database and compared
patients decision and physician referral has been between education delivered in primary care and
positively associated with patient participation in hospital programmes. A two to threefold increase in
health services. Our ndings reected this and the proportion of patients receiving DSME when
afrmed the need to develop processes to improve delivered in primary care (24.7 % versus hospital-
physician referrals and receipt of DSME [30]. based 8.3 %, P < 0.0001) was reported. In addition,
Efforts to explore our communitys needs con- diabetes educators were able to demonstrate their
tinued by examining our progress towards meeting ability to generate revenue. At programme initiation,
the diabetes objectives in Healthy People 2010 diabetes education services were a loss leader to the
goals (increase receipt of diabetes education for health system. In contrast, at conclusion, diabetes
adults) [49] and potential challenges that still educators were covering costs through reimburse-
needed to be addressed in primary care [50]. ment [53]. The intent was not to suggest that DSME
UPMC primary care providers reported several is a large revenue source or that hospital programmes
barriers to outpatient DSME services that are con- will all be replaced by primary care, but rather
sistent with other reports [44, 51]. Providers found that opportunities exist to expand this cost-neu-
the referral process difcult, expressed fears that tral service and should be explored.
patients are told to do things that they do not agree
with, and that they may lose patients to specialists Phase 3: Examining Self-Management
following DSME. In addition, providers and dia- Support: An Expanded Role
betes educators reported using different Electronic for Diabetes Educators? (20112012)
Health Record systems (EHR), thus eliminating While arguments could be made for the nancial
opportunities for tracking patient DSME partici- benets of diabetes educators in the short and
pation and direct communication between the pro- long-term, evidence for sustained clinical effec-
tiveness was still needed. There was also an addi- received usual care. This was a cluster random-
tional need to explore ongoing DSMS, which ized trial that took place in community-based pri-
was necessary to sustain improved outcomes mary care practices in Southwestern Pennsylvania
following DSME [54]. In a comparative effec- among 240 type 2 diabetes participants who had
tiveness research study DSME/DSMS was exam- at least one of the ABC levels above the ADA-
ined in three geographically/socioeconomically recommended goals. At the end of the 1-year
diverse primary care practice communities [53]. intervention, there was a signicant difference in
All type 2 patient-participants rst received HbA1c between the intervention and usual care
DSME from a diabetes educator over the course groups. The HbA1c in the intervention group
of 6 weeks. Participants were then randomized to decreased from 8.8 % to 7.3 %, while the HbA1c
receive 6 months of DSMS delivered by super- in the usual care group increased slightly from
vised supporter (peer, practice staff, or the dia- 8.2 % to 8.3 % (p = 0.001). There was no signi-
betes educators). This supporter was trained and cant difference between groups for LDLc or sys-
supervised by the diabetes educator. DSMS tolic blood pressure. Findings suggested that
groups were compared to determine which sup- CDEs following standardized protocols is a fea-
port agent helped participants to maintain/ sible strategy and can effectively intensify treat-
improve clinical and behavioural outcomes. ment and improve glycaemic control.
Patients experienced a signicant reduction in In an effort to organize a system-wide approach
HbA1c values (0.9 %, p = 0.0001) and signicant to integrate and expand the role of the diabetes
improvements in empowerment [49], self-care educator in primary care, we developed the
[55] and distress scores [39] following DSME in Glucose to Goal. programme. Glucose to
the primary care practices. Although those in the Goal is a novel diabetes educator-driven,
diabetes educator group best sustained improved population-based management programme for
HbA1c, all groups maintained glycaemic patients with diabetes. The purpose of this project
improvements regardless of supporter. Study was to assess the feasibility of Glucose to Goal,
ndings reafrmed that DSME provided by dia- offered within the constructs of the ADA-
betes educators in primary care is effective, recognized DSME programme within the UPMC
DSMS supported by various agents trained and network, and evaluate its impact on patient gly-
supervised by diabetes educators is feasible and caemia. Three diabetes educators were introduced
has potential to be sustained. into primary care practices in their respective
urban, suburban, and rural communities. Through
Phase 4: Maximizing the Role the primary care EHR, diabetes educators proac-
of the Diabetes Educators in Primary tively identied patients, reviewed lists with pro-
Care (20122014) viders for diabetes educator referral, arranged
The next phase in our demonstration projects diabetes educator visits and worked collabora-
moved towards the diabetes educators maximiz- tively with providers on treatment plans. HbA1c
ing their expertise in both the behavioural and values were collected 13 months prior to DSME
clinical aspects of diabetes care. Since our efforts (baseline, HbA1c) to establish patient glycaemic
to integrate diabetes educators into primary care control pre-DSME. Patients were categorized by
were successful, it was logical to enhance the role HbA1c: 7 %, >7 % to 9 %, and >9 %. These val-
of the diabetes educator in the practice setting. ues were compared to HbA1c levels at 36 and
We implemented the Redesigning Effectiveness 912 months post DSME. HbA1c values were
for Treatment in Diabetes Study (REMEDIES available in the EHR for 78 % of 143 patients
4D) [56] to assess the differences in A1C, blood (61.3 years, 51 % male) who met with a diabetes
pressure, and LDLc levels (ABC) in participants educator during the study period.
to whom a diabetes educator intensied diabetes Average HbA1c values during the study
management by following the standardized, period are shown in Fig. 4.1. For patients with
preapproved protocols compared with those who baseline HbA1c 7 % (n = 32), HbA1c levels
Fig. 4.1 HbA1c over time in UPMC Glucose to Goal programme
were maintained over time. For patients with Assessing Barriers to Practice
baseline HbA1c >79 % (n = 41), HbA1c was sig- Transformation
nicantly reduced initially,, but trended upward
at 12 months. For patients with baseline HbA1c During the implementation of the demonstration
>9 % (n = 39), HbA1c was signicantly reduced projects, the CCM was gaining widespread atten-
between prediabetes education visit and 6 months tion; however, adoption of the elements of the
post- diabetes education visit(s) (1.9; p < 0.001) CCM was not universal within our health system.
and this reduction was sustained at 12 months It seemed that the early adopters were on board,
post-diabetes educator visit(s) (1.7; p < 0.001). while challenges remained for others. From
Findings demonstrate the feasibility of this dia- March through October 2009, we sought to
betes education-led primary care approach and explore and enumerate the barriers to adopting
conrm its benets in all patients with type 2 dia- the Model. Our research team partnered with dia-
betes, particularly for those at higher risk [57]. betes educators in rural areas of western
This study also reafrms the need for ongoing Pennsylvania to organize and participate in a
patient support by diabetes educators in order to qualitative study to gain insights into adoption of
maintain glycaemic improvements over time. the CCM by primary care practices. These areas
In summary, the ndings from our demonstra- are shown on the map in the darker shaded areas
tion projects reect challenges with referrals, the (Fig. 4.2).
importance of referrals on patient participation, Our target geography was challenged by
and the benet the diabetes educator can have on resources and lacked the infrastructure and sup-
patient outcomes in the primary care setting. port provided by a large health system as they
Additionally, maximizing the role of diabetes were outside of the UPMC network. The ratio-
educators to facilitate therapeutic management is nale for exploring the barriers in resource poor
feasible and effective. Our work provides the evi- areas was to identify methods for adopting the
dence for elevating the role of diabetes educators CCM from the most challenged areas that could
in primary care and assuring their place as a vital be applied across a variety of primary care
member of the health-care team. settings.
Fig. 4.2 Map of rural Southwestern Pennsylvania. The shaded area represents the general location of rural primary care
practices
Following identication of 13 practices by the for implementation were outlined. Following the
local diabetes educator, a trained qualitative discussion, the elements of the CCM that were of
interviewer conducted discussion groups in these interest to the primary care providers in a
practices. Groups ranged in size from three to particular practice were rolled out into that
ve and included physicians, ofce managers, practice. Details of the themes identied in the
Licensed Practical Nurses (LPN) and medical discussion groups are summarized in Table 4.1.
assistants. A total of 49 health-care providers and The most common element of the CCM identi-
ofce staff participated across the 13 practices. ed by the practices for immediate implementa-
The format for the discussion group included tion was DSMS/Delivery System Design or
brief introductions (10 min), the purpose of the easier access to a diabetes educator. Providers
discussion group (5 min), the CCM conceptual believed that access to diabetes educators was
framework (2030 min), discussion by the group essential to enhancing patient self-management
(3060 min), feedback and ndings from the and behaviour change. Practices also identied
researchers (15 min), and question and answer the need for resources to implement the Decision
(as needed). During the discussion, participants Support element of the CCM by having posters
were asked to identify at least three elements of available as reminders to patients (e.g., foot care),
the CCM that could be implemented in their ow sheets to track patient data, and easier access
practice within the next month, 6-months and the to current practice guidelines. The elements iden-
next year. They were also asked prioritize imple- tied for long-term implementation were support
mentation of the elements and enumerate the from the Health System and Community. The
methods they would use to implement these element of least interest to the practices was
elements. Researchers asked them to identify Clinical Information Systems given nancial bar-
possible barriers to implementation. Problem- riers and lack of time or interest (EHRs were not
solving strategies were discussed and action steps in widespread use at this time).
Table 4.1 Themes identied by primary care practices was facilitated by the deployment of practice
during focus groups about the chronic care model
coaches. These coaches were nurses and mid-
Element of the level providers with experience in diabetes edu-
CCM Themes identied
cation and quality improvement processes. These
Health system One staff person takes on
individuals worked with the research staff and
numerous and sometimes
disparate responsibilities met with the ofce managers, providers and
The length of the visit was not ofce staff as needed. Their expertise allowed
long enough to address all patient diabetes educators in rural sites to rely on those
needs who had signicant experience in troubleshoot-
Hosting other providers like ing quality improvement strategies. Having this
CDEs was not feasible due to
space limitations
expertise available to practices as issues arose
Lack of knowledge on
was also important for addressing barriers in real
reimbursement for services time. Some of the topics with which the coaches
provided in the ofce were asked to assist included: methods for incor-
Community Issues of territory or boundaries porating diabetes education into the practice,
prohibited collaboration team-based diabetes care, requirements for ADA
Patients were often unaware of Recognition, National Committee for Quality
opportunities to promote healthy
lifestyles like group walks Assurance (NCQA) certication, and billing and
Community initiatives should not reimbursement for diabetes education services.
be one-time things, but rather The researchers also offered resources to the
they should aim for consistent, practices that included diabetes ow sheets, one-
continuous efforts to offer page informational sheets for diabetes self-
education and raise awareness
management, time and location of local DSME/
Existing community events could
partner to offer wellness activities DSMS programmes, and educational materials
Decision support No quick way to examine for patients.
longitudinal patterns in clinical
outcomes
Lack of knowledge about Perceived Patient Psychosocial
guidelines for diabetes care
Barriers
Self-management Lack of reimbursement for
diabetes education activities
Patient acknowledging that they
Throughout the discussions it became clear that
need to manage their diabetes staff and providers perceived that there were
Dispel myths about diabetes characteristics within the patient population that
Delivery system Have a diabetes educator in the contributed to challenges in meeting expectations
design practice so that patients could see for self-management and clinical management of
them at the time of their visits diabetes.
Group appointments Most participants from the practices con-
Lack of administrative support for
tended that patients did not know enough about
change
diabetes to manage it properly. They recognized
Clinical Interoperability of computers-
information hospitals had different computer that some of this may be due to the failure of
systems systems leading to confusion and insurance companies to reimburse for diabetes
frustration education, but they also indicated that some of
No electronic health record the limited knowledge is due to patients not
listening to what they are told by their doctor.
Some participants admitted that, upon diagnosis,
The research team provided assistance for patients are inundated with too much information
implementation of the elements of the CCM pri- to digest it all at once. There were also miscon-
oritized by the primary care practices. Assistance ceptions about diabetes noted during the discus-
sions. For example, staff noted that patients think patients unique needs and preferences [58].
that because their mother has diabetes, they The UPMC Health Plan initiated a PCMH
know how to manage it. These and other local pilot. Ten practices with 162 primary care phy-
attitudes hampered the success of educational sicians serving 23,930 Health Plan members
initiatives. The staff and providers also recog- participated in the pilot programme through
nized that their patient populations are for the 2010. The UPMC Health Plan provided six
most part poor, which affects the kind of food practice-based nurse care managers for the ten
that they can buy and their access to transporta- practices. Care managers received comprehen-
tion and insurance coverage. sive training on diabetes and related co-mor-
bidities from two diabetes educators. Three of
the practice-based nurse care managers focused
Summary of Findings on helping patient members with one or more
chronic conditions. Risk-stratication was
Providers and staff have characteristics that may used to identify high-risk patients. The care
impact how they care for patients with diabetes. managers expanded their focus to address pre-
Two important characteristics to consider are vention, self-management of chronic condi-
their attitudes towards change in the practice and tions, reduction of clinical gaps in care (e.g.,
their tendency to blame patients for their inability laboratory testing, eye exams), unplanned care,
to take action and adhere to recommended treat- and use of duplicative services [59].
ment regimens. The former is seen in remarks Practices provided ofce space, integration of
like my patients would never take to group the care manager into their work ow, and access
appointments; the latter is reected in state- to the EHR for scheduling and documentation. At
ments such as they wont comply with their the end of the pilot period, the ten PCMH prac-
medication schedule. On the other hand, many tices were compared with the remainder of the
of the providers and staff are quite committed to practices in the primary care network. Overall,
helping their patients better manage their diabe- the practices participating in the PCMH pilot
tes and to providing the best care that they can. achieved signicantly lower costs and reduced
A common barrier identied from the focus hospital readmissions. While emergency room
groups is that engaging the administration at (ER) visits and hospital admissions increased in
the Health System level early on and obtaining both groups, the increase was proportionately
the support from them was crucial for success lower in the PCMH practices (Fig. 4.3). When
of a sustainable programme. Administrative Healthcare Effectiveness Data and Information
inertia would need to be overcome in order to Set (HEDIS) indicators were compared, the
establish a business model for sustainability of PCMH sites outperformed the other sites on all
DSME services. measures between 2008 and 2010 (HbA1c tests:
+6.6 % vs +3.4 %, eye exam: +23.2 vs 7.1 %,
LDLc screen: +9.7 % vs +2.9 %, nephropathy
Driving Change into the Future monitoring: +6.8 % vs +4.8 %), although results
were not statistically signicantly different
The Patient-Centered Medical Home (Table 4.2). The return on investment was 160 %
(PCMH) [56]. Those participating in the pilot also received
support for completing the National Committee
As described in this chapter, the PCMH model for Quality Assurances (NCQA) application.
is grounded in the concepts of the CCM [13 Currently, some of the practices are NCQA des-
16]. In this model, practices seek to improve ignated PCMH; however, many of the elements
the quality, effectiveness, and efciency of the of the PCMH are present throughout the UPMC
care they deliver while responding to each primary care network.
Fig. 4.3 Change in Hospital service use per 1000 members of UPMC Health Plans Primary Care Network, 2008
2010. PCMH= patient centered medical home
Table 4.2 Change in performance on HEDIS measures sicians to implement an electronic record. MU is
in UMPC Health Plan Primary Care Network 20082010
being rolled out in three stages at UPMC. Stage 1
PCMH No PCMH (20112012): data capture and sharing; Stage 2
2008 2010 2008 2010 (2014): advance clinical processes (2016):
HbA1c tests (%) 84.1 90.7 82.7 86.1 improved outcomes. CMS provides incentive
Eye exams (%) 60.6 83.8 50.6 57.7 payments as MU parameters are adopted.
LDLc screening (%) 80.3 90.0 80.6 83.5 Approximately $44,000 is available per eligible
Nephropathy monitoring 82.6 89.4 75.4 80.2 provider. Hospitals may receive $2 million or
(%)
more. Currently, UPMC is in Stage 2 of MU,
PCMH= Patient-
Centered Medical Home though UPMC has put several ongoing initiatives
Data from Rosenberg et al. [59] in place to meet the all of the parameters of
MU. These included, but are not limited to, initia-
tives addressing quality of care, nance, and
Meaningful Use information technology. The primary barrier to
implementation continues to be resources includ-
Meaningful Use (MU), a programme adminis- ing funds, personnel, and support for rollout.
tered by the Center for Medicare and Medicaid While this barrier exists, UPMC has managed to
Services (CMS), is designed to improve quality, overcome these barriers and roll out the required
safety, efciency, and reduce health disparities. It elements.
also proposes to engage patients and their fami-
lies, improve care coordination, population and
public health while maintaining patient privacy. Infrastructure
The ultimate goal of MU is to improve clinical
outcomes, increase transparency and efciency, UPMC continues to move forward with initia-
empower individuals, and provide more robust tives to improve the quality of care and outcomes
research data on health systems [60]. The MU in people with diabetes. Over the course of 15
programme was also designed to incentivize phy- years of integrating care, the EpicCare medical
record system [61] was implemented in all CMI 2. LIPID Study Group. Long-term effectiveness and
safety of pravastatin in 9014 patients with coronary
practices. This system has core metrics for qual-
heart disease and average cholesterol concentrations:
ity of care. These metrics are directly tied to the LIPID trial. 2002.
incentives for the providers and the health sys- 3. The ALLHAT Ofcers and Coordinators for the
tem. The physicians need to successfully obtain ALLHAT Collaborative Research Group. Major out-
comes in moderately hypercholesterolemic, hyperten-
all of the metrics in order to receive the incen-
sive patients randomized to pravastatin vs usual care:
tives. This EHR has a tool called Healthy Planet the antihypertensive and lipid-lowering treatment to
which is a population-based tool for disease man- prevent heart attack trial. JAMA.
agement. To facilitate population management, 2002;288(23):29983007.
4. Ridker PM, Rifai N, Pfeffer MA, Sacks FM, Moye
there are currently patient registries for asthma,
LA, Goldman S, et al. Inammation, pravastatin, and
diabetes, preventive care and hypertension. the risk of coronary events after myocardial infarction
Registries for cardiovascular, renal, and liver dis- in patients with average cholesterol levels. Cholesterol
ease are in development. and Recurrent Events (CARE) investigators.
Circulation. 1998;98(9):83944.
CMI is also in the process of establishing
5. UK Prospective Diabetes Study Group. Tight blood
PODS to drive population management. These pressure control and risk of macrovascular and micro-
pods are determined geographically and serve vascular complications in type 2 diabetes: UKPDS 38.
approximately 25,000 patients. These Pods pro- BMJ. 1998;317:70313.
6. Intensive blood-glucose control with sulphonylureas
vide support to the primary care providers
or insulin compared with conventional treatment and
through care management teams. These teams risk of complications in patients with type 2 diabetes
consist of an advanced practice nurse, a behav- (UKPDS 33). The Lancet. 1998;352(9131):83753.
ioural specialist, and CDEs. This team works 7. Stark Casagrande S, Fradkin JE, Saydah SH, Rust KF,
Cowie CC. The prevalence of meeting A1C, blood
together with the primary care provider to address
pressure, and LDL goals among people with diabetes,
chronic conditions. Their focus is to decrease 19882010. Diabetes Care. 2013;36(8):22719.
variability in care, improve quality and give the 8. Janes GR. Ambulatory medical care for diabetes. In:
physician more time to focus on complex patients. Group NDD, editor. Diabetes in America. Bethesda:
National Institutes of Health; 1995. p. 54152.
There are currently 5 Pods with the eventual goal
9. American Diabetes Association. Standards of medical
of 20 Pods throughout the UPMC health system. care in diabetes, 2015. Diabetes Care. 2015;38 Suppl
In summary, diabetes care and integration of 1:S587.
nontraditional approaches to deliver that care 10. Bloom BS. Effects of continuing medical education
on improving physician clinical care and patient
have evolved over a 15-year periods at
health: a review of systematic reviews effects of
UPMC. There continues to be enthusiasm, sup- continuing medical education on improving physician
port, and dedication to improving the care and clinical care and patient health_0.pdf [Internet]. [cited
outcomes for people with diabetes using cutting 2016 Apr 19]. Available from: https://aetcnec.ucsf.
edu/sites/aetcnec.ucsf.edu/files/Effects%20of%20
edge strategies. There also continues to be chal-
continuing%20medical%20education%20on%20
lenges at the patient, provider and system levels; improving%20physician%20clinical%20care%20
however, innovative approaches are continually and%20patient%20health_0.pdf
implemented to meet the needs of the rapidly 11. VanNieuwenborg L, Goossens M, De Lepeleire J,
Schoenmakers B. Continuing medical education for
growing population of people with diabetes in a
general practitioners: a practice format. Postgrad Med
dynamic health-care environment. J. 2016;92(1086):21722.
12. Wagner EH, Austin BT, Korff MV. Improving out-
comes in chronic illness. Manag Care Q.
1996;4(2):1225.
References 13. Wagner EH, Austin BT, Korff MV. Organizing care
for patients with chronic illness. Mill Bank Q.
1. Hansson L, Zanchetti A, Carruthers SG, Dahlof B, 1996;74(4):51144.
Elmfeldt D, Julius S, et al. Effects of intensive blood- 14. Wagner EH, Bennett SM, Austin BT, Greene SM,
pressure lowering and low dose aspirin in patients Schaefer JK, Vonkorff M. Finding common ground:
with hypertension: principal results of the patient-centeredness and evidence-based chronic ill-
Hypertension Optimal Treatment (HOT) randomised ness care. J Altern Complement Med. 2005;11(sup-
trial. Lancet. 1998;351:175562. plement 1):s715.
15. Wagner EH, Glasgow RE, Davis C, Bonomi AE, rural primary care practice. Diabetes Educ.
Provost L, McCulloch D, et al. Quality improvement 2005;31(2):22534.
in chronic illness care: a collaborative approach. Jt 29. Siminerio L, Ruppert K, Huber K, Toledo
Comm J Qual Improv. 2001;27(2):6380. FGS. Telemedicine for Reach, Education, Access, and
16. Piatt GA, Orchard TJ, Emerson S, Simmons D, Treatment (TREAT): linking telemedicine with diabetes
Songer TJ, Brooks MM, et al. Translating the chronic self-management education to improve care in rural
care model into the community: results from a ran- communities. Diabetes Educ. 2014;40(6):797805.
domized controlled trial of a multifaceted diabetes 30. Ruppert K, Uhler A, Siminerio L. Examining patient
care intervention. Diabetes Care. 2006;29(4):8117. risk factors, comorbid conditions, participation, and
17. University of Pittssburgh Medical Center. UPMC physician referrals to a rural diabetes self-management
Facts and Figures [Internet]. 2015. Available from: education program. Diabetes Educ.
http://www.upmc.com/about/facts/numbers/Pages/ 2010;36(4):60312.
default.aspx 31. Tshiananga JKT, Kocher S, Weber C, Erny-Albrecht
18. University of Pittsburgh Medical Center Health Plan. K, Berndt K, Neeser K. The effect of nurse-led diabe-
UPMC health plan: our company [Internet]. 2014. tes self-management education on glycosylated
Available from: http://www.upmchealthplan.com/ hemoglobin and cardiovascular risk factors: a meta-
about/who-we-are/about-our-company.aspx analysis. Diabetes Educ. 2012;38(1):10823.
19. Brown SA. Studies of educational interventions and 32. Welch G, Zagarins SE, Feinberg RG, Garb
outcomes in diabetic adults: a meta-analysis revisited. JL. Motivational interviewing delivered by diabetes
Patient Educ Couns. 1990;16(3):189215. educators: does it improve blood glucose control
20. Norris SL, Lau J, Smith SJ, Schmid CH, Engelgau among poorly controlled type 2 diabetes patients?
MM. Self-management education for adults with type Diabetes Res Clin Pract. 2011;91(1):5460.
2 diabetes: a meta-analysis of the effect on glycemic 33. Gary TL, Genkinger JM, Guallar E, Peyrot M,
control. Diabetes Care. 2002;25(7):115971. Brancati FL. Meta-analysis of randomized educa-
21. Haas L, Maryniuk M, Beck J, Cox CE, Duker P, tional and behavioral interventions in type 2 diabetes.
Edwards L, et al. National standards for diabetes self- Diabetes Educ. 2003;29(3):488501.
management education and support. Diabetes Care. 34. Cooke D, Bond R, Lawton J, Rankin D, Heller S,
2013;36(Supplement_1):S1008. Clark M, et al. Structured type 1 diabetes education
22. Powers MA, Bardsley J, Cypress M, Duker P, Funnell delivered within routine care: impact on glycemic
MM, Hess Fischl A, et al. Diabetes self-management control and diabetes-specic quality of life. Diabetes
education and support in type 2 diabetes: a joint posi- Care. 2013;36(2):2702.
tion statement of the American Diabetes Association, 35. Cochran J, Conn VS. Meta-analysis of quality of life
the American Association of Diabetes Educators, and outcomes following diabetes self-management train-
the Academy of Nutrition and Dietetics. Diabetes ing. Diabetes Educ. 2008;34(5):81523.
Care. 2015;38(7):137282. 36. Toobert DJ, Glasgow RE, Strycker LA, Barrera M,
23. Brunisholz K, Briot P, Hamilton S, Joy E, Lomax M, Radcliffe JL, Wander RC, et al. Biologic and quality-
Barton N, et al. Diabetes self-management education of-life outcomes from the Mediterranean lifestyle
improves quality of care and clinical outcomes deter- program: a randomized clinical trial. Diabetes Care.
mined by a diabetes bundle measure. J Multidiscip 2003;26(8):228893.
Healthc. 2014;533. 37. Funnell MM, Anderson RM. Empowerment and self-
24. Weaver RG, Hemmelgarn BR, Rabi DM, Sargious management of diabetes. Clin Diabetes.
PM, Edwards AL, Manns BJ, et al. Association 2004;22(3):1237.
between participation in a brief diabetes education 38. Thorpe CT, Fahey LE, Johnson H, Deshpande M,
programme and glycaemic control in adults with Thorpe JM, Fisher EB. Facilitating healthy coping in
newly diagnosed diabetes. Diabet Med. patients with diabetes: a systematic review. Diabetes
2014;31(12):16104. Educ. 2013;39(1):3352.
25. Steinsbekk A, Rygg L, Lisulo M, Rise MB, Fretheim 39. Fisher L, Hessler DM, Polonsky WH, Mullan J. When
A. Group based diabetes self-management education is diabetes distress clinically meaningful?:
compared to routine treatment for people with type 2 Establishing cut points for the diabetes distress scale.
diabetes mellitus. A systematic review with meta- Diabetes Care. 2012;35(2):25964.
analysis. BMC Health Serv Res. 2012;12(1):213. 40. de Groot M, Doyle T, Kushnick M, Shubrook J,
26. Fan L, Sidani S. Effectiveness of diabetes self- Merrill J, Rabideau E, et al. Can lifestyle interven-
management education intervention elements: a meta- tions do more than reduce diabetes risk? Treating
analysis. Can J Diabetes. 2009;33(1):1826. depression in adults with type 2 diabetes with exercise
27. Ellis SE, Speroff T, Dittus RS, Brown A, Pichert JW, Elasy and cognitive behavioral therapy. Curr Diabet Rep.
TA. Diabetes patient education: a meta-analysis and meta- 2012;12(2):15766.
regression. Patient Educ Couns. 2004;52(1):97105. 41. Healy SJ, Black D, Harris C, Lorenz A, Dungan
28. Siminerio LM. Implementing the chronic care model KM. Inpatient diabetes education is associated with
for improvements in diabetes care and education in a less frequent hospital readmission among patients
with poor glycemic control. Diabetes Care. 51. Siminerio LM, Ruppert K, Emerson S, Solano FX,
2013;36(10):29607. Piatt GA. Delivering Diabetes Self-Management
42. Robbins JM, Thatcher GE, Webb DA, Valdmanis Education (DSME) in primary care: the Pittsburgh
VG. Nutritionist visits, diabetes classes, and hospital- Regional Initiative for Diabetes Education (PRIDE).
ization rates and charges: the urban diabetes study. Dis Manag Health Outcomes. 2008;16(4):26772.
Diabetes Care. 2008;31(4):65560. 52. Siminerio LM. Deploying the chronic care model to
43. Duncan I, Ahmed T, Li Q, Stetson B, Ruggiero L, implement and sustain diabetes self-management train-
Burton K, et al. Assessing the value of the diabetes ing programs. Diabetes Educ. 2006;32(2):25360.
educator. Diabetes Educ. 2011;37(5):63857. 53. Siminerio L, Ruppert KM, Gabbay RA. Who can pro-
44. Peyrot M, Rubin RR, Funnell MM, Siminerio vide diabetes self-management support in primary
LM. Access to diabetes self-management education: care? Findings from a randomized controlled trial.
results of National surveys of patients, educators, and Diabetes Educ. 2013;39(5):70513.
physicians. Diabetes Educ. 2009;35(2):24663. 54. Anderson RM, Fitzgerald JT, Gruppen LD, Funnell
45. Li R, Shrestha SS, Lipman R, Burrows NR, Kolb LE, MM, Oh MS. The diabetes empowerment scale-short
Rutledge S, et al. Diabetes self-management educa- form (DES-SF). Diabetes Care. 2003;26(5):16412.
tion and training among privately insured persons 55. Toobert DJ, Hampson SE, Glasgow RE. The sum-
with newly diagnosed diabetes United States, 2011 mary of diabetes self-care activities measure: results
2012. MMWR Morb Mortal Wkly Rep. from 7 studies and a revised scale. Diabetes Care.
2014;63(46):10459. 2000;23(7):94350.
46. American Medical Association (AMA) and the 56. Zgibor JC, Kuo S, Emerson S, Gittinger P, Solano FX,
National Committee for Quality Assurance (NCQA) Tilves D, et al. Rationale, design, and implementation
[Internet]. Adult diabetes: performance measures. of a cluster randomized trial using certied diabetes
AMA-convened physician consortium for perfor- educators to intensify treatment for glycemia, blood
mance Improvement; 014; Available from: http:// pressure and lipid control: REMEDIES 4D. Contemp
www.ama-assn.org/apps/listserv/x-check/qmeasure. Clin Trials. 2014;39(1):12431.
cgi?submit=PCPI 57. Siminerio L. Improving glycemia in primary care: a
47. Siminerio L, Zgibor J, Solano FX. Implementing the model integrating diabetes education into practice.
chronic care model for improvements in diabetes Boston: American Diabetes Association; 2015. 75th
practice and outcomes in primary care: the University Scientic Sessions.
of Pittsburgh Medical Center experience. Clin 58. American Academy of Family Physicians. The patient
Diabetes. 2004;22(2):548. centered medical home [Internet]. 2015. Available
48. Fitzgerald JT, Funnell MM, Hess GE, Barr PA, from: http://www.aafp.org/practice-management/
Anderson RM, Hiss RG, et al. The reliability and transformation/pcmh.html
validity of a brief diabetes knowledge test. Diabetes 59. Rosenberg CN, Peele P, Keyser D, McAnallen S,
Care. 1998;21(5):70610. Holder D. Results from a patient-centered medical
49. U.S. Department of Health and Human Services. home pilot at UPMC health plan hold lessons for
Healthy people 2010: understanding and improving broader adoption of the model. Health Aff (Millwood).
health. 2nd ed. Washington, DC: U.S Government 2012;31(11):242331.
Printing Ofce; 2010. 60. US Department of Health & Human Services. EHR
50. Krall J, Gabbay R, Zickmund S, Hamm ME, Williams incentives & certication [Internet]. 2015. Available
KR, Siminerio L. Current perspectives on psychologi- from: https://www.healthit.gov/providers-
cal insulin resistance: primary care provider and professionals/meaningful-use-denition-objectives
patient views. Diabetes Technol Ther. 61. Epic [Internet]. [cited 2016 Apr 19]. Available from:
2015;17(4):26874. https://www.epic.com/Software
Integrated Diabetes Care in Hong
Kong: From Research to Practice 5
to Policy
Roseanne O. Yeung, Junmei Yin,

and Juliana C.N. Chan
Introduction public hospitals and institutions, 47 Specialist

Outpatient Clinics, and 73 General Outpatient
Hong Kong is a unique densely populated metro- Clinics. Most recent estimates suggest the adult
politan city with heavy Western influences in a prevalence of diabetes is 9.9 % 1.
Chinese society. Given its colonial history, This chapter discusses the evolution of diabe-
Western acculturation has been occurring in tes care delivery at the Prince of Wales Hospital
Hong Kong before most other areas in China, and (PWH) and the Chinese University of Hong Kong
so it provides a window into the potential future (CUHK) into the current Joint Asia Diabetes
of the rest of China given its rapid economic and Evaluation (JADE) programme, a multipronged
industrial development over the last century. quality improvement initiative. It also highlights
Over seven million people live in the 1104 km2 the studies carried out at PWH and CUHK that
city-state, and health care is provided by both the have examined the data of Hong Kong citizens
public and private sectors. The public health-care with diabetes, and how this information has and
system is managed under the Hospital Authority continues to influence diabetes care delivery to
(HA), a statutory body that provides acute hospi- meet the local needs.
tal and ambulatory care services including 42
R.O. Yeung (*)

Division of Endocrinology, Department of Medicine, J.C.N. Chan
University of Alberta, Edmonton, Alberta, Canada Department of Medicine and Therapeutics,
The Chinese University of Hong Kong,
Hong Kong Institute of Diabetes and Obesity,
Hong Kong, China
The Chinese University of Hong Kong,
Hong Kong, China Hong Kong Institute of Diabetes and Obesity,
e-mail: [email protected] The Chinese University of Hong Kong,
Hong Kong, China
J. Yin
Department of Medicine and Therapeutics, International Diabetes Centre of Education,
The Chinese University of Hong Kong, The Chinese University of Hong Kong, Prince of
Hong Kong, China Wales Hospital, Satin, Hong Kong, China
e-mail: [email protected] e-mail: [email protected]

66 R.O. Yeung et al.
Clinical Care: The CUHK-PWH protocol was continued for another 4 years.
Integrated Diabetes Care During these 5 years, the study clinicians com-
Programme pared their experiences between usual care in the
HA clinics to that within the clinical trial clinic.
Adapting Lessons from Clinical Trials: Usual care was poorly coordinated and frag-
Protocol-Driven Care as the Standard mented, as the clinician could only spend an
of Care average of 812 min per patient and had little
support to educate and reinforce the patient on
PWH is the public teaching hospital affiliated lifestyle changes, treatment adherence and self-
with the CUHK and serves more than 1.2 million management. These patients were usually seen
people from all socio-economic backgrounds, every 46 months by different generalists or spe-
and is governed by the HA. Hong Kongs public cialists within a busy and crowded clinic setting.
health-care system is heavily subsidized where At each visit, blood pressure, fasting blood glu-
90 % of patients with chronic disease, including cose and fructosamine were measured by clinic
diabetes, receive care from the HA clinics and nurses. Annual assessments were recommended,
hospitals. All patients have access to medica- but not dictated. By contrast, with the salary of a
tions, investigations, and professional consulta- research nurse, the clinical trial clinic was set
tions for a nominal fee (US$10 per clinic visit, up in a 200 ft2 room within the teaching block
US$1.5 per drug item each lasting 34 months). where patients were seen by doctors according to
Since 1995, the HA has introduced a territory- the protocol with visits every four months once
wide Clinical Management System (CMS) which stable. On average, each consultation lasted
captures all critical clinical information including 1520 min with predefined assessment for
hospitalization, emergency room attendance, lab- advancement of therapy, as needed. The research
oratory investigations, and drug dispensing infor- nurse served as a coordinator and liaison between
mation. Authorized personnel can retrieve patient the doctors and patients. Apart from performing
data by using the Hong Kong Identity Card num- clinical assessments (e.g. blood pressure, body
ber, a unique identification number issued by the weight, laboratory sample collection), the nurses
Hong Kong Government to all Hong Kong also reinforced treatment plans and provided
residents. education and support which contributed towards
Led by CUHK clinician scientists and HA a trusting relationship between patients and care
health-care professionals, the Diabetes and team. Guided by the protocol, biochemical test-
Endocrine Centre in PWH is a multidisciplinary ing was more rigorous with measurement of gly-
centre that provides diabetes complication assess- cated haemoglobin and lipid profile at least every
ment, diabetes education and endocrine investi- 6 months and renal function including tests for
gations to patients referred to PWH. It is also a albuminuria at least annually. Annual foot and
training centre for medical doctors, nurses and retinal examination were also mandated in the
other health-care professionals on diabetes man- trial clinic.
agement and education. It is dedicated to provide After 5 years, the follow-up study demon-
comprehensive multidisciplinary care to patients strated the durability of these medications in
with diabetes through risk assessment and strati- blood pressure control with fewer clinical events,
fication, patient education and empowerment, which motivated the team to explore the broader
periodic monitoring to prevent diabetes-related impact of this structured clinical setting on
complications, and improving quality of life. patient outcomes [3]. After 7 years of implement-
In 1989, the CUHK researchers carried out a ing the enalapril versus nifedipine protocol, a
1-year randomized trial of enalapril versus nife- case-cohort study was undertaken to compare
dipine in 102 patients with type 2 diabetes and those in the trial versus those receiving usual care
hypertension [2]. Both groups had improvements [4]. The latter patients were not eligible for the
in blood pressure and albuminuria, so the study original trial due to lower blood pressure at the
5 Integrated Diabetes Care in Hong Kong: From Research to Practice to Policy 67
time (<160/90 mmHg). They were matched on a rienced hypoglycaemia during the previous 3
1:1 basis with all those in the trial on age, sex, months.
and duration of diabetes, among other clinical Blood pressure and anthropometric measure-
characteristics. At baseline, patients in the trial ment including body mass index (BMI) and waist
had higher prevalence of neuropathy and retinop- circumference (WC) were taken using standard-
athy. However, by the end of 7 years, the usual ized procedures. Visual acuity and retinal photog-
care group had higher blood pressures and fasting raphy were performed by trained nurses and later
blood glucose than those in the trial group. More technicians, and reviewed by endocrinologists
strikingly, 24.7 % of the usual care group died, for typical features of diabetic retinopathy.
compared to 8.8 % of those in the trial group with Standardized monofilament and graduated tuning
an adjusted relative risk of death of 0.21 (95 % CI fork examinations were used to detect sensory
0.07, 0.65, P = 0.006) in favour of the trial group. neuropathy. Blood and urine samples were col-
The team concluded that diabetes management lected for fasting plasma glucose, HbA1c, lipid
by protocol-driven care improved survival and profile (total cholesterol, low density-lipoprotein
clinical outcomes in patients. cholesterol LDL-C, high density-lipoprotein cho-
lesterol HDL-C and triglycerides), renal function
and urinary albumin-to-creatinine ratio (ACR).
Standardization of Workow
Given the large improvements seen with struc- Big Data to Drive Improvements:
tured care in this setting, the CUHK-PWH team The Hong Kong Diabetes Registry
created a standardized and structured intake pro-
cess based on the St. Vincents Declaration Recognizing the importance of quality data in
Eurodiab Protocol in 1995 to improve care deliv- improving care, the Hong Kong Diabetes Registry
ery [5]. Due to the shortage of manpower, nurses (HKDR) was established in 1995 using the data
were trained to perform protocol-driven diabetes from the structured assessments. Resources from
assessment and screening for common complica- both HA and CUHK were put towards meticu-
tions of diabetes. Changes were made to the lously prospectively collecting the data, and
workflow by creating two to three additional ses- clerks were hired to enter and manage the data
sions at the Diabetes Centre where 2025 patients for quality assurance purposes. This rich clinical
would have comprehensive assessments (CA) dataset has provided the means to examine clini-
carried out by one to two nurses and a few health- cal research questions in a timely and efficient
care assistants in each session. manner. This registry has led to many dozens of
In detail, patients attended the Diabetes Centre publications which provide numerous descriptive
for CA after fasting for at least 8 h. Trained analyses of the local diabetic population, which
nurses collected information on socio-economic previously had not been well defined. This
status, medical history, family history, medica- included the rarity of autoimmune diabetes even
tion use and adherence, lifestyle, and self-care in the young population; low BMI with propen-
behaviours from the patients using structured sity for central obesity; secular changes of clini-
preprinted forms. Significant medical history of cal outcomes, initially dominated by stroke and
co-morbidities was cross-checked with either the renal failure in the 1990s, coronary heart disease
referring physician and/or the CMS. Lifestyle in the 2000, and more recently heart failure and
factors included smoking status, self-monitoring cancer as survival continues to improve with bet-
of blood glucose (SMBG) frequency, exercise ter risk factor management and interventions.
frequency, and adherence to a balanced diet These epidemiological findings have subse-
within the previous 3 months. Hypoglycaemia quently been reported by many Asian investiga-
was assessed by asking patients if they had expe- tors which have led to the acceptance by the
scientific community regarding the Asian phe-
notype, or better referred as a phenotype in have contributed to the corporate strategy of the
transition, characterized by early onset of dis- HA, which is the major health-care provider in
ease with beta cell insufficiency, metabolic syn- Hong Kong, to develop career paths for nurse
drome, and a propensity for renal disease and specialists and establishment of diabetes centres
cancer 6. Given the paucity of randomized clini- and pharmacist-led adherence clinics to provide
cal trials in Asian populations, researchers have assessment, education and integrated care pro-
been able to replicate results from other parts of grammes in major public hospitals in order to
the world in order to validate questions in their reduce the burden of chronic disease. These pro-
own local population, in addition to examining grammes were also timely in light of the develop-
novel research questions. ment of the territory-wide CMS which enabled
Given the volume of patients in the clinics, the the HA to adapt the HKDR care protocol and
teams earliest work from the HKDR prioritized develop standard templates for data collection for
the development of prediction models, to allow all diabetes centres. All centres use the same tem-
for more efficient, data-driven risk stratification plate for diabetes assessment, derived from the
of patients. After accruing data for a decade on HKDR, which has now enrolled 350,000 patients
over 7000 patients, the team established 5-year with diabetes, accounting for 90 % of people
probabilities for major diabetes-related compli- diagnosed with diabetes in HK with a population
cations as defined by the International Code for of seven million and an estimated prevalence of
Diseases retrieved from the CMS. These included 10 %, where only 5070 % have been diagnosed.
end stage renal disease [7], stroke [8], coronary More recently, this diabetes complication screen-
heart disease [9], heart failure [10], and mortality ing service was extended to the publicly-funded
[11]. These risk equations have a 7090 % sensi- community-based family clinics for risk stratifi-
tivity and specificity of predicting outcomes cation and triage purposes, and evaluation has
based on the parameters collected in the registry. shown a reduction of mortality (adjusted HR
0.363; 95 % CI, 0.3080.428) using this pro-
gramme [12]. However, due to the heavy service
From Pilot Projects to Public Health demands and resource constraints, there are still
Programmes in the Hospital variations among different hospitals and clinics
Authority (HA) in delivering diabetes care despite the same
Electronic Medical Record (EMR) templates. As
Among the over 40 public hospitals run by HA, a teaching hospital, the CUHK-PWH team con-
there are 17 public-funded diabetes centres coor- tinues to leverage the academic resources includ-
dinated by diabetologists and nurse educators ing research funding and postgraduate students to
that provide two to three weekly sessions of com- test pilot programmes and use research results to
plication screening to all patients referred by the influence clinical practice. Here are a few exam-
public hospitals and community-based clinics. ples of how research data can be used to change
The benefits of team-based periodic comprehen- practice and influence policy.
sive assessments, disease management and peer
support programmes, designed by specialists and Medication Adherence and Follow-Up
coordinated by trained nurses, pharmacists and In 1998, the CUHK-PWH team carried out a
health-care assistants on clinical outcomes 2-year randomized trial evaluating the impact of
(death, cardiovascular-renal disease, hospitaliza- a pharmacist-led telephone counselling interven-
tions) had been rigorously evaluated in peer- tion to promote medication adherence on mortal-
reviewed publications. These results were ity in non-adherent patients prescribed at least
presented at hospital management conferences five chronic medications attending the usual care
with mass media coverage to inform payers and specialist clinics [13]. As part of her MPhil the-
the public. Along with efforts from other col- sis, a pharmacist identified eligible non-adherent
leagues, these quality improvement programmes patients based on a 2030 min medication assess-
ment interview, and defined adherence as taking design of the JADE programme (see later section
80120 % of the prescribed daily medications. in this chapter).
Adherence was assessed at screening, random-
ization, and after 2 years. The intervention con- Negative Emotions
sisted of a pharmacist making six to eight phone The lifelong commitments to medication adher-
calls lasting 1015 min per call to the patient ence and lifestyle modification make diabetes
between physician visits. The pharmacist clari- self-management both physically and emotion-
fied misconceptions of medications, asked about ally taxing. The psychological burdens result
side effects, reinforced the importance of adher- from insulin injection, self-monitoring of blood
ence, and provided basic counselling on self-care glucose, dietary restriction, as well as fear of
and lifestyle management. Patients were encour- complications, which may significantly increase
aged to report side effects, self-initiated changes negative emotions in patients with diabetes.
in medication, and concerns to their attending Depression, anxiety, and distress are prevalent
doctors. The control group received no phone mental afflictions found in patients with diabetes
calls. After 2 years, 38 (17 %) patients had died in [1619]. In a survey involving 189 patients with
the control group compared to 25 (11 %) in the type 2 diabetes attending the PWH clinic,
intervention group (adjusted relative risk: 0.59, 2050 % experienced diabetes-related distress
95 % CI 0.350.87, P = 0.039). Notably, half of using validated questionnaires. These symptoms
the patients who initially consented but defaulted of distress closely correlate with obesity and
on follow-up died within the 2-year period. These HbA1c [20]. Another study with a consecutive
defaulters had similar clinical and demographic cohort of 586 outpatients with type 2 diabetes
characteristics as the patients who were followed, showed that the prevalence of depression was
except that their adherence scores were lower at 18.3 % in Hong Kong Chinese patients with type
baseline. 2 diabetes. Furthermore, depression was associ-
In another PhD project, the pharmacist candi- ated with poor glycaemic control and self-
date implemented a structured care programme reported hypoglycaemia, in part due to poor
in patients with types 2 diabetes and renal impair- adherence [21]. With the aid of the HKDR, the
ment and demonstrated the marked benefits in team conducted a prospective study involving
reducing cardiovascular disease and end stage 7835 patients with type 2 diabetes without car-
renal disease after 2 years compared to usual care diovascular disease (CVD) at baseline, and found
[14]. These results were replicated in a multicen- that 153 patients were diagnosed with major
tre study where the use of a doctor-nurse team, depression by psychiatrists in public hospitals.
guided by a protocol with predefined care pro- After adjusting for conventional risk factors,
cesses and treatment targets (A1c <7 %, BP depression was independently associated with a
<130/80 mmHg, LDL-C <2.6 mmol/L, triglycer- two to threefold increase in the risk of incident
ide <1.7 mmol/L, persistence of renin angioten- CVD [22].
sin system (RAS) inhibitors), increased the rate
of attaining 3 targets by threefold (61 % versus Treatment Gaps in Young Onset
28 %). Patients who attained 3 targets was Diabetes
translated into relative risk reduction of 0.43 As diabetes prevalence increases in younger pop-
(95 % CI: 0.210.86) for end stage renal disease ulations and based on clinical observations
after 2 years [15]. These findings reinforced the regarding the poor outcomes in these subjects,
importance of using protocols and frequent fol- the CUHK-PWH team used the registry to
low-up to treat to multiple targets and reinforce explore the impact of young-onset diabetes, a
treatment adherence in these patients with silent therapeutic challenge which has not been well
disease, which have provided the rationale for the defined. Using the HKDR, the team reported that
one in five patients with type 2 diabetes were
diagnosed before the age of 40. Compared to Critical importance of albuminuria and renal
those with later onset of disease (40 years), dysfunction as prognostic markers for
patients with young-onset diabetes had worse cardiovascular-renal complications and their
metabolic control, under-prescription of life sav- multiple determinants including metabolic
ing drugs such as statins and RAS inhibitors, syndrome [39]
with a higher cumulative event rate driven by lon- Association of HbA1c variability (mean stan-
ger disease duration [23]. These data also indi- dard deviation) with development of chronic
cate that the rate of cardiovascular-renal kidney disease and cardiovascular disease
complications in young type 2 diabetes patients independent of mean HbA1c and other con-
have eclipsed those in type 1 diabetes, due to the founders [40]
amplifying effects of silent risk factors including Association of severe hypoglycaemia requir-
obesity, hypertension, and dyslipidaemia [24]. ing hospitalization with increased mortality
The focus on preventing and delivering early pre- mainly due to cancer and chronic kidney dis-
ventive care has sparked the team to conduct ease [41]
community outreach programmes to engage Association of cancer risk with (1) copresence
tech-savvy younger people to assess and monitor of low triglycerides and low LDL-C and (2)
their risk for diabetes using mobile phone appli- copresence of high HDL-C and low LDL-C
cations (see section in this chapter on Yao Chung plus albuminuria, suggesting the importance
Kit Diabetes Assessment Centre for discussion of of dysregulation of lipid metabolism in cancer
Community Outreach). development in type 2 diabetes [42]
Patients with low C-peptide who received
Genetic Risk insulin had lower cardiovascular events and
The establishment of the HKDR was accompa- mortality than those on insulin with high
nied by a biobank from consenting participants, C-peptide highlighting subphenotype-
and has allowed for novel genomic research (see treatment interactions and the possible use of
section in this chapter on CUHK Diabetes: C-peptide to guide therapy [43]
Genomics Research and Biobanking for further Both white blood cell count and haematocrit
details). Areas of active research include discov- were independent predictors of mortality in
ery for genetic markers of diabetes and its com- type 2 diabetes which raise the hypothesis
plications including cancer. Apart from regarding the roles of inflammation and eryth-
conducting linkage analysis using family-based ropoiesis in diabetes [44, 45]
cohorts recruited through the registry [25] and
validating the first wave of genetic variants dis- These findings help to inform the pathophysi-
covered in genome wide association studies in ological understanding of this complex disease
our Chinese population [26], the group has also by better characterizing phenotypes, and provid-
used these resources to discover genetic variants ing the premise for formulating hypothesis for
associated with increased cancer risk [27], end mechanistic and interventional studies with high
stage renal disease [28], cardiovascular-renal dis- clinical relevance.
ease [2933], and young-onset diabetes [3437],
some of which have been validated in a recent Cancer and Diabetes
meta-analysis [38]. Diabetes has been associated with increased can-
cer risk, but the underlying mechanism is poorly
Biochemical Markers understood. The linkage between the longitudi-
The registry contains many biochemical parame- nal clinical data within the HKDR and the cancer
ters linked to the development of complications, outcome data in the CMS has provided important
and have allowed exploration of a number of observational findings to help elucidate these
associations that are informing areas of future connections. Detailed pharmacoepidemiological
interventional research. Examples include: analyses revealed attenuated cancer risk in
patients treated with insulin and oral antidiabetic The JADE Programme: Evolving
drugs compared with non-users of these drugs Health Information Technology
[4648]. The team has further observed signifi-
cant drug-subphenotype interactions with attenu- In 2007, the Asia Diabetes Foundation (ADF)
ated cancer risk in: was founded as a non-profit research-promoting
entity under the governance of the CUHK
1. metformin users with low HDL-C Foundation to transfer this large body of knowl-
2. RAS inhibitor users with high WBC count edge through technological enhancements to fur-
3. statin users with copresence of low LDL-C ther improve the efficiency of care delivery and
plus albuminuria or low triglyceride [49, 50]. conduct of research. The ADF is a charitable
organization dedicated to develop and validate
These observations corroborate with experi- innovative chronic disease management models
mental findings of possible consequences of with the aim to make quality care sustainable,
hyperglycaemia on dysregulation of cholesterol affordable and accessible. The prime achieve-
metabolism as well as activation of RAS and ment of the ADF is the development of the JADE
adenosine 5-monophosphate-activated protein programme, a state-of-the-art web-based infor-
kinase pathways, all of which may be implicated mation technology that incorporate diabetes care
in cancer development. In support of these protocols, validated risk engine, and clinical
notions, the registry was used to confirm the decision support tools to maximize the delivery
additive effects of optimal glycaemic control and of quality care and enable both health-care pro-
use of RAS inhibitors and statins on reduced can- viders and patients to make informed decisions.
cer risk [51]. By combining the experimental and Supported by an educational grant, an endo-
epidemiological observations, the CUHK-PWH crinologist was seconded for 2 years to develop
diabetes team has formulated the hypothesis that the JADE programme, alongside a programming
early identification and optimization of multiple team and a project coordinator. The JADE pro-
risk factors including early use of statins and gramme began enrolling patients in 2007 as a
RAS inhibitors may normalize the internal milieu quality improvement initiative to engage both
to reduce cancer risk in type 2 diabetes, although physicians and patients. The JADE software con-
large-scale, randomized clinical trials will be sists of a web-based portal that uses technology
needed to confirm these hypothesis [52]. to digitize risk stratification and protocol-driven
care. The existing paper templates based on pro-
Diabetes and Chronic Hepatitis B tocols recommended by international bodies
Infection were transformed into online forms, and the data-
Given the high prevalence of chronic hepatitis B base was upgraded to a larger, more robust sys-
infection in the region, the CUHK-PWH diabetes tem. JADE assessment relies on a specified
team was among the first to document the clinical workflow that integrates multiple care
increased risk of renal and cardiovascular com- components to enable health-care providers to
plications in people with chronic hepatitis B establish a diabetes registry with built in matrixes
infection and diabetes compared to having diabe- for documentation of key performance indexes.
tes alone [53, 54]. Furthermore, the rich dataset The key features of this web-based quality
allowed the team to investigate the impact of improvement programme include (Figs. 5.1 and
treatment of risk factors in hepatitis B infected 5.2):
patients and found that the use of statins and
insulin were associated with greatly reduced risk 1. Templates to guide standardized assessment
of development of hepatocellular carcinoma [55], workflow and data capture
providing the basis for designing clinical trials in 2. Validated risk equations to estimate 5-year
this area. probability of major clinical events
Fig. 5.1 Integrated diabetes

care protocol recommended by
the Plan of Action (2006
2010) for the Western Pacific
Declaration on Diabetes
(WPDD) (Adapted by the
Joint Asia Diabetes Evaluation
programme, designed by the
Asia Diabetes Foundation
(ADF). The protocol was
based on Alberti et al. [74];
and Asian-Pacific Type 2
Diabetes Policy Group. Type 2
diabetes practical targets and
treatments. International
Diabetes Institution and In
Vivo Communications Asia
Pte Co. Ltd., 2005)
3. Personalized reports with graphs and bar 5. Built in matrixes which displays the propor-
charts illustrating risk categories, current con- tions of patients attaining various treatment
trol of risk factors with easy communication targets and risk categories and their changes
of target values, notably A1c, blood pressure, over time
LDL-C (ABC) and body weight
4. Individualized decision support for both doc- These data are entered into a computer pro-
tors and patients, triggered by attained ABC gramme with predefined definitions of risk fac-
targets and body weight, to empower self- tors, complications and treatment targets to
management, reduce clinical inertia, and pro- generate a one-page summary report which
mote shared decision-making enable the physicians to efficiently identify gaps
entered by clerical staff to be mailed to the refer-

ring physician and patient to promote shared
decision-making. A prospective JADE research
registry has been established through these peri-
odic assessments.
Facilitating Knowledge Transfer

and Clinical Decision-Making
The process of knowledge transfer in clinical

medicine includes multiple steps and stakehold-
ers, the simplest being:
(i) The patient conveys information to one or

more health-care providers
(ii) The health-care provider(s) investigates
with additional history, physical examina-
tion, or specialized investigations
(iii) The health-care provider(s) synthesizes the
data to determine a potential diagnosis and
treatment plan
(iv) The health-care provider(s) shares the
potential diagnosis and treatment plan with
the patient
(v) The patient and health-care provider(s)
engage in shared decision-making to carry
out plan
Each health-care system has its unique chal-

lenges of knowledge transfer and communica-
tion. Nevertheless, the challenges faced by the
Hong Kongs public health-care system, mod-
elled after the UK National Health System, are
likely to be shared by many developing areas in
Asia where universal health-care coverage for
chronic disease like diabetes is becoming an
urgent government priority. During the life jour-
ney of a person with diabetes, s/he will have mul-
tiple encounters with many health-care
professionals including primary care physicians,
Fig. 5.2 A sample of comprehensive assessment (CA) specialists, nurses, other allied health profession-
report for patient generated by the Joint Asia Diabetes als working in different settings including private
Evaluation programme (Used with permission of Asia
and public as well as hospital and community
Diabetes Foundation, Hong Kong)
clinics. Failed communication at any of these
steps may result in the patient slipping through
in treatment and make recommendations for each the cracks of the fragmented care continuum.
patient. Depending on the setting and workflow, The CUHK-PWH team recognizes the chal-
these data collected by nurses or doctors, can be lenges at each step and continues to make great
efforts, based on feedback from users, to enhance the JADE portal based on the published and
the functionality of the JADE portal and design internally validated risk equations for diabetes
the JADE report to promote collaborative, multi- complications. A simple, practical clinical
disciplinary care, resulting in the creation of two assessment consisting of four JADE risk catego-
JADE reports: one for the health-care provider, ries was based on the learnings from the HKDR
and one for the patient. including the various risk equations (Fig. 5.3).
These internally-validated risk categories
respectively predict an annual all-event rate of
Communicating 8 % (level 4, very high risk), 5 % (level 3, high
Between Professionals: JADE Health- risk), 3 % (level 2, moderate risk), 1 % (level 1,
Care Provider Reports low risk). The CA report displays the patients
5-year probabilities of clinical events (stroke,
The JADE Health-Care Provider Report addresses coronary heart disease, heart failure and ESRD)
the first three steps of the knowledge transfer based on these validated risk equations.
process: Based on the risk category, a physician must
then identify appropriate unmet gaps to prevent
(i) The patient conveys information to one or development or worsening of disease by control-
more health-care providers, ling the four most modifiable risk factors (ABC
(ii) The health-care provider(s) performs his- and body weight). In order to help physicians
tory taking, physical examination, or spe- identify evidence-based gaps and reduce clinical
cialized investigations, inertia, clinical decision support based on the
(iii) The health-care provider(s) synthesizes the International Diabetes Federation (IDF) clinical
data to create a problem list and treatment practice guidelines were integrated into the
plan. reports.
In addressing the first two steps, the team

established the minimum data set required for Communicating with the Patient:
collection using the template-guided process of JADE Patient Reports
information gathering between the patient and
the health-care providers. By standardizing the Multi-target clinical care focusing on glycaemic
type of information gathered and the format in control, blood pressure, and lipids had been
which the information is conveyed, all team shown to reduce the development of diabetes-
members can quickly and easily find and inter- related complications and mortality in the land-
pret information, decreasing the amount of mark STENO-2 study [56]. The findings, based
searching through records. on data collected in a controlled randomized clini-
Once the information is collected, the chal- cal trial setting, were subsequently confirmed in
lenge lies in synthesizing the data for diagnostic the HKDR which showed that attainment of ABC
and treatment purposes given the amount of targets reduced events in a real-world setting [57].
medical knowledge required and the ever- To promote health literacy and empowerment,
expanding scientific literature. One of the big- patients are informed about their ABC and body
gest clinical obstacles is properly risk-stratifying weight targets with decision support to help them
patients to determine the appropriate level of achieve these goals.
intervention. The CUHK-PWH team capitalized The JADE patient report was created in order
on the local knowledge generated from the to complete steps 4 and 5 of the knowledge trans-
HKDR, and built a risk stratification engine in fer process:
Fig. 5.3 (a, b) KaplanMeier

analysis of survival and
clinical event rates in 7534
Chinese type 2 diabetic
patients categorized into
different risk levels using the
Risk Engine of the Joint Asia
Diabetes Evaluation (JADE)
programme. Log rank test:
P < 0.001 in both analyses.
Risk level: level 1, low risk
(n = 452), level 2, moderate
risk (n = 1468), level 3, high
risk (n = 4476), level 4, very
high risk (n = 1138) (Reprinted
with permission from Chan
et al. [75])
(i) The health-care provider(s) shares the poten- The JADE report was designed to educate and
tial diagnosis and treatment plan with the empower the patient to self-manage with clear
patient goal-directed feedback. After electronic entry of
(ii) The patient and health-care provider engage the data collected during the annual or biannual
in shared decision-making to carry out plan CA, the JADE portal generated a CA report for
the patient (Fig. 5.2). This report shows the 0.48 % reduction in mean HbA1c [63]. Moreover,
results of the assessments with emphasis on key patient empowerment, which motivates patients
modifiable risk factors including HbA1c, BP, with diabetes to actively participate in decision-
LDL-C and body weight, and trend of these risk making regarding their self-care rather than sim-
factors over time in a graphic representation. The ply comply with physicians instruction, is a
ABC targets were set at HbA1c <7 %, BP <130/80 further step to increase ones ability to think criti-
mmHg and LDL-C <2.6 mmol/L, based on the cally and act autonomously.
IDF recommendations [58]. The patient report In order to efficiently empower patients, group
also includes the risk stratification engine, and classes were created to educate and reinforce
informs the patient of their 5-year probabilities of patients on the basic of diabetes management.
clinical events. The report also contains practical Patients are asked to return to the Diabetes Centre
suggestions in layman terms to promote patient 68 weeks after their initial CA to collect their
self-care and medication adherence in order to reports and attend a diabetes empowerment class
reach multiple treatment targets. led by diabetes nurse specialist. During this 2-h
The JADE e-portal also has the functionality group session, the patients are informed of their
to generate follow-up (FU) reports to help track assessment results and educated about self-
quality of diabetes care with decision support management knowledge on diet, medication,
between CAs. As long as the key modifiable risk physical activity, self-monitoring of blood
factors (HbA1c, BP, LDL-C and body weight) glucose, and psychological health. They are also
are assessed and entered into the portal, a person- taught to interpret the JADE reports and
alized FU report displaying the patients trends of emphasize the importance of attaining multiple
ABC (HbA1c, BP, LDL-C) control and body treatment targets. The nurse uses both didactic
weight with individualized reminders for self- and interactive approaches to teach and motivate
care can be generated as a reinforcement tool to the patients to be more engaged in
empower the patient. This FU report provides an self-management.
important tool to track progress and promote This care delivery model combining logistics,
ongoing dialogue between patients and health- knowledge transfer, and information technology
care professionals for shared decision-making. enables integrated and holistic care to patients
with diabetes (Fig. 5.4). Starting with the primary
care physician or other referring physician, a
Closing the Loop: Patient Education patient with diabetes is identified and referred for
CA by the specialist team. A multidisciplinary
Among the many challenges of patient self- team performs the CA and inputs the results into
management, lack of education and empower- the JADE portal, generating the two personalized
ment are the two most cited barriers [59]. feedback reports as mentioned above. Patients
Sufficient knowledge is unquestionably impor- are asked to return to PWH 46 weeks after the
tant in self-care, especially in people with low CA to attend a nurse-led empowerment class
health literacy and limited access to diabetes edu- where they are educated about self-management
cation. Several systematic reviews showed that knowledge and taught to interpret the JADE
self-management education with comprehensive patient report. Patients follow up with their pri-
lifestyle interventions improved glycaemic and mary care physicians, who are empowered with
cardiovascular risk factor control [6062]. In a the JADE professional report recommendations
meta-analysis, quality improvement measures and arrange follow-up with the patients at appro-
targeted at patients, systems, and care providers priate time periods based on risk assessment (the
all had positive impacts on metabolic control majority being 34 months), where higher risk
among patients with diabetes with patient educa- patients are booked in more frequently. Periodic
tion focusing on personalized goal-setting and specialist-reviewed CA are suggested every
action planning, having the largest effect size of 1224 months to help primary care physicians
Fig. 5.4 Workflow of the multicomponent integrated diabetes care programme at the Chinese University of Hong
Kong Prince of Wales Hospital International Diabetes Federation Centre of Education, Hong Kong
and patients stay connected to the latest evidence- tion the validity of the data given the industrys
based treatments, clinical trial opportunities, and inherent conflict of interest over profitability ver-
specialty programmes. sus best patient care. As more medications come
to market, there is a growing need for compara-
tive effectiveness studies looking at medications
Innovation in JADE: More Efcient within the same class and multiple medication
Data Infrastructure for Clinical Trials class combinations, yet this research is difficult,
time consuming, and often unpalatable to indus-
Prospective cohort studies are always limited by try funding because of unclear benefits. And
bias, and evidence-based medicine requires when evaluating cost-effectiveness, many of the
planned, prospective evaluation of interventions existing studies are limited due to forced assump-
to demonstrate efficacy. However, the costs of tions, since robust longitudinal clinical data are
obtaining efficacy data in diabetes are growing often unavailable [66]. Furthermore, few systems
exponentially as regulatory bodies demand data are in place for quality ongoing clinical surveil-
demonstrating superiority over existing treat- lance once medications are approved.
ments, long-term non-inferiority data in regards
to cardiovascular outcomes, and cost- Data Collection
effectiveness [64, 65]. At present, many trials are Clinical trials are expensive because of the detail
funded by industry as major funding bodies con- and depth of data required on each patient, which
tinue to have budget cuts, leading many to ques- often require separate databases to be developed
outside of the usual-care electronic medical interactions in the diabetes education class.
records or paper-based chart systems. These These observations led to the development of
databases must be built, managed, and main- the Peer support, Empowerment, And Remote
tained from scratch every time, often requiring communication Linked by information technol-
double-entry of data by research staff. The JADE ogy (PEARL) Study, based on a Train-the-
programme provides a more efficient means of Trainer model [68]. The team identified and
collecting the key clinical variables in its com- invited friendly and knowledgeable patients with
prehensive assessments, and allows researchers type 2 diabetes with an HbA1c less than 8 % to
to add new fields as necessary for research pur- become peer supporters. A multidisciplinary
poses. This obviates the need for redundant entry team designed a curriculum of training, which
into non-clinical systems, as the JADE pro- consisted of four 8-h workshops incorporating
gramme is simultaneously a clinical care tool and tutorials, case sharing, reflections, role-playing,
prospective database. and games. Training focused on basic diabetes
self-management information on diet, exercise,
Recruitment poor sleep, stress, changes in daily routines, body
The identification of patients for clinical trial weight, medications, concurrent illnesses, and
recruitment starts with informing either health- the importance of self-monitoring of blood
care professionals or patients of the trial inclu- glucose. Training was also provided regarding
sion criteria. Professionals must identify patients communication and empathic listening. Peer sup-
by relying on memory while sifting through porters were encouraged to share their positive
paper charts, or using filtering capabilities if experiences to assist their peers to manage diabe-
available in existing electronic medical records. tes on a day-to-day basis.
A large number of trials fail because of inade- All patients undergoing comprehensive
quate recruitment [67]. The JADE programme assessments were invited to join the study, and
has allowed for ready identification of eligible those who agreed were randomized to receive
clinical trial participants because of its detailed either peer support or usual care within the JADE
clinical database. programme where patients received periodic CA
and quarterly FU reports through the mail. After
Longitudinal Benet 1 year, those managed by the JADE programme
One of the greatest challenges in clinical trials is had significant improvement in risk factor control
maintaining the contact between researchers and through reduced clinical inertia and improved
patients over many years. By pairing a study self-care. Although patients who received addi-
platform with clinical care, JADE facilitates tional peer support did not have further improve-
long-term contact with the patient, as part of rou- ment in cardio-metabolic control, they had
tine periodic follow-up. This also allows research- reduced hospitalization rates after one year of
ers to evaluate longer term outcomes than many intervention. In a post hoc analysis, patients with
previous trials, given the great expense in main- significant negative emotions who had poor
taining databases for the tracking of longitudinal cardio-metabolic risk factor control such as
outcomes. hyperglycaemia, high BMI, and CKD, benefited
most from peer support with hospitalization rate
similar to those without negative emotions in
Case Example of Enabling Research whom peer support did not have effects on hospi-
and Quality Improvement: PEARL talization. In these high risk patients with multi-
Study ple risk factors and complications as well as
negative emotions, peer support further improved
Clinicians on the CUHK-PWH team recognized psychological well-being and drug adherence.
the potential of empowering people with diabetes Furthermore, the peer supporters showed sus-
to help each other after seeing some of the positive tained glycaemic control, and improvements in
self-care and health-related quality of life over 4 learned into a clinical effectiveness setting, and
years of being involved with the programme [69]. adhering to principles of continuous quality
Lessons learned from that trial have allowed improvement.
for the development of the second iteration of the
PEARL study, currently underway. In the second
iteration, there has been the additional systematic CUHK Diabetes Clinical Research
evaluation of providing periodic personalized FU Centre
reports on metabolic control and hospitalization
in patients with both type 1 and type 2 diabetes Established since 1999, the CUHK Diabetes
managed by the JADE programme. It includes Clinical Research Centre has conducted nearly
almost all the patients referred to the PWH 100 phase two to four clinical trials of novel com-
Diabetes Centre for annual CA from February 26 pounds and devices, to address unmet needs in
to December 1, 2013. A group of high-risk diabetes, obesity, cardiovascular and renal dis-
patients was offered peer support and random- eases. Conducted trials included pharmaceutical-
ized to receive two JADE FU reports by mail sponsored multicentre international studies and
after their clinic visits. Patients not offered peer investigator-initiated studies. Apart from address-
support were also randomized to receive two ing the primary research questions, execution of
JADE FU reports. The primary outcome was clinical trials provide direct clinical benefits
change in HbA1c at month 12, as well as rate, through education, intensified monitoring and
frequency, and length of stay of hospitalization structured care. Established principles of the
during the 12 months. The secondary outcomes research centre include:
include risk factor control, attainment of treat-
ment targets, rate and frequency of emergency (i) Protecting human rights
room visits, and changes in cognitive- (ii) Ensuring that studies are conducted in
psychosocial-behavioural parameters. The pri- accordance to Good Clinical Practice
mary results show that patients who received (iii) Ensuring data quality and integrity for pub-
additional FU reports had greater reduction in lic dissemination
HbA1c compared to those under usual care, and (iv) Adhering to Standards of Operation
patients with peer support had further improve-
ments in psychosocial well-being, self- Apart from fundraising to support other
empowerment, and quality of life [70]. This research programmes, the Diabetes Clinical
programme forms the basis of another PhD the- Research Centre also serves as a training centre
sis, which highlights the benefits of using an where fellows, nurses, health care, research and
academic-health-care institution-foundation administrative personnel work in a collaborative
partnership to generate new knowledge and and cohesive manner to gather and translate evi-
inform clinical practice, as advocated in the dence to clinical practice through a bedside-to-
emerging field of improvement science [71]. bench-to-beside approach.
The third iteration of the PEARL programme
is in the planning stages. However, this serves as
an example of an iterative quality improvement CUHK Diabetes: Genomics Research
programme that has been greatly augmented by and Biobanking
the JADE data collection platform, allowing for
quality research at reduced cost given much of Since 1994, the unit has embarked on an over-
the data platform and clinical processes are arching strategy to combine epidemiology,
already in place. In this programme, the team has applied genomics, clinical trials, and transla-
been able to translate knowledge from the clini- tional research with the ultimate goal to use clini-
cal efficacy setting of a randomized controlled cal, biochemical, and genetic markers to identify
trial, to adapting and implementing lessons high risk subjects for early intervention and pre-
vention of complications. A biobank was estab- cuhk.edu.hk/Programmes/MScinEndocrinolo

lished, containing multiple prospective gy,DiabetesandMetabolism/
case-control, family-based, adolescent/youth and OverviewandObjectives.aspx
mother-offspring cohorts. These interlinking Diploma in General Endocrinology and
research programmes which aim to discover Metabolism since 2004
markers to predict diabetes and its complications, Diabetes Preventing the Preventable Forum
have formed the basis of large scale epidemio- held annually since 2011 http://www.idfce-hk.
logical studies to examine the prevalence and org/dpp2015/
natural history of childhood obesity, multiple
forms of diabetes (e.g., young-onset diabetes,
gestational diabetes), and possible interventions. Yao Chung Kit Diabetes Assessment
This rich dataset has also facilitated global and Centre
region-wide research with the CUHK-PWH team
being a key member in the Global Diabetes In 2007, the Yao Chung Kit (YCK) Diabetes
Consortium funded by the National Institute of Assessment Centre http://www.yckdac.hkido.
Health in the United States and the Asian Genetic cuhk.edu.hk/en/index.html was established
Epidemiology Network (AGEN) Consortium, through a generous donation from the Yao Yiu
which, for example, has enabled the discovery of Sai Education and Charitable Memorial Fund to
new loci for type 2 diabetes in East Asians [72]. the CUHK to increase the accessibility, afford-
ability and sustainability of this much needed
risk stratification programme to benefit the
Building Professional Capacity: growing population of patients with diabetes in
Hong Kong Institute of Diabetes the community. Given the dual private and pub-
and Obesity lic health-care systems in Hong Kong, the YCK
Diabetes Assessment Centre offers an affordable
The Hong Kong Institute of Diabetes and Obesity private option for expedited care, as waiting lists
(HKIDO) is an education and research institute for a CA at the HA Diabetes Centre can be up to
established under the CUHK in 2005, funded by a year or more. Services provided in YCK
donations, grants, contract research, to deliver Diabetes Assessment Centre include compre-
training and education courses to health-care pro- hensive JADE-based diabetes CA and risk evalu-
fessional from both locally and in the Asia region, ation, with available 24 h ambulatory blood
with the aim to strengthen professional capacity. pressure monitoring and 24 h continuous glu-
In turn, proceeds from these conferences and cose monitoring. Lifestyle-focused diabetes edu-
courses have provided the much needed seed fund- cation classes are also provided, including
ing to sustain the continuous data analysis and exercise workshops, cooking classes, and diabe-
genomic programme which are not inexpensive. tes conversation map workshops. Referrals can
The key education programmes and confer- be made by physician or self-referral by the
ences conducted include: patient, allowing motivated patients readier
access to services. By supporting private doctors
Hong Kong Diabetes and Cardiovascular Risk in the community to provide a more holistic,
Factors East-Meets-West Symposium held quality-assured and affordable care, the YCK
annually since 1999 http://www.hkido.cuhk. Diabetes Assessment Centre aims to provide an
e d u . h k / S y m p o s i u m s Wo r k s h o p s / alternative option to patients who can afford a
EMWSymposium2015/WelcomeMessage.aspx more personalized service to reduce the growing
Diploma in Diabetes Management and burden on the public system, while at the same
Education since 2002 time building a network of like-minded doctors
Masters Course in Endocrinology, Diabetes and care professionals to improve diabetes care
and Metabolism since 2004 http://www.hkido. in the community.
Community Engagement: OPAL comparisons were made between those with

and RUBY young-onset diabetes (diagnosis before age 40)
and late-onset diabetes (diagnosis at age 40 or
Supported by university and government grants, older) [16]. Of the 41,029 patients with data
the YCK Diabetes Assessment Centre takes on an available, 18 % had young-onset diabetes. The
additional role of reaching out to the community young-onset group had longer disease duration
to raise awareness and detect high risk subjects than those with late-onset diabetes (10 versus 5
for early intervention. In the nurse-led Outreach years), with worse glycaemic control (mean
Program to raise Awareness and Lifestyle HbA1c 8.32 % [SD 2.03] vs 7.69 % [1.82];
Modification (OPAL), volunteers, peer support- p <0.001). Despite their worse risk profiles, these
ers and health-care workers use simple tools and young onset patients were also less likely to
point of care tests to screen for prediabetes, dia- receive statins for documented dyslipidaemia,
betes and metabolic syndrome in workplaces and and less likely to be on antihypertensive medica-
public spaces, often in partnership with interested tions for documented hypertension. These pro-
employers and non-governmental organizations. spectively collected data, using the same protocol,
Using various community and family-based provide a reasonable estimate of real-world prac-
cohorts, a validated risk equation for developing tices. By identifying the variation in practice
diabetes has been developed [73] and made into a across settings, clinicians, policy makers, and
web-based engine and mobile phone application, public health experts are better equipped to
known as Risk Understanding By Yourself examine the system and population differences to
(RUBY). These technologies aim to engage web- account for these differing gaps. Adoption of
site visitors and mobile-users to self-assess their broad interventions or secular changes, such as
risk of diabetes with recommendations including the introduction of new practice guidelines, can
periodic monitoring and connection to health- be evaluated using more detailed clinical data, on
care providers if they are at risk (http://rubyapp. a much larger population level, and used to guide
adf.org.hk/#). future interventions, which can be continuously
tracked in a quality improvement cycle. The clin-
ical decision algorithms and reporting tools
Beyond Hong Kong: Experience embedded in the JADE programme provide a
to Date means to disseminate the latest knowledge and
provide a closed-loop approach to knowledge
Though started in Hong Kong, the vision of the translation from the findings directly obtained
JADE programme is to create a virtual environ- from the included population.
ment to enable sharing of best practices and
empower informed decisions and self-
management in people with diabetes across Asia. JADE: Future Considerations
This platform, with multiple languages, has now
been adopted by more than 300 sites in ten other As technology advances, the JADE programme
Asian countries/regions: China, India, Indonesia, will endeavour to keep up with the demands of
Malaysia, Philippines, South Korea, Singapore, improving user interface and user experience for
Taiwan, Thailand, and Vietnam. This has enabled diabetes care delivery. Increasingly, the private
regional collaborations with cross-national com- and public sectors are recognizing the potential of
parisons and gap identification. using technology to deliver self-reported and self-
As an example, the prevalence of diabetes is tracked data for chronic disease management, with
increasing in young adults across Asia, but little ever-expanding mobile applications and medical
is known about the metabolic control or burden devices. Our team hopes to build interoperability
of diabetes in this population. Data were extracted between web-based JADE programme and
from 245 sites for cross sectional evaluation, and mobile devices which measure objective physical
Fig. 5.5 The JADE

programme advocates to use
logistics, information
technology and team-based
care to change practice and
influence policy through
ongoing collaborative research
efforts (Reprinted with
permission from Chan et al.
[76])
data (e.g., glucometers, insulin pumps, blood pres- mote international collaborative research efforts,
sure) to enable even better risk assessment and focused on providing more holistic, personalized,
monitoring to improve clinical decision-making. evidence-based care.
A focus on longitudinally capturing self-reported
data such as quality of life and patient satisfaction
through convenient technological methods will References
also allow for better clinical care and more accu-
rate health economic evaluation. 1. Information Services Department, Hong Kong SAR
Government. Feb 2016. Available from: http://www.
gov.hk/en/about/abouthk/factsheets/docs/public_
health.pdf
JADE Summary 2. Chan JC, Cockram CS, Nicholls MG, Cheung CK,
Swaminathan R. Comparison of enalapril and nifedip-
ine in treating non-insulin dependent diabetes associ-
Since 1989, the CUHK-PWH diabetes team has
ated with hypertension: one year analysis. BMJ.
amassed considerable knowledge, from the 1992;305(6860):9815.
microscopic world of molecular biology to the 3. Chan JC, Ko GT, Leung DH, Cheung RC, Cheung
macroscopic realm of clinical epidemiology and MY, So WY, et al. Long-term effects of angiotensin-
converting enzyme inhibition and metabolic control
translational medicine. With a strong focus on
in hypertensive type 2 diabetic patients. Kidney Int.
local needs, the CUHK-PWH team has been able 2000;57(2):590600.
to identify and develop approaches to overcome 4. So WY, Tong PCY, Ko GTC, Leung WYS, Chow
many of the barriers of chronic disease manage- C-C, Yeung VTF, et al. Effects of protocol-driven care
versus usual outpatient clinic care on survival rates in
ment through the use of structured workflow,
patients with type 2 diabetes. Am J Manag Care.
information technology, team-based care, and 2003;9(9):60615.
rigorous evaluation (Fig. 5.5). The team recog- 5. Piwernetz K, Home PD, Snorgaard O, Antsiferov M,
nizes the importance of facilitating communica- Staehr-Johansen K, Krans M. Monitoring the targets
of the St Vincent Declaration and the implementation
tion of the evidence at all levels, from patients to
of quality management in diabetes care: the
care providers to policy makers. The JADE pro- DIABCARE initiative. The DIABCARE Monitoring
gramme serves as an important prototype to dem- Group of the St Vincent Declaration Steering
onstrate how collection of registry data can Committee. Diabet Med. 1993;10(4):3717.
6. Chan JCN, Malik V, Jia W, Kadowaki T, Yajnik CS,
enhance care, identify gaps, and improve out-
Yoon K-H, et al. Diabetes in Asia: epidemiology, risk
comes, while all in the spectre of usual clinical factors, and pathophysiology. JAMA.
practice. This serves as a hopeful example to pro- 2009;301(20):212940.
7. Yang X, So W, Kong A, Clarke P, Ho C, Lam C, et al. 20. Ting RZW, Nan H, Yu MWM, Kong APS, Ma RCW,
End-stage renal disease risk equations for Hong Kong Wong RYM, et al. Diabetes-related distress and phys-
Chinese patients with type 2 diabetes: Hong Kong ical and psychological health in Chinese type 2 dia-
Diabetes Registry. Diabetologia. betic patients. Diabetes Care. 2011;34(5):10946.
2006;49(10):2299308. 21. Zhang Y, Ting R, Lam M, Lam J, Nan H, Yeung R,
8. Yang X, So W-Y, Kong APS, Ho C-S, Lam CWK, et al. Measuring depressive symptoms using the
Stevens RJ, et al. Development and validation of Patient Health Questionnaire-9 in Hong Kong Chinese
stroke risk equation for Hong Kong Chinese patients subjects with type 2 diabetes. J Affect Disord.
with type 2 diabetes: the Hong Kong Diabetes 2013;151(2):6606. Elsevier.
Registry. Diabetes Care. 2007;30(1):6570. 22. Ting RZW, Lau ESH, Ozaki R, Lau WWY, Kong
9. Yang X, Ma RC, So W-Y, Ko GT, Kong AP, Lam CW, APS, Luk AOY, et al. High risk for cardiovascular dis-
et al. Impacts of chronic kidney disease and albumin- ease in Chinese type 2 diabetic patients with major
uria on associations between coronary heart disease depression a 7-year prospective analysis of the
and its traditional risk factors in type 2 diabetic Hong Kong Diabetes Registry. J Affect Disord.
patients the Hong Kong Diabetes Registry. 2013;149(13):12935.
Cardiovasc Diabetol. 2007;6:37. 23. Chan JC, Lau ES, Luk AO, Cheung KK, Kong AP, Yu
10. Yang X, Ma R, So W-Y, Kong A, Ko GT, Ho C-S, LW, et al. Premature mortality and co-morbidities in
et al. Development and validation of a risk score for young-onset diabetes a 7 year prospective analysis.
hospitalization for heart failure in patients with type 2 Am J Med. 2014;127(7):61624.
diabetes mellitus. Cardiovasc Diabetol. 2008;7(1):9. 24. Luk AOY, Lau ESH, So WY, Ma RCW, Kong APS,
11. Yang X, So WY, Tong PCY, Ma RCW, Kong APS, Ozaki R, et al. Prospective study on the incidences of
Lam CWK, et al. Development and validation of an cardiovascular-renal complications in Chinese
all-cause mortality risk score in type 2 diabetes. Arch patients with young-onset type 1 and type 2 diabetes.
Intern Med. 2008;168(5):4517. Diabetes Care. 2014;37(1):14957.
12. Jiao F, Fung CSC, Wan YF, McGhee SM, Wong 25. Ng MCY, So W-Y, Cox NJ, Lam VKL, Cockram CS,
CKH, Dai D, et al. Long-term effects of the multidis- Critchley JAJH, et al. Genome-wide scan for type 2
ciplinary risk assessment and management program diabetes loci in Hong Kong Chinese and confirmation
for patients with diabetes mellitus (RAMP-DM): a of a susceptibility locus on chromosome 1q21-q25.
population-based cohort study. Cardiovasc Diabetol. Diabetes. 2004;53(6):160913.
2015;14:105. 26. Ng MCY, Park KS, Oh B, Tam CHT, Cho YM, Shin
13. Wu JYF, Leung WYS, Chang S, Lee B, Zee B, Tong HD, et al. Implication of genetic variants near
PCY, et al. Effectiveness of telephone counselling by TCF7L2, SLC30A8, HHEX, CDKAL1, CDKN2A/B,
a pharmacist in reducing mortality in patients receiv- IGF2BP2, and FTO in type 2 diabetes and obesity in
ing polypharmacy: randomised controlled trial. BMJ. 6,719 Asians. Diabetes. 2008;57(8):222633.
2006;333(7567):522. 27. Ma RCW, So WY, Tam CHT, Luk AO, Ho JSK, Wang
14. Leung WYS, So W-Y, Tong PCY, Chan NN, Chan Y, et al. Genetic variants for type 2 diabetes and new-
JCN. Effects of structured care by a pharmacist- onset cancer in Chinese with type 2 diabetes. Diabetes
diabetes specialist team in patients with type 2 dia- Res Clin Pract. 2014;103(2):32837.
betic nephropathy. Am J Med. 2005;118(12):1414. 28. Ma RCW, Tam CHT, Wang Y, Luk AO, Hu C, Yang
15. Chan JC, So W-Y, Yeung C-Y, Ko GT, Lau I-T, Tsang X, et al. Genetic variants of the protein kinase C-beta
M-W, et al. Effects of structured versus usual care on 1 gene and development of end-stage renal disease in
renal endpoint in type 2 diabetes: the SURE study: a patients with type 2 diabetes. JAMA.
randomized multicenter translational study. Diabetes 2010;304(8):8819.
Care. 2009;32(6):97782. 29. Wang Y, Luk AOY, Ma RCW, So W-Y, Tam CHT, Ng
16. Das-Munshi J, Stewart R, Ismail K, Bebbington PE, MCY, et al. Predictive role of multilocus genetic poly-
Jenkins R, Prince MJ. Diabetes, common mental dis- morphisms in cardiovascular disease and
orders, and disability: findings from the UK National inflammation-related genes on chronic kidney disease
Psychiatric Morbidity Survey. Psychosom Med. in type 2 diabetes an 8-year prospective cohort anal-
Jan;69(6):54350. ysis of 1163 patients. Nephrol Dial Transplant.
17. Anderson RJ, Freedland KE, Clouse RE, Lustman 2012;27(1):1906.
PJ. The prevalence of comorbid depression in adults 30. Ho JSK, Germer S, Tam CHT, So W-Y, Martin M, Ma
with diabetes: a meta-analysis. Diabetes Care. RCW, et al. Association of the PPARG Pro12Ala
2001;24(6):106978. polymorphism with type 2 diabetes and incident coro-
18. Grigsby AB, Anderson RJ, Freedland KE, Clouse RE, nary heart disease in a Hong Kong Chinese popula-
Lustman PJ. Prevalence of anxiety in adults with dia- tion. Diabetes Res Clin Pract. 2012;97(3):48391.
betes: a systematic review. J Psychosom Res. 31. So W-Y, Wang Y, Ng MCY, Yang X, Ma RCW, Lam
2002;53(6):105360. V, et al. Aldose reductase genotypes and cardiorenal
19. Golden SH. A review of the evidence for a neuroen- complications: an 8-year prospective analysis of
docrine link between stress, depression and diabetes 1,074 type 2 diabetic patients. Diabetes Care.
mellitus. Curr Diabetes Rev. 2007;3(4):2529. 2008;31(11):214853.
32. Wang Y, Ng MCY, Lee S-C, So W-Y, Tong PCY, 44. So W, Yang X, Ma R, Kong A, Lam C, Ho C, et al.
Cockram CS, et al. Phenotypic heterogeneity and Risk factors in V-shaped risk associations with all-
associations of two aldose reductase gene polymor- cause mortality in type 2 diabetesThe Hong Kong
phisms with nephropathy and retinopathy in type 2 Diabetes Registry. Diabetes Metab Res Rev.
diabetes. Diabetes Care. 2003;26(8):24105. 2008;24(3):23846.
33. Baum L, Ng MCY, So W-Y, Lam VKL, Wang Y, Poon 45. Tong PCY, Kong APS, So W-Y, Ng MHL, Yang X,
E, et al. Effect of hepatic lipase -514C->T polymor- Ng MCY, et al. Hematocrit, independent of chronic
phism and its interactions with apolipoprotein kidney disease, predicts adverse cardiovascular out-
C3482C->T and apolipoprotein E exon 4 polymor- comes in Chinese patients with type 2 diabetes.
phisms on the risk of nephropathy in Chinese type 2 Diabetes Care. 2006;29(11):243944.
diabetic patients. Diabetes Care. 2005;28(7):17049. 46. Yang X, So WY, Ma RCW, Yu LWY, Ko GTC, Kong
34. Ma RCW, Lee HM, Lam VKL, Tam CHT, Ho JSK, APS, et al. Use of sulphonylurea and cancer in type 2
Zhao H-L, et al. Familial young-onset diabetes, prediabetes-the Hong Kong Diabetes Registry. Diabetes
diabetes and cardiovascular disease are associated Res Clin Pract. 2010;90(3):34351.
with genetic variants of DACH1 in Chinese. PLoS 47. Yang X, So W-Y, Ma RCW, Yu LWL, Kong APS, Lee
ONE. 2014;9(1):e84770. HM, et al. Use of thiazolidinedione and cancer risk in
35. Lam VKL, Ma RCW, Lee HM, Hu C, Park KS, Furuta type 2 diabetes: the Hong Kong Diabetes Registry.
H, et al. Genetic associations of type 2 diabetes with Diabetes Res Clin Pract. 2012;97(1):e137.
islet amyloid polypeptide processing and degrading 48. Yang X, Ko GTC, So WY, Ma RCW, Yu LWL, Kong
pathways in Asian populations. PLoS ONE. APS, et al. Associations of hyperglycemia and insulin
2013;8(6):e62378. usage with the risk of cancer in type 2 diabetes: the
36. Ma RCW, Hu C, Tam CH, Zhang R, Kwan P, Leung Hong Kong Diabetes Registry. Diabetes.
TF, et al. Genome-wide association study in a Chinese 2010;59(5):125460.
population identifies a susceptibility locus for type 2 49. Yang X, Ma RCW, So WY, Ko GTC, Kong APS, Zhao
diabetes at 7q32 near PAX4. Diabetologia. H, et al. White blood cell count and renin-angiotensin
2013;56(6):1291305. system inhibitors for the risk of cancer in type 2 diabe-
37. Ng MCY, Lam VKL, Tam CHT, Chan AWH, So W-Y, tes. Diabetes Res Clin Pract. 2010;87(1):11725.
Ma RCW, et al. Association of the POU class 2 50. Yang X, So W-Y, Ma RCW, Ko GTC, Kong APS,
homeobox 1 gene (POU2F1) with susceptibility to Wang Q, et al. Predicting values of lipids and white
type 2 diabetes in Chinese populations. Diabet Med. blood cell count for all-site cancer in type 2 diabetes.
2010;27(12):14439. Endocrinol Relat Cancer. 2008;15(2):597607.
38. Mooyaart AL, Valk EJJ, van Es LA, Bruijn JA, de 51. Kong APS, Yang X, So W-Y, Luk A, Ma RCW, Ozaki
Heer E, Freedman BI, et al. Genetic associations in R, et al. Additive effects of blood glucose lowering
diabetic nephropathy: a meta-analysis. Diabetologia. drugs, statins and renin-angiotensin system blockers
2011;54(3):54453. on all-site cancer risk in patients with type 2 diabetes.
39. So WY, Kong APS, Ma RCW, Ozaki R, Szeto CC, BMC Med. 2014;12:76.
Chan NN, et al. Glomerular filtration rate, cardiorenal 52. Yang X, So W-Y, Ma RCW, Kong APS, Xu G, Chan
end points, and all-cause mortality in type 2 diabetic JCN. Diabetes and cancer: the mechanistic implica-
patients. Diabetes Care. 2006;29(9):204652. tions of epidemiological analyses from the Hong
40. Luk AOY, So W-Y, Ma RCW, Kong APS, Ozaki R, Kong Diabetes Registry. Diabetes Metab Res Rev.
Ng VSW, et al. Metabolic syndrome predicts new 2012;28(5):37987.
onset of chronic kidney disease in 5,829 patients with 53. Cheng AYS, Kong APS, Wong VWS, So WY, Chan
type 2 diabetes: a 5-year prospective analysis of the HLY, Ho CS, et al. Chronic hepatitis B viral infection
Hong Kong Diabetes Registry. Diabetes Care. independently predicts renal outcome in type 2 dia-
2008;31(12):235761. betic patients. Diabetologia. 2006;49(8):177784.
41. Kong APS, Yang X, Luk A, Ma RCW, So WY, Ozaki 54. Lo MKW, Lee K-F, Chan NN, Leung WYS, Ko GTC,
R, et al. Severe hypoglycemia identifies vulnerable Chan W-B, et al. Effects of gender, Helicobacter
patients with type 2 diabetes at risk for premature pylori and hepatitis B virus serology status on cardio-
death and all-site cancer: the Hong Kong Diabetes vascular and renal complications in Chinese type 2
Registry. Diabetes Care. 2014;37(4):102431. diabetic patients with overt nephropathy. Diabetes
42. Yang X, So WY, Ma RCW, Kong APS, Lee HM, Xu Obes Metab. 2004;6(3):22330.
G, et al. Synergistic effects of low LDL cholesterol 55. Yang X, Wang Y, Luk AOY, So WY, Ma RCW, Kong
with other factors for the risk of cancer in type 2 dia- APS, et al. Enhancers and attenuators of risk associa-
betes: the Hong Kong Diabetes Registry. Acta tions of chronic hepatitis B virus infection with hepa-
Diabetol. 2012;49 Suppl 1:S18593. tocellular carcinoma in type 2 diabetes. Endocrinol
43. Ko GT, So WY, Tong PC, Chan WB, Yang X, Ma RC, Relat Cancer. 2013;20(2):16171.
et al. Effect of interactions between C peptide levels 56. Gaede P, Vedel P, Larsen N, Jensen GV, Parving HH,
and insulin treatment on clinical outcomes among Pedersen O. Multifactorial intervention and cardio-
patients with type 2 diabetes mellitus. CMAJ. vascular disease in patients with type 2 diabetes. N
2009;180(9):91926. Engl J Med. 2003;30;348(5):38393.
57. Kong APS, Yang X, Ko GTC, So W-Y, Chan W-B, 67. McDonald AM, Treweek S, Shakur H, Free C, Knight
Ma RCW, et al. Effects of treatment targets on subse- R, Speed C, et al. Using a business model approach
quent cardiovascular events in Chinese patients with and marketing techniques for recruitment to clinical
type 2 diabetes. Diabetes Care. 2007;30(4):9539. trials. Trials. 2011;12(1):74.
58. International Diabetes Federation. Global guideline 68. Chan JCN, Sui Y, Oldenburg B, Zhang Y, Chung
for type 2 diabetes: recommendations for standard, HHY, Goggins W, et al. Effects of telephone-based
comprehensive, and minimal care. Diabet Med. peer support in patients with type 2 diabetes mellitus
2006;23(6):57993. receiving integrated care. JAMA Intern Med.
59. Ahola AJ, Groop P-H. Barriers to self-management of 2014;174(6):97281.
diabetes. Diabet Med. 2013;30(4):41320. 69. Yin JM, Wong R, Au S, Chung HY, Lau M, Lin L,
60. Loveman E, Cave C, Green C, Royle P, Dunn N, Tsang CC, Lau KP, Ozaki R, So WY, Ko G, Luk A,
Waugh N. The clinical and cost-effectiveness of Yeung R, Chan JC, Yin JM, Wong R, Au S, Chung HY,
patient education models for diabetes: a systematic Lau M, Lin L, Tsang CC, Lau KP, Ozaki R, So WY,
review and economic evaluation. Health Technol Ko G, Luk A, Yeung RCJ. Effects of being a peer sup-
Assess. 2003;7(22):iii. 1190. porter on diabetes management in people with type 2
61. Loveman E, Frampton GK, Clegg AJ. The clinical diabetes. Ann Fam Med. 2015;13 Suppl 1:S429.
effectiveness of diabetes education models for type 2 70. Yin J. A quality improvement project evaluating the
diabetes: a systematic review. Health Technol Assess. effect of personalized feedback report and Peer sup-
2008;12(9):1116. iii. port in patients with diabetes in Hong Kong. PhD dis-
62. Jarvis J, Skinner TC, Carey ME, Davies MJ. How can sertation. The Chinese University of Hong Kong.
structured self-management patient education 2015. ProQuest LLC.
improve outcomes in people with type 2 diabetes? 71. Marshall M, Pronovost P, Dixon-Woods M. Promotion
Diabetes Obes Metab. 2010;12(1):129. of improvement as a science. Lancet.
63. Tricco AC, Ivers NM, Grimshaw JM, Moher D, 2013;381(9864):41921.
Turner L, Galipeau J, et al. Effectiveness of quality 72. Cho YS, Chen C-H, Hu C, Long J, Ong RTH, Sim X,
improvement strategies on the management of diabe- et al. Meta-analysis of genome-wide association stud-
tes: a systematic review and meta-analysis. Lancet. ies identifies eight new loci for type 2 diabetes in east
2012;379(9833):225261. Elsevier Ltd. Asians. Nat Genet. 2012;44(1):6772.
64. U.S. Department of Health and Human Services 73. Ko G, So W, Tong P, Ma R, Kong A, Ozaki R, et al. A
Food and Drug Administration Center for Drug simple risk score to identify Southern Chinese at high
Evaluation and Research (CDER). Guidance for risk for diabetes. Diabet Med. 2010;27(6):6449.
Industry Diabetes Mellitusevaluating cardiovascu- 74. Alberti KG, Zimmet P, Shaw J. International Diabetes
lar risk in new antidiabetic therapies to treat type 2 Federation: a consensus on type 2 diabetes preven-
diabetes Internet; 2008 cited 2015 May 17. Available tion. Diabet Med. 2007;24:45163.
from: http://www.fda.gov/downloads/drugs/guid- 75. Chan J, So W, Ko G, Tong P, Yang X, Ma R, et al. The
ancecomplianceregulatoryinformation/guidances/ Joint Asia Diabetes Evaluation (JADE) Program: a
ucm071627.pdf web-based program to translate evidence to clinical
65. Avorn J. The $2.6 billion pill methodologic and practice in Type 2 diabetes. Diabet Med.
policy considerations. N Engl J Med. 2009;26(7):6939.
2015;372(20):18779. 76. Chan JCN, Ozaki R, Luk A, Kong APS, Ma RCW,
66. Hay J, Jackson J. Panel 2: methodological issues in Chow FCC, et al. Delivery of integrated diabetes care
conducting pharmacoeconomic evaluations model- using logistics and information technology the Joint
ing studies. Value Health. 1999;2(2):7881. Asia Diabetes Evaluation (JADE) program. Diabetes
Res Clin Pract. 2014;106(Suppl):S295304.
Diabetes Care: A South African 6
Approach
Larry A. Distiller and Michael A.J. Brown
Part I: Background to the South supplies excellent care, but faces constant media
African Health Care System and opportunistic political accusations of pro-
teering off the health burden of South Africa.
South Africa has an estimated population of 54 Healthcare disparities are worsened by the
million people [1]. The Department of Health fact that around 70 % of all doctors and most spe-
holds overall responsibility for healthcare, with a cialists only work in the private sector; the
specic responsibility for the public sector. remaining 30 % serve the public sector [2, 3].
Because of high levels of poverty and unemploy- A health intelligence report [4] on the future
ment, the bulk of the burden of healthcare is of healthcare in Africa [4], considers South Africa
borne by the state, with 84 % of the population by many health measures, as the most
receiving some portion of their healthcare from advanced of the Sub-Saharan nations, with the
the public (Government) sector. Sixty-eight per- biggest and most well developed, high-quality,
cent of the population does not use any private private health insurance sector, and the largest
care at all, and a further 16 % of the population and best-trained health workforce in Africa. It is
rely on the public sector for hospital care, but use also formulating a universal national health
the private sector for primary care, paying out of insurance (NHI) system, one of the rst and most
their own pockets. Despite this burden, ambitious on the continent, in attempt to bring
Government spending on healthcare comprises healthcare equality to all. However, the same
less than half of total health expenditure. In 2013, report indicates that South Africa also experi-
the remaining 16 % of the population (8.64 mil- ences many healthcare problems facing other
lion people) paid for private health insurance African countries, including high rates of mater-
cover (often with a monthly contribution from nal, infant and child mortality, chronic conditions
their employers), from 87 Registered Medical including diabetes, hypertension and obesity,
Insurance companies or Medical Schemes (down injuries and violence, and communicable dis-
from 93 schemes in 2012, as schemes are battling eases like HIV and tuberculosis. Additionally,
to maintain the legislated monetary reserves and many health services underperform on service
amalgamate or fold). The private sector generally delivery, with a background of poor manage-
ment, deteriorating infrastructure, and under-
funding. This has increased healthcare inequality.
L.A. Distiller (*) M.A.J. Brown The private-sector health insurance system is
Centre for Diabetes and Endocrinology (Pty) Ltd, seen as both an asset and a potential obstacle to
Johannesburg, South Africa implementing an NHI system [4]. Based on many
e-mail: [email protected];
[email protected] patient reports, treatment of patients with diabetes

88 L.A. Distiller and M.A.J. Brown
in the public health sector is also under-resourced Health Insurance companies, are desperately
and underfunded. Some Academic Hospital dia- looking for a way to reduce costs without reduc-
betes clinics endeavour to provide full care, but ing quality of care. A number of commonly used
overall, these clinics are understaffed and overex- interventions have proved to be cost-effective [7].
tended by large patient numbers. While they do These include:
offer comprehensive diabetes education, and eye
and foot screening, circumstances limit their the use of angiotensin-converting-enzyme
reach and fragment their care. For example, inhibitors (ACE inhibitors (ACE-I)) for inten-
patients verbalize that they often choose not to sive blood pressure control
lose their place in 68 h pharmacy queues, ACE-I or angiotensin receptor blockers (ARB)
thereby missing potentially useful consultations for renal disease
with dieticians and other diabetes team comprehensive foot care
members. intensive risk-factor control
intensive insulin therapy for type 1 diabetes
life-style modication and
Part II: The Growing Burden screening for and early treatment of
of Diabetes retinopathy.
Diabetes imposes a massive economic burden on While these are all important components of
all healthcare systems, accounting for 11 % of long-term care, the economic consequences of
total global healthcare expenditure on adults in such treatment protocols can be overwhelming
2013. In the next 20 years, the developing for any healthcare funder. Consequently, many
world is expected to be affected most by the dia- funders, be they Private or Government, resort to
betes pandemic, with 77 % of people with diabe- developing Managed Care Programmes to con-
tes living in low- and middle-income countries. tain costs while trying to provide quality, afford-
Africa is, and will be, particularly hard-hit with able care to their patients.
76 % of deaths due to diabetes occurring in
people under 60 years of age, the highest conti-
nental proportion of people with diabetes being Part III: Managed Care Programmes
undiagnosed (62.5 %) and the largest predicted in South Africa
continental increase in prevalence (109.1 % by
2035) [5]. Regulation 8 of the Medical Schemes Act of
With the spreading diabetes pandemic and the 1998 [8] requires South African Medical Schemes
advent of newer, more expensive drugs to treat to pay in full for the costs of diagnosis, treatment
the condition, it can be anticipated that costs of and care of 270 medical conditions and 25 com-
diabetes care over the coming decades will mon chronic conditions, including diabetes mel-
increase incrementally. While the exact preva- litus. These Prescribed Minimum Benets
lence of diabetes in South Africa is unknown, the (PMB) are a set of dened benets to ensure that
2014 International Diabetes Federation (IDF) all medical scheme members have access to cer-
estimated prevalence of diabetes for South Africa tain minimum health services, regardless of the
was 8.39 % [6], which suggests a possible total option they have chosen. Treatment protocols
of up to 4,530,000 patients with diabetes, which provide guidelines for appropriate treat-
737,000 in the private healthcare sector, and a ment for each of the chronic PMB conditions
further 3,800,000 in the public sector. have been published in the Government Gazette.
Worldwide funders, whether they are govern- To contain the costs of providing such cover,
ments, National Health Services, or private while ensuring good quality treatment, certain
6 Approaches to Integrated Diabetes Care: A South African Approach 89
measures have been taken to ensure that schemes and Council for Medical Scheme PMB guid-
can cover those members who need it, without ance [12] only species:
putting the scheme at nancial risk. Accordingly Consultations with your treating provider
schemes are entitled to contract with designated (GP or specialist if authorized by your
service providers (specied groups of hospitals, scheme)
clinics, doctors, retail pharmacies, etc.) to pro- Lifestyle modication interventions such
vide treatment for PMB conditions. However, as dietary and disease education.
this fact must be stated in the scheme rules and The type of provider is unspecied and thus
patients must be informed about where and how the essential role of the Diabetes Nurse
they can get medication and treatment from that Educator in diabetes care [13] is ignored.
provider. Patients who do not abide by the rules Programmes that restrict the number of labo-
about which providers to use, may face having to ratory investigations that can be performed
pay all or part of the costs of their treatment annually. A typical limit of services in seen in
themselves. Fig. 6.1, as per a form sent to patients who are
Often, designated service providers institute on this particular funders Management
managed care programmes to standardize and Programme.
control care rendered in a safe and cost-effective Programmes that make use of telephonic case
manner. However, many of these programmes manager contact from time to time to check
concentrate on cost-savings rather than patient how patients are doing.
service utilization and improved clinical out-
comes. These programmes fall into several While these approaches may save some costs in
categories: the short-term, Managed Care Programmes which
do not address patient outcomes nor reduce long
Programmes driven by drug formularies, term complications, ignore the fact that that the
which may exclude or restrict some more majority of the costs for treating diabetes, even in
expensive and newer pharmaceutical agents. the medium term, are due to the treatment of acute
In many instances, this also extends to restrict- and chronic complications and for inpatient hospi-
ing (in patients with type 1 diabetes) or disal- tal care [14]. Additionally, it is well established
lowing entirely (in type 2 patients) testing that poor long-term clinical outcomes increase the
strips for home glucose monitoring. However, cost burden of managing the patient with diabetes
overall, the costs of medication, including by up to 250 %. Outpatient care provided in clinics
insulin, accounts for just 7 % of all healthcare or doctors ofces, accounts for less than one-
costs related to diabetes [9]. quarter of the costs of accruing to a cohort of
Programmes that restrict the frequency of patients with diabetes [15]. Despite this, a recent
patient visits to doctors and that restrict access study of 11 different funders in South Africa [16]
to specialist care. The number of visits to showed that utilization of necessary outpatient ser-
other healthcare providers such as dietitians, vices to monitor diabetes control and screen for
podiatrists and ophthalmologists are also lim- potential complications is grossly inadequate. On
ited as a cost-cutting exercise. Not surpris- average, only 48.37 % of patients have an HbA1c
ingly, these funders do not even acknowledge measured annually, 35.08 % have their lipids mon-
the need for, or the role of, the Diabetes Nurse itored, 31.55 % are tested for microalbuminuria,
Educator and do not fund education sessions. 20.79 % see an ophthalmologist annually and
This passes PMB muster because the 2.39 % see a podiatrist (Fig. 6.2).
Treatment Protocols [10, 11] focus on Clearly, if the economic costs of diabetes are
attainment of glycaemic targets, using mainly to be contained, any approach needs to incorpo-
an algorithmic pharmacological approach, rate a managed care initiative that will promote
Tariff Code Description Number of Services

000192 GP - Consultation/Visit: long duration 2
000192 Specialist Consultation/Visit: long duration 1
(Physician)
000192 Specialist Consultation/Visit: long duration 1
(Ophthalmologist)
001232 ECG Without Effort 1
003003 Fundus contact lens or 90D lens examination 1
003009 Basic capital equipment used in Specialist rooms 1
003014 Test:Tonometry 1
004025 Blood Test: Cholesterol HDL/LDL/Trig 1
004032 Blood Test: Creatinine level 1
004050 Blood Test: Glocose Strip Test 2
004064 Blood Test: Haemoglobin A1C measurement 2
004113 Blood Test: Potassium level 1
004114 Blood Test: Sodium level 1
004151 Blood Test: Urea level 1
004188 Urine Test: Dipstick 2
068302 Podiatrist Consultation 11 - 20 minutes 1
084205 Dietician Consultation 1
Fig. 6.1 A typical diabetes managed care programme as promoted by a medical funder
Fig. 6.2 Patient utilization of services across 11 medical aid schemes (funders) in South Africa (Used with permission
from HQA [16])
better care and control of diabetes and other importantly, none of this will be implementable
co-morbidities, reduce both acute and all-cause unless the system of care is designed around the
hospitalization rates, and be proactive in promot- patients perspective of their diabetes and the
ing patient health rather than reactively treating needed care, the so-called integrated care of
complications and problems as they arise. Most diabetes [17].
Part IV: The CDE Diabetes hospitalization (with the Centre being responsi-
Management Programme (DMP): ble for the costs thereof if we failed to prevent
Past to Present this), our diabetes care team would receive fair
professional remuneration for proactive diabetes
With the intensive care results and the other care management. The medical aid scheme would
insights provided by the Diabetes Control and receive state-of-the art care (and improved out-
Complications Trial (DCCT) [18], fresh in our comes) for their members with diabetes and the
minds, the Centre for Diabetes and ability to budget for their reduced diabetes risk.
Endocrinology was initially established as a This was groundbreaking thinking at that time;
single Centre of Excellence in 1994. It was the concept of managed healthcare and the idea
staffed by two endocrinologists, two nurse edu- of ring-fencing, capitating and managing a condi-
cators, a registered dietician, a podiatrist, a clini- tion like diabetes was alien in South Africa. In
cal psychologist, a pharmacist and a biokineticist 1995, in a great leap of faith, and possibly with a
and effectively provided a one-stop shop for glimpse into the future, this medical aid scheme
our patients. With all services in one place, and a contracted the services of the CDE. In the rst
well-managed appointment system, patients month of operations, the CDE had 13 patients
experienced minimal waiting and optimal con- under management.
sultation times. They could continue with their We recognized that to provide good diabetes
lives with minimal disruption. This was in stark care across South Africa, more than one Centre
contrast to the prevailing situation diabetes care of Excellence would be required. This was also
resources available to South Africans were gener- needed to meet the expectations of our rst
ally grossly inadequate. Additionally, medical aid funder, which had members across South Africa.
schemes did not appropriately fund private sector As a result, we established a founding preferred
diabetes care and the resultant outcomes were provider network of 14 CDE Centres within
generally suboptimal. We had a vision to create months. At the helm at each of these initial
an all-encompassing and comprehensive diabetes Centres of Excellence, was either an
treatment and management Centre, which Endocrinologist or a specialist Diabetologist.
allowed us to implement correct and appropriate Over the past 20 years, the Centre for Diabetes
diabetes care principles. However, within a very and Endocrinology has expanded from these 14
short time after opening our Centre, we faced Centres, to a national network offering the ser-
bankruptcy as the salaries of the allied health pro- vices of 31 Endocrinologists/Diabetologists, 48
fessionals and the costs our ancillary services Specialist Physicians (Internists), 165 Centres
could only be funded from the consultation fees of Excellence run by trained and dedicated
of the two founding medical practitioners. General Practitioners and Family Physicians, and
We had to make a plan to survive. With our 610 contracted primary care doctors (Fig. 6.3).
current crisis being the muse of innovation, we This network effectively offers primary, second-
approached a medical aid scheme with our care ary and tertiary levels of expertise and care
offering and a simple but compelling nancial nationally in the private sector.
equation. We knew that we could manage the In tandem with the growth of the CDE
monthly treatment costs of a person with diabetes Provider Network, the number of patients under
for X. We also knew that with the current our management has risen steadily At the end
hospital-centric diabetes management approaches of May 2015, our national network of 220 Centres
of the time, the medical funders were paying a (some Centres have more than one CDE-
higher gure Z. Could we not agree to meet accredited doctor) were responsible for the care
somewhere in-between at a mutually agreed of 20,569 patients. Two thousand eight hundred
monthly, per patient capitation fee, Y? With a ninety one (14.1 %) had type 1 diabetes and
contractually bound promise of community- 17,678 (85.9 %), type 2 diabetes. Of the people
based, holistic care that prevented unnecessary with type 2 diabetes, 50.4 % (8903) were on oral
Fig. 6.3 The CDE network pyramid
glucose-lowering agents alone and 49.6 % (8775) this approach have yet to be ascertained. In this
required insulin therapy, with or without the chapter, we discuss the CDE Comprehensive
addition of oral agents. Care Plan.
Persons with diabetes covered by the medical
schemes and contracted to receive care via the
CDE DMP encompass all of the multiple ethnic Part V: Structure and Principles
groups found in South Africa and much of the of the CDE DMP and Diabetes Care
socioeconomic spectrum from blue-collar work- Network
ers to company directors. We are keen to expand
our best practice care to all South Africans with The contracted servicing doctor (Centre) receives
diabetes. a set monthly capitation fee, in advance, for car-
In response to the needs of funders who can- ing for patients contracted to receive their diabe-
not afford the CDE traditional Comprehensive tes care from that Centre. A two-tier fee structure
Care Plan, from 2015 the CDE has introduced a exists, with a lower fee being paid for patients on
Standard Care Plan, which has a capitation fee oral glucose-lowering agents alone and a higher
more palatable to funders with tight nancial fee for those requiring insulin, irrespective of
margins, albeit with reduced benets. Within whether the insulin-requiring patient has type 1
this new model, payment responsibility for medi- or type 2 diabetes. The fee is negotiated annually
cines and some services falls away from the based on the anticipated costs of providing all
ambit of the capitation fee. Our Centres are, how- guaranteed services, plus a fair margin for the
ever, expected to refer to the core members of the Centre taking the risk for those patients who
diabetes team as usual, and maintain the highest require more intensive management, the cost of
possible standard of care, even though they will acute diabetes related hospital admissions, ina-
not be paying for all aspects directly. With this tion and the choice of more expensive treatment
approach, we hope to enable access to improved modalities for selected patients. Centres are paid
diabetes care to many people for whom this was according to the number of patients they have
previously not possible. The clinical outcomes of contracted to service. Funds must be utilized to
provide all services, including clinical care, sup- Table 6.1 Minimum Care Guidelines as contractually
agreed to between CDE diabetes centres and contracted
ply of all diabetes medication and accessories
funders
(including meters and testing materials for self-
Service Minimum frequency
monitoring of blood glucose), and specied labo-
Consultations:
ratory investigations (The annual measurement
Doctor 2 annually
of lipids, renal function, and microalbuminuria is
Nurse educator 2 annually
part of the guaranteed services, and careful moni-
Dietitian (nutritional 1 annually
toring of blood pressure is expected at every guidance)
visit). Should it be found necessary, the Medical Podiatrist (foot care) 1 annually (screening)
Scheme funds treatment for co-morbid condi- Ophthalmologist (eye 1 annually (screening)
tions outside of the monthly capitation fee. We care)
guarantee all medical schemes that contract the Exercise physiologist If required
CDE Network an absolute minimum level of ser- (Biokineticist)
vicing for their patients as stated in the CDE Clinical psychologist If required
Minimum Care Guidelines (Table 6.1). Laboratory tests to monitor diabetes:
However, subjects often receive additional ser- HbA1c 6-monthly
vices in excess of these Guidelines, depending on Lipogram 1 annually
their individual clinical circumstances. Renal function, 1 annually
microalbuminuria
Each Centre of Excellence is required to have
24/7/365 Hotline for emergencies and advice
on staff, in addition to the responsible doctor, a
All diabetes medications and monitoring equipment
trained Diabetes Nurse Educator (DNE) to act as
Insulins and tablets As prescribed
the primary team contact for their DMP members
Blood glucose metres As prescribed
and to facilitate the process of patient empower- and test strips
ment and self-management. Each Centre must Glucagon hypo kit All patients on insulin
also contract with or employ both a dietitian and therapy replaced on
a podiatrist to provide the guaranteed services to useexpiry
the patients. For diabetes-related emergencies, Ketone test strips All patients with type 1
diabetes replaced on
every Centre must provide a direct 24-h emer- useexpiry
gency telephone number (Hotline). Insulin pens/syringes, As prescribed
Via this novel diabetes-care model, CDE needles, lancets
Centres are empowered to decide on medication Risk assumption for hospitalization costs for acute
for the optimal treatment of their patients with diabetes emergencies
diabetes no formulary restrictions are imposed.
Since each Centre must pay for the medication
and insulin from the set capitation fee, the treat- Furthermore, each contracted Centre is
ing doctor is responsible for deciding which directly responsible for all additional costs
treatment regimen is the most cost- and quality- incurred should a person with diabetes under that
effective for each patient. Thus, the onus falls on Centres care be admitted to hospital for a diabe-
the individual Centre to absorb the costs of pre- tes emergency (dened as a primary admission
scribing more expensive treatment modalities diagnosis of hypoglycaemia, hyperglycaemia,
should they be deemed clinically advantageous. hyperosmolar non-ketotic coma (HONK) or dia-
Our Centres are generally happy to do this, betic ketoacidosis (DKA)). This is based on the
because they are clearly taught that the CDE experience that admissions for such acute meta-
DMP has an in-built level of cross subsidization bolic events are largely avoidable in patients who
(like medical aid schemes) where any available have received adequate diabetes education and
funds remaining from patients on less-expensive who are sufciently self-empowered and have
regimens cover the decit accrued by the opportunity to call their Centre via the
patients on more expensive regimens. contractually mandated 24-h emergency Hotline
should they develop any acute problems. Again, of 7.7 % (1.2 % SD, Median 7.6 %) after 5 years
our Centres are happy to assume this risk (once is not at the recommended target of 7 %, it
they have a nancially viable minimum of 20 approaches the 7.6 % suggested by the VISS
capitated patients under their care), because the (Vascular diabetic complications In Southeast
contracted doctors soon learn that with good Sweden) study to be the cut-off for the prevention
care, they have the power to avoid these events. of proliferative retinopathy and macroalbumin-
This voluntary risk assumption for the costs of uria [20]. It is signicantly better than the mean
hospitalization is a powerful guarantee to our HbA1c seen at many diabetes clinics, which may
funding partners that the CDE and its Provider be in excess of 8 % [21, 22]. The mean HbA1c of
Network will do the job they have contracted to 7.4 % (1.36 % SD, median 7.1 %) achieved in
do. The DMP, however, does not cover the treat- our type 2 patients, is considerably better than
ment cost of any chronic complication or the that seen in many surveys in people with type 2
treatment of other co-morbidities or risk factors, diabetes including the United Kingdom
and the hospital admission costs for diabetes Prospective Diabetes Study (UKPDS) 10-year
complications or non-diabetes-related illness. follow-up data (mean HbA1c of both the inten-
Our contracts with Funders clearly state these sive and the conventional therapy groups evened
exclusions, as the funding risk for these events out at about 8 %) [23].
would exceed the capacity of the DMP budget,
which is clearly based on the daily management
costs of diabetes. Hospital Admissions
Hospital admission remains the top healthcare

Part VI: The CDE DMP: Outcomes cost in the private healthcare sector in South
Africa, accounting for 39.1 % of the total paid out
Note: Since people tend to change medical schemes by Medical Schemes in the 2013 nancial year
from time to time, with membership of the DMP [3]. Specically for diabetes, the all cause hospi-
depending on whether their current medical scheme tal admission rate for patients with diabetes is
was contracted to the DMP or not, long term fol- threefold higher than that seen in the nondiabetic
low-up for all patients is not always possible. population [24]. Additionally, a study of the
Economic Costs of Diabetes in the U.S. in 2012
[25] showed that for people with diabetes, hospi-
Glycaemic Control tal inpatient care accounted for 43 % of the total
medical cost of diabetes. Therefore, any pro-
A sustained reduction in HbA1c of approxi- gramme that can result in even a slight reduction
mately 1.5 % in subjects with both type 1 and in the number of acute diabetes-related admis-
type 2 diabetes has been achieved (Fig. 6.4) [19]. sions has the potential of substantial cost savings
Although the HbA1c assays were not stan- in any healthcare system.
dardized across the country, each patient had his In this context, we have seen a signicant
or her HbA1c performed at the same laboratory overall reduction in all acute diabetes-related
longitudinally, so that the starting HbA1c for hospital admissions for patients on the CDE
each patient effectively acted as its own control. DMP. Our previously reported hospital admis-
Reasons for the sustained improvement might sion rates [19] for patients requiring admission
relate to the high compliance rate ensured by the for acute metabolic decompensation (where the
CDE Programme as well as responsiveness to entire hospital bill was paid by their treating CDE
person-centred advice on lifestyle modication doctor) were 6 admissions per 1000 patient-years
needed and an aggressive treat-to-target for type 1 and 1 admission per 1000 patient-years
approach taught to and adopted by the CDE for type 2 subjects. This was achieved with no
Centres. While a mean HbA1c for type 1 patients patient mortality resulting from acute metabolic
Fig. 6.4 Five-year follow-up of 2726 type 1 and 14,317 type 2 patients with diabetes who were part of the CDE
Diabetes Management Programme for more than 5 years (Used with permission from Distiller et al. [19])
causes (hypoglycaemia, hyperglycaemia or Since mortality rates were not taken into
DKA). We have had no admissions for the crisis account, it could be argued that those patients with
of HONK in over 10 years, a condition that is still major illness or extensive arterial disease when
prevalent in many academic hospital settings joining the DMP could have died subsequently,
[26]. Hospital admission rates for these condi- leaving those in better health and with less reason
tions in this group of patients prior to joining the for hospital admission on the Programme over
DMP were not obtainable, nor are there any pub- subsequent years. However, the phenomenon of
lished gures for hospital admissions for acute managed, better-controlled persons with diabetes
diabetes-related causes available in South Africa. requiring hospital admission less often than
The CDE DMP cohort also showed a 40 % over- unmanaged and uncontrolled subjects is well-
all reduction in hospital admission rates and a described. One managed care approach in
20 % reduction in length of hospital stay for hos- Pennsylvania (USA) [27] was associated with a
pital admission diagnoses related directly or major reduction in the total number of admissions
indirectly to the diabetes (acute or chronic com- per patient per year, down from 0.16 to 0.12 over a
plications). All-cause hospital admission rates 2-year period. They also documented less inpa-
were reduced from 210 admissions per 1000 tient days and fewer emergency room visits.
patient-years for the rst year on the DMP to Another integrated diabetes disease management
<100 admissions per 1000 patient years in subse- programme across ve States in the USA also
quent years. This reduction was sustained for the reported a 22 % reduction in hospital admission
full 5 years of the study [19]. rates [28] and several other studies have conrmed
this [29, 30]. Attempts at an Integrated Care patient care and attention to and aggressive treat-
Initiative in the UK have been less successful [31], ment of other risk factors such as hypertension
with an increase in hospital admission rate in the and dyslipidaemia. Data on macrovascular out-
rst year of the Programme, although thereafter, comes would have been of interest, but were not
costs appear to have started reducing. The authors sufciently robust.
speculate that this may have been due to initial
difculties in implementing the initiative with
difculty in assuring participation of all local staff, Part VII: The CDE DMP: A South
amongst other problems. African Example of Integrated
Diabetes Care
Microvascular Disease Outcomes The DMP has provided the CDE with over two
decades of experience in many aspects of the
Prevalence data for diabetic microvascular dis- managed care of diabetes. Our clinical outcomes
ease are not available for South Africa and are include long-term improvements in glycaemic
difcult to source internationally. The National control, delay in the progression of microvascu-
Health and Nutrition Examination Survey lar complications and reductions in hospital
(NHANES) 19992004 survey reported chronic admissions for both acute metabolic emergencies
kidney disease to be present in 27.8 % and eye and all other causes.
disease in 18.9 % of people with diabetes in the The cost savings and resulting improvements
USA [32]. Microalbuminuria was present in in quality of life for the patients served are self-
2040 % of patients with diabetes [33]. The evident. Although no formal quality of life
Wisconsin Epidemiologic Study of Diabetic assessments have been performed, we have, how-
Retinopathy (WESDR) reported some retinopa- ever, been recognized for excellence in managed
thy in nearly all persons who had had type 1 dia- healthcare, by being awarded eight PMR.africa
betes for 20 years [34] and in nearly 80 % of Managed Healthcare Awards since 2002 (the
those who had had type 2 diabetes for the same majority being Diamond Arrow Awards
duration [35]. Up to 21 % of newly diagnosed (ranked 1st overall and rated at least 4, 10 out of
type 2 patients have some degree of retinopathy 5, 00 equivalent to outstanding). The PMR.
at time of diagnosis [36]. A series of patients who africa Awards are designed to recognize and
were assessed for retinopathy at the time of join- enhance excellence in a range of industries and to
ing the CDE DMP, showed a prevalence of 35.2 % set a benchmark in each sector. These externally
for the type 1 patients (background retinopathy and independently adjudicated awards are the
26 % and referable retinopathy 9.2 %) and 20.5 % culmination of a research process by PMR.africa,
in the type 2 diabetes (14.1 % background reti- whereby companies and institutions are rated
nopathy and 6.4 % referable retinopathy) [37]. based on respondents perceptions with a strong
This was in line with the internationally reported focus on evaluating and measuring customer ser-
gures. In individuals who were on the CDE vice and customer satisfaction. Importantly, a
DMP for over 5 years, the prevalence of retinopa- company, department, institution and individual
thy was 28 % for the type 1 patients and 26.6 % cannot enter the research process, but must
for the type 2 patients [19]. The incidence of always be nominated and rated by the respon-
nephropathy (15.8 % in patients with type 1 and dents. In the case of managed care companies,
22.6 % in patients with type 2 diabetes) was excellence is rated by input from a random,
clearly lower than might have been expected for national sample of 100 respondents (Chairmen
a mean duration of diabetes of 15.2 years in the and Principal Ofcers of listed/large companies
type 1 group and 9.3 years for the type 2 subjects. as well as Fund Managers, Trustees, Medical
This is probably attributable to the improved lev- Advisors/Directors and Assessors representing
els of glycaemic control, but also to better overall Medical Aid Schemes and Administrators).
Table 6.2 Differences between the CDE diabetes man- perspective of care and the needs that accrue as a
agement programme and usual disease management
result. Diabetes mellitus is a complex, chronic,
programmes for diabetes
physical/psychosocial/spiritual condition that
CDE Diabetes Management
affects every part of the human experience. This
Usual programmes Programme
gives rise to a number of challenges that we have
Maximum visits per Minimum visits guaranteed.
year laid down Maximum unlimited experienced over the years, which make the
Drug formularies Drugs used depends only on understanding of the patient perspective even
doctors judgement more important [38, 39] Healthcare professionals
Success measured Success measured by outcomes, (and people with diabetes) often need experience
by cost savings not cost-savings and specialized training and knowledge to
Clinical outcomes Clinical outcomes are key achieve this mutual understanding. The resultant
largely ignored performance indicators to
justify DMP existence
care principles from these insights are role-
No transfer of risk Risk of acute hospital modelled in all that we do and are included in all
admission costs transferred to our healthcare-provider training courses. Once
provider we understand these challenges and care princi-
Task orientated Person-centred ples, insight into the patient perspective and pro-
cess of care becomes easier for all involved:
Important inherent differences exist between Chronicity: People with diabetes battle to
the CDE DMP and more conventional Managed accept the life-long nature of their condition
Care programmes for diabetes (Table 6.2). in the mind of the patient, the traditional
However, we believe that our focus on and atten- expectation of cure is insufciently replaced
tion to the provision of integrated diabetes care, by the concept of a lifetime of control of
since our inception, has been one of the main rea- lifestyle, blood glucose, blood pressure, serum
sons for our many successes. cholesterol and body weight. Chronicity is not
only a major task for the patient to deal with,
but also for the caregiver who needs to assume
A Denition of Integrated a new professional identity as a chronic care
Diabetes Care specialist. Treatment of a chronic condition
with a physical domain bias using an acute-
Diabetes UK (2014) [17] offers the following care approach will inevitably lead to imbal-
denition: Integrated care is about designing a ance, non-compliance and failure to control.
system that focuses on the patients perspective This can be a large spiritual, psychosocial and
of care. The delivery of integrated care is facili- nancial burden.
tated by integration of the processes, methods Change: Diabetes is a life-changing condition.
and tools, which enable patients to move between One cannot hope to facilitate the process of
services according to need. Integrated diabetes change (a major developmental task in diabe-
care means vertical integration between primary, tes) and adjust to a condition that must be
community and specialist care. This is distinct mainly self-managed and that requires life-
from the wider agenda of horizontally integrated long care and control, if one cannot identify
health and social care. with and manage this process oneself. This
applies equally to patients and their caregiv-
ers. Change, however, is not an on-off switch,
Important Patient/Provider Principles but often a long and stop-start-relapse process
of Integrated Diabetes Care requiring a high degree of reection, self-
awareness, and the testing of the validity of
This denition tells us that the delivery of diabe- our own attitudes, values and beliefs about
tes services must be designed around the patients something (in this case diabetes). It is these
internal drives that determine our eventual has to balance the demands of life, diabetes and
behaviour and what our patients/clients expe- diabetes management with the emotional, spiri-
rience as a result. tual, structural, nancial, and social resources
One size does not fit all: One practitioner does available to them. Everybody must understand
not have the necessary knowledge and skills to and accept the degree of tension that is being
treat all the different effects of diabetes. Team- experienced and of the ability of the person
facilitated management was shown in the with diabetes to cope with it. This will help to
Diabetes Control and Complications Trial ensure an open, trusting care process and
(DCCT) to be a vital element in the control of achieve maximum adherence to therapy.
type 1 diabetes by Intensive Therapy [40]. Communication: Good communication is a
In addition to the doctor, input from the diabe- prerequisite to concordance patient and dia-
tes educator, dietitian, podiatrist, ophthalmol- betes team must be open and truthful at all
ogist, pharmacist, biokineticist and times and most importantly must listen to each
psychologist (amongst others), is necessary at other.
different times to maximize insight, care and Conceptions of diabetes: previous experiences
quality of life. The critical role of the diabetes of diabetes treatment successes and failures
nurse educator is reinforced by the CDE expe- result in a set of values, attitudes and beliefs,
rience over the years that every CDE Centre which guide future feelings and behaviours
that has won one of our eight Annual Clinical towards diabetes. We must assess these and
Excellence Awards, designed to recognize factor them into any treatment plan.
outstanding patient care in various categories,
has had as its coordinator, an outstanding dia-
betes nurse educator. Important Structural Principles
Self-care is vital and is a major challenge for of Integrated Diabetes Care
Health Professionals to facilitate when they
and their patients are used to conventional For the person with diabetes to self-manage their
(acute) care approaches. The focus is on the condition, they need the support of a diabetes
prevention of ill health and not its treatment. team that provides care responsive to their needs
Continuity: Care must be organized around a from diagnosis to the management of chronic
person who has a life full of events, both good complications of diabetes. Diabetes UK lists the
and bad. The aim is to build up a coherent pic- following ve key enablers of integration [17]:
ture of their needs and their health status over
time. Ideally, the same caregivers should facil- 1. Integrated Information Technology Systems
itate this care at each visit to engender trust 2. Aligned nances and responsibility
and to improve the continuity of thought, pro- 3. Care planning
cess and action. 4. Clinical engagement and leadership
Congruence in care: everyone in the Health 5. Robust clinical governance
Team should not only be giving the same infor-
mation, but should also have the same insight-
ful approach (based on a set of commonly How Does the CDE DMP Approach
shared and communicated values attitudes and Measure Up?
beliefs) towards diabetes management. The
person with diabetes will be reassured by the 1. Integrated Information Technology Systems:
agreement and harmony they see and be more All Centres on the CDE Network are obliged
condent (a feeling of self-efcacy) to prac- to use a customized internet-based clinical
tise what they have been taught. management programme to enter all patient
Concordance (or agreement within the Team, contacts, ndings, diagnosed complications,
including the patient): A person with diabetes key clinical outcomes, medications dispensed
and laboratory results. Several of the Private needs, develop and implement action plans
Pathology Laboratories in South Africa, serv- and monitor progress. People with diabetes
ing up to half the DMP patients, are able to should have active involvement in the care
upload the relevant results directly into the planning process of deciding, agreeing and
CDE system, but Centres that elect to use owning how their diabetes will be managed.
other certied laboratories need to enter their Many of the problems surrounding the pro-
results manually. Patient condentiality is vision of adequate person-centred care for
assured by a multilevel, role-dependent pass- those with diabetes revolve around the pres-
word system. The Central Administrative sures of clinical practice and a lack of time.
Ofce has real-time access to all data. This Good diabetes management requires attention
facility is utilized to download, collate and to a number of clinical parameters
check patient and Centre compliance with the 1. (Near) Normalization of blood glucose
Minimum Care Guidelines outcomes 2. Control of co-morbidities and risk factors
including HbA1c trends and complication 3. Attainment of normal growth and
rates and to present this outcomes data to development
funding organizations. This ensures that all 4. Prevention of Acute Complications
accredited providers providing care are able to 5. Screening for Chronic Complications
access and add to a patients data, obviating To t all this and a holistic, patient-centred
the need for le transfers. We can identify at collaborative approach into a busy general
risk and defaulting patients using data thresh- practice, the servicing doctor and other team
olds and follow them up. members must understand that diabetes can-
2. Aligned finances and responsibility: The not be dealt with coincidently during a
unique structure of the CDE doctor network patient consultation for an acute condition. It
allows for appropriate patient referrals to spe- requires a specic individual consultation of
cialists as required and a patient specic spe- at least half-an-hour. This can be achieved by
cialist clinical advisory service. The CDE the doctor setting aside a specic time for a
specialists make themselves available for Diabetes Clinic, be it a morning a week, a
telephonic/e-mail advice and face-to-face day a week, or a day a month, depending on
referral consultations, at no charge, for the number of patients with diabetes being
patients registered on the CDE Managed Care serviced. Each patient should have a pre-
Programme. Because of this, complicated booked half-hour appointment. Patients
patients, and their attending doctors, have should regard this as the equivalent of taking
easy access to higher levels of expertise within themselves in for a routine service every 6
the same network. months. In a resource-poor environment, one
In addition, the new CDE Managed Care can make use of group education sessions. A
Programme Model makes provision for an Registered Nurse (ideally a Diabetes Nurse
annual clinical review of every single patient, Educator) performs the vital roles of team
by a CDE specialist. For this review, the CDE coordinator and patient advocate, mentor,
specialist accesses the electronic health record counsellor, coach and self-management
of patient, and provides clinical and therapeu- facilitator. In more stretched settings, the
tic advice, to the CDE treating doctor. This DNE can also assist with measuring blood
approach ensures optimization of every pressure, weight and abdominal circumfer-
patients therapy and care strategy in a highly ence of the patients, ordering the relevant lab-
cost-effective manner. oratory investigations, and providing basic
3. Care planning: Diabetes UK (2014) [17] foot screening. The DNE can also download
denes care planning as a continuous pro- and check home blood glucose monitoring
cess, in which clinicians and patients work records. With the average number of patients
together to agree goals, identify support at any one CDE Centre seldom exceeding 200,
CDE healthcare providers have the time to are of a high standard. In the past two
provide individualized care to patients and the decades, we have trained over 6000 health-
patients feel recognized as individuals. care professionals from all over the world
CDE patients are also encouraged to regard in the principles of best-practice diabetes
their Diabetes Centre as a place where they are care.
welcome and can present for advice or discus- As an annual follow-up to the CDE 5-Day
sion with their nurse educator at short notice. Course, the CDE also hosts an annual
Patients are always asked to present their national Postgraduate Forum in Diabetes
agenda at every consultation so that their Management which all members of the
perspective is always respected and recog- CDE Network are obliged to attend. All
nized. This practical application of the other interested healthcare professionals
Medical Home concept [4143] and the are also welcome to attend. This weekend
central role of the patient in the diabetes team event, in its 18th iteration for 2016, has a
have been an integral part of the CDE philoso- busy academic programme consisting of
phy since our inception. lectures, discussion groups and workshops
4. Clinical engagement and leadership: We on current and new concepts and modali-
founded the CDE on a robust background of ties in diabetes management. A CDE
healthcare provider training and clinical and Faculty of senior endocrinologists in the
academic support, which has developed con- CDE Network presents and facilitates the
tinuously over the past two decades. Our Programme. No honorariums are offered or
Central Ofce Team in Johannesburg is pas- paid for this service and companies
sionate about teaching diabetes care to anyone involved in the provision of diabetes-
interested and we are active in exchanging related pharmaceuticals and diagnostics
diabetes knowledge with colleagues across (although welcome to participate in a con-
the world. The following mechanisms are in current trade exhibition and offer
place to attract and retain the best-skilled and Company-branded pre-Forum satellite
most passionate people in diabetes to our events) have no say or part in the develop-
Network: ment, content and presentation of any
General practitioners who are not accred- aspect of the Forum academic programme.
ited endocrinologists, and all DNEs who This assures participants that the pro-
wish to join the CDE Network, are obliged gramme will provide an objective and
to attend a comprehensive and person- unbiased review of the latest in diabetes
centred 5-Day Advanced Course in care.
Diabetes Care for Health Professionals. For any healthcare professional nationwide
We present this Course, covering all who has demonstrated a keen interest in
aspects of practical diabetes management, furthering his or her diabetes knowledge
several times yearly. All practitioners inter- and skill we facilitate subsidized
ested in diabetes are welcome to attend. (Sponsorship is sought by CDE Central
During the 5 days of the Course, the Ofce for a portion of the fees) attendance
Faculty has opportunity to identify those of Masters level Postgraduate University
attendees who show exceptional passion Diplomas in Diabetes Care (University of
for and insight into diabetes. These practi- Cardiff Diabetes Diploma/University of
tioners are encouraged to take their skill South Wales Diabetes Diploma). Both are
and interest further, as part of the CDE net- online distance learning Courses with
work or not. Many of these practitioners annual face-to-face introductory lectures at
choose to approach the CDE to accredit as the CDE Central Ofce in Johannesburg
a CDE Provider. This may be one of the and can be extended for an extra year to
reasons why practitioners in our Network earn an MSc in Diabetes. This initiative has
provided a major boost to diabetes care and hypertension are aggressively moni-
competence in South Africa. The local reg- tored and treated. Each patient is seen by
ulatory environment has effectively halted the same named team of allied healthcare
the development of local diabetes courses professionals and the same doctor. A key
(even though the expertise exists in South aspect of the CDE Network is that, while
Africa), so the CDE sought out interna- outcomes and standard of care are moni-
tional methods of up skilling our local tored, each trained doctor and Centre is
healthcare providers. A number of CDE free to treat their patients in any way they
Faculty members have Recognized prefer, with their choice of any medication
Teaching Status with these Universities or insulin, as long as adequate outcomes
and are Tutors on the Courses. and patient safety are assured. Generally,
In addition, all members of the CDE insulin therapy is started early and aggres-
Network receive complimentary copies of sively to attain and sustain glycaemic
our quarterly in-house extract of current targets.
diabetes literature, produced by one of the With these interventions, our accredited
senior endocrinologists in the Network, as practitioners are generally highly empow-
well as our Ofcial Journal for Diabetes ered, motivated, insightful and knowledge-
Healthcare Professionals, the South able individuals who really understand the
African Journal of Diabetes. demands of integrated, patient-centred
Recently, the CDE has established an care.
online Forum, which allows any CDE While Diabetes UK views engaging
Centre, or team member to post questions people with diabetes in a direct role in
and cases for comment and advice from the planning education and training needs as
CDE Faculty. vital in this key enabler of integration, peo-
The main motivation for improving out- ple with diabetes in South Africa are
comes amongst the CDE Centres appears extraordinarily apathetic in getting
to be a combination of concern to demon- involved in diabetes care at a political, gov-
strate good care and outcomes with the ernance or advocacy level. As a result, we
need to avoid hospital admission at the have not made any progress here.
doctors own expense. Overall, however, 5. Robust clinical governance: Diabetes UK
most of the Centres enjoy the opportunity (2014) [17] denes clinical governance as a
to become involved in long-term chronic system through which organizations are
disease management and in being part of a accountable for continuously improving the
successful nationwide network regarded as quality of their services and safeguarding high
Centres of Excellence by the medical standards of care by creating an environment
funders. in which excellence in clinical care will
A key aspect of the CDE Programme is ourish.
that the treating doctor and not the funder, The overriding philosophy of the DMP is
is the gate-keeper and is wholly respon- to provide total patient care for persons with
sible for all related costs. This includes diabetes utilizing a trained team of healthcare
employing or paying for the services of a professionals, including doctors specically
DNE, podiatrist and dietician. The clinical trained in diabetes management and
diabetes care given by the DMP is closely encompassing a signicant component of
aligned to the International Diabetes nancial risk sharing.
Federation denition of a standard level We assure compliance across the CDE
of care [44]. Although the DMP concen- Network to our Minimum Care Guidelines
trates on diabetes (glycaemic) control, and the principles of Good Clinical Practice
other risk factors such as dyslipidaemia in two ways. In addition to the internet-based
clinical management programme previously Medical Schemes (CMS), a statutory body

discussed, the CDE employs a full-time medi- established by the Medical Schemes Act (131
cal practitioner who conducts ongoing peer of 1998) to provide regulatory supervision of
review and audit of the participating Centres private health nancing through medical
on the CDE Network. This both by monitoring schemes. Council governance is vested in a
the data entered onto the online database and board appointed by the Minister of Health,
by visiting the Centres regularly and unan- consisting of a Non-executive Chairman,
nounced to inspect patients written records, Deputy Chairman and 13 members. The
laboratory results and reports from outside Executive Head of the Council is the Registrar,
healthcare providers. Our Auditor visits each also appointed by the Minister in terms of the
Centre on at least an annual basis. Any Centre Medical Schemes Act. The Council deter-
team found to be underperforming by not ful- mines overall policy, but day-to-day decisions
lling the CDE Minimum Care Guidelines or and management of staff are the responsibility
not following principles of Good Clinical of the Registrar and the Executive Managers.
Practice is counselled, coached on improve- The CDE is registered as an Accredited
ments required and then placed on probation Managed Care Organization with the
for 3 months. On re-audit, if the Centre has not CMS. Every 3 years we have to submit to an
rectied the deciencies, it is closed. Regular intensive assessment of the extent to which we
peer review and monitoring are part of our meet the conditions set out for accreditation
ethos. This has been an accreditation criterion by the Medical Schemes Act, including if we
for our Network since its inception. are t and proper, if we have the necessary
No incentives are offered to CDE Centres infrastructure and are nancially sound. This
for improvements in HbA1c, as this is an ensures that entities contracting with medical
expected outcome. schemes have been duly accredited as required
We have relatively few difculties in by the Act. It has been of interest to the CDE
implementing our Programme requirements that the CMS have modelled their accredita-
with our Providers. We achieve this by having tion and audit criteria for Managed Care
committed buy-in from all stakeholders before Organizations on the extensive clinical,
a CDE Centre is allowed to operate. This con- administrative, healthcare provider training
tention may be supported by a study by Pringle and mentoring and network commissioning,
et al. [45], who attempted to assess variables coordination and auditing competencies that
of process of care to determine their relative exist in the CDE Network and our Diabetes
effects. They report that patients who attended Management Programme.
a practitioner with an interest in diabetes and
those that saw a dietician had a marked posi-
tive effect on HbA1c values. This corresponds Part VIII: Conclusion
well to the principles inherent in the CDE
DMP. The CDE trains, administers and audits the big-
Since each team is responsible for not only gest network of diabetes providers in Africa, pro-
the costs of hospital admission for acute viding care excellence to many people with
diabetes-related emergencies, but also ensuring diabetes. Our capitation-based, fully integrated
good clinical outcomes, there is little resistance care model has excellent clinical and cost-
from Centres to providing the best possible efcacy outcomes, achieved largely by a
medication, even if it is more costly, to achieve geographically and economically diversied net-
these outcomes in a particular patient. work, run primarily by primary care physicians,
The CDE does not practise in a vacuum with the backing and support of a small group of
and is subject to oversight by the Council for certied endocrinologists. This demonstrates that
improved glycaemic control and better outcomes steady expansion in a nancially and politically
are achievable in a wider primary care setting. hostile environment, based on sustainable busi-
Primary care Practitioners can be trained in the ness principles (no grants or government assis-
core principles of diabetes management and tance) and person-centred, ethical care, much
attain satisfactory outcomes. However, to achieve more remains to be done. As a self-aware organi-
this, adequate training, a holistic team approach, zation, we know that we only care for a very
ongoing oversight and review and adequate small part of the South African population with
nancial reward are required to ensure service diabetes; multitudes do not receive adequate care
sustainability. Furthermore, it is apparent that and support. This we believe is a tragedy. Many
Managed Care Programmes for diabetes which of the contracted medical schemes have made
pass on the risk and gate keeping to the doctor membership of the CDE DMP voluntary, result-
and which focus on outcomes rather than cost- ing in a negative selection bias; those with newly
containment, are successful and cost-effective in diagnosed type 2 diabetes who perceive them-
both the shorter and longer terms. Fears that cap- selves to be well and whose primary care doc-
itation-based programmes may result in under tors consider them to be controlled often elect
servicing are unfounded, provided the servicing not to join. This has limited the potential growth
doctors understand the principles of chronic dis- of and inuence of this model in providing better
ease management and are judged on outcomes care to more South Africans that make use of pri-
and take risk for failure to attain these. vate healthcare.
We have achieved international recognition A recent report by the International Finance
for our work and our model of care. We are work- Corporation [46] noted that, while the role of the
ing hard to set up networks of providers and private sector in African healthcare continues to
funders to ensure our continued existence and, if be contentious, better collaboration between
possible, to help make nancial, clinical and both the public and private sectors would be the
moral sense to any future NHI model. We work most efcient way of extending high-quality
on low prot margins, enough to sustain and healthcare across the continent and crucial to
grow us as a business, but not to milk the improving healthcare provision in Africa. A
healthcare system of vital funds. recent academic study by Volminck et al. [47]
We have also provided diabetes education to looked at a cost-effectiveness analysis and
thousands of healthcare providers through the potential utility of applying the private sector
medium of evidence-based, IDF aligned diabetes CDE DMP capitation model to the management
training courses since our inception. This has of type 2 diabetes in the South African public
helped to ll a huge void in diabetes competence sector versus usual practice. Probabilistic
left open by local Universities and healthcare modelling showed all iterations of the CDE
policies. The training of healthcare providers is DMP to fall below the accepted Willingness-to-
also essential to the success of diabetes care in Pay (WTP) threshold (i.e., it was cost-effective)
the public health sector in this vein we have and that it could contribute to increased life
already provide free training via attendance at expectancy in South Africa. The study recom-
our 5-Day Courses for nearly 50 public health mended that a pilot study of the CDE DMP be
employed health professionals. In addition to our done to explore the practical translation of this
local Courses and national meetings, we are cur- analysis. Currently, however, the CDE lacks the
rently enabling healthcare providers to obtain political recognition and acceptance to enable a
Masters level diabetes qualications at very little public-private partnership (PPP) with the South
cost to themselves. We have changed diabetes African Department of Health. We trust that this
care in South Africa and beyond. status quo will change and open an exploration
Although we have record of more than two of the possibility of better diabetes care to our
decades of successful community practice and population at large.
Abbreviations 2013. Available from: http://www.idf.org/

diabetesatlas.
6. International Diabetes Federation. IDF diabetes atlas
ACE-I Angiotensin-converting- enzyme update poster. 6th ed. Brussels: International Diabetes
inhibitors Federation; 2014. Available from: http://www.idf.org/
ADA American Diabetes Association sites/default/les/Atlas-poster-2014_EN.pdf.
7. Li R, Zhang P, Barker LE, Chowdhury FM, Zhang
ARB Angiotensin receptor blockers
Z. Cost-effectiveness of interventions to prevent and
CDE Centre for Diabetes and control diabetes mellitus: a systematic review.
Endocrinology Diabetes Care. 2010;33:187294.
CMS Council for Medical Schemes 8. Department of Health Schemes Act 131 of 1998.
Regulations in terms of the medical schemes act, 1998
DCCT Diabetes Control and Complications
(Act No. 131 of 1998). Chapter 3, Contributions and
Trial benets, Regulation 8. Prescribed Minimum Benets.
DNE Diabetes Nurse Educator Available from: http://www.selfmed.co.za/PDF/
DMP Diabetes Management Programme GNR1262_20_October_1999.pdf
9. Jnsson B, CODE-2 Advisory Board. Revealing the
DKA Diabetic ketoacidosis
cost of type II diabetes in Europe. Diabetologia.
HONK Hyperosmolar non-ketotic coma 2002;45:S512.
HQA Health Quality Assessment 10. Council for Medical Schemes. Chronic disease algo-
NHANES National Health and Nutrition rithms diabetes mellitus type 2 (Undated). Available
from: http://www.medicalschemes.com/les/
Examination Survey
Prescribed%20Minimum%20Benefits/
IDF International Diabetes Federation DiabetesMellitus1_2.pdf
NHI National Health Insurance 11. Council for Medical Schemes. Chronic disease algo-
PMB Prescribed Minimum Benets rithms diabetes mellitus type 2 (Undated). Available
from: http://www.medicalschemes.com/les/
PPP Public-Private Partnerships
Prescribed%20Minimum%20Benefits/
UKPDS United Kingdom Prospective DiabetesMellitus2_3.pdf
Diabetes Study 12. Council for Medical Schemes. How prescribed mini-
VISS Vascular diabetic complications In mum benets help you manage diabetes. cmscript
Issue 8 of 20102011. Available from: http://www.
Southeast Sweden
medicalschemes.com/files/CMScript/
WESDR Wisconsin Epidemiologic Study of CMScript8Of2010_2011.pdf
Diabetic Retinopathy 13. Amod A, Ascott-Evans BH, Berg GI, et al. The 2012
SEMDSA guideline for the management of type 2
diabetes (revised). Diabetes self-management educa-
tion. JEMDSA. 2012;17(2 Suppl 1):S134.
References 14. Williams R, Van Gaal L, Lucioni C, CODE-2
Advisory Board. Assessing the impact of complica-
1. Statistics South Africa. Mid-year population esti- tions on the costs of type II diabetes. Diabetologia.
mates, South Africa. Statistical release P0302; 2014. 2002;45:S137.
Available from: http://www.statssa.gov.za/publica- 15. American Diabetes Association. Economic costs of diabe-
tions/P0302/P03022014.pdf tes in the US in 2012. Diabetes Care. 2013;36:103346.
2. Bridging the gap in South Africa. Bull World Health 16. HQA. 2012 Health quality assessment industry report.
Org. 2010 Nov;88(11):797876. Available from: Towers Watson Actuaries and Consultants (Pty) Ltd;
h t t p : / / w w w. w h o . i n t / b u l l e t i n / v o l - 2013 Jan.
umes/88/11/10-021110/en/ 17. Improving the delivery of adult diabetes care through
3. Annual Report of the Council of Medical Schemes integration. Diabetes UK. 2014. Available from:
(CMS) 2013/2014. Available from: https://www.med- h t t p s : / / w w w. d i a b e t e s . o r g . u k / D o c u m e n t s /
icalschemes.com/files/Annual%20Reports/ About%20Us/What%20we%20say/Integrated%20
AR2013_2014LR.pdf diabetes%20care%20(PDF,%20648KB).pdf
4. The Economist Intelligence Unit Limited. The future 18. The Diabetes Control and Complications Trial
of healthcare in Africa. The Economist. 2012. Research Group. The effect of intensive treatment of
Available from: http://www.economistinsights.com/ diabetes on the development and progression of long-
sites/default/files/downloads/EIU-Janssen_ term complications in insulin-dependent diabetes
HealthcareAfrica_Report_Web.pdf mellitus. The Diabetes Control and Complications
5. International Diabetes Federation. IDF diabetes atlas. Trial Research Group. N Engl J Med. 1993. Sep
6th ed. Brussels: International Diabetes Federation; 30;329(14):97786.
19. Distiller LA, Brown MA, Joffe BI, Kramer BD, 33. Remuzzi G, Schieppati A, Ruggenenti P. Nephropathy
Organisation and Delivery of Care. Striving for the in patients with type 2 diabetes. N Engl J Med.
impossible dream: a community-based multi-practice 2002;346:114551.
collaborative model of diabetes management. Diabet 34. Klein R, Klein BEK, Moss SE, Davis MD, DeMets
Med. 2010;27:197202. DL. The Wisconsin epidemiologic study of diabetic
20. Nordwall M, Abrahamsson M, Dhir M, Fredrikson M, retinopathy. II. Prevalence and risk of diabetic reti-
Ludgsson J, Arnqvist HJ. Impact of HbA1c, fol- nopathy when age at diagnosis is less than 30 years.
lowed from onset of type 1 diabetes, on the develop- Arch Ophthalmol. 1984;102:5206.
ment of severe retinopathy and nephropathy: the VISS 35. Klein R, Klein BEK, Moss SE, Davis MD, DeMets
study (vascular diabetic complications in Southeast DL. The Wisconsin epidemiologic study of diabetic
Sweden. Diabetes Care. 2015;38:30815. retinopathy. III. Prevalence and risk of diabetic reti-
21. The Diabetes Control and Complications Trial/ nopathy when age at diagnosis is 30 or more years.
Epidemiology of Diabetes Interventions and Arch Ophthalmol. 1984;102:52732.
Complications Research Group. Retinopathy and 36. American Diabetes Association. Retinopathy in dia-
nephropathy in patients with type 1 diabetes four betes (position statement). Diabetes Care. 2004;27
years after a trial of intensive therapy. N Engl J Med. Suppl 1:S847.
2000;42:3819. 37. Thomas RL, Distiller LA, Luzio SD, Chowdhury SR,
22. Saunders SA, Wallymahmed M, MacFarlane Melville VJ, Kramer B, Owens DR. Ethnic differences
IA. Glycaemic control in a type 1 diabetes clinic for in the prevalence of diabetic retinopathy in persons
younger adults. Q J Med. 2004;97:57580. with diabetes when rst presenting at a diabetes clinic
23. Holman RR, Paul SK, Angelyn Bethel M, Matthews in South Africa. Diabetes Care. 2013;36:33641.
DR, Neil HAW. 10-year follow-up of intensive glu- 38. Brown MAJ. Core concepts in diabetes mellitus. S
cose control in type 2 diabetes. N Engl J Med. Afr J Diabetes. 2010;3(1):103.
2008;359:157789. 39. Brown MAJ. Changing diabetes care. S Afr
24. Ettaro L, Songer TJ, Zhang P, Engelgau MM. The J Diabetes. 2010;3(2):235.
economic burden of diabetes. Pharmacoeconomics. 40. Diabetes Control and Complications Trial Research
2004;22:14964. Group. The effect of intensive treatment of diabetes
25. American Diabetes Association. Economic Costs of on the development and progression of long-term
Diabetes in the U.S. in 2012. Diabetes Care. complications in insulin-dependent diabetes mellitus.
2013;36:103346. N Engl J Med. 1993;329:97786.
26. Kitabchi AE, Umpierrez GE, Fisher JN, Murphy MB, 41. American Academy of Family Physicians, American
Stentz FB. Thirty years of personal experience in College of Physicians, American Osteopathic
hyperglycemic crises: diabetic ketoacidosis and Association, American Academy of Pediatrics. Joint
hyperglycemic hyperosmolar state. J Clin Endocrinol principles of the patient-centered medical home (PC-
Metab. 2008;93:154152. MH) [Internet], 2007 Mar 5. Available from: http://
27. Siderov J, Shull R, Tomcavage J, Girolami S, Lawton www.acponline.org/pressroom/pcmh.htm
N, Harris R. Does diabetes disease management save 42. Iglehart JK. No place like home-testing a new model
money and improve outcomes? Diabetes Care. of care delivery. N Engl J Med. 2008;359:12002.
2002;25:6849. 43. Fisher ES. Building a medical neighbourhood for the
28. Steffans B. Cost effective management of type 2 dia- medical home. N Engl J Med. 2008;359:12025.
betes: providing quality care in a cost constrained 44. International Diabetes Federation Clinical Guidelines
environment. Am J Manag Care. 2000;6:S697703. Task Force. Global guideline for type 2 diabetes.
29. Wagner EH, Sandhu S, Newton KM, McCulloch DK, Brussels: International Diabetes Federation; 2005.
Ramsey SD, Grothaus LC. Effect of improved glyce- 45. Pringle M, Stewart-Evans C, Coupland C, Williams I,
mic control on health care costs and utilization. Allison S, Sterland J. Inuences on control in diabetes
JAMA. 2001;285:18290. mellitus: patient, doctor, practice, or delivery of care?
30. Menzin J, Langley-Hawthorne C, Friedman M, BMJ. 1993;306:6304.
Boulanger L, Cavanaugh R. Potential short-term eco- 46. International Finance Corporation. Healthy partner-
nomic benets of improved glycemic control. ships: how governments can engage the private sector
Diabetes Care. 2001;24:515. to improve health in Africa. Washington, DC: World
31. Simmons D, Yu D, Wenzel H. Changes in hospital Bank; 2011. Available from: https://openknowledge.
admissions and inpatient tariff associated with a dia- worldbank.org/handle/10986/2304.
betes integrated care initiative: preliminary ndings. 47. Volmink HC, Bertram MY, Jina R, Wade AN, Hofman
J Diabetes. 2014;6:819. KJ. Applying a private sector capitation model to the
32. Mitka M. Report quanties diabetes complications. management of type 2 diabetes in the South African
JAMA. 2007;297:23378. public sector: a cost-effectiveness analysis. BMC
Health Serv Res. 2014;14:444.
English Approaches to Integrated
Diabetes Care: The East 7
Cambridgeshire and Fenland
Diabetes Integrated Care Initiative:
A Multiple Provider Approach
David Simmons, Dahai Yu, and Helmut Wenzel
Background However, increasingly, another driving force

has been to reduce payments to hospitals for the
In England, there have been several approaches tariff, the sum payable per clinic appointment
to replace the hospital based, medical consultant- under the English NHS internal market system.
led diabetes clinic. The rationale behind these Essentially, the tariff is a mandatory payment for
new models has varied. In some cases, there was the clinical service including the facility cost (i.e.,
a genuine desire to improve access to more spe- the costs of, e.g., management overheads, any
cialist levels of care by reducing demand, through underpaid activity such as diabetes inpatients [5],
diversion of patients with less complex needs to health professional training and safety activities).
other clinicians with diabetes management skills, A GPSI or nursing service could therefore run a
beyond those of the average general practice. In clinic in an inexpensive community facility,
other cases, initiatives have upskilled and sup- sometimes with the staffing costs covered by a
ported existing practices across an area to deliver prior block payment making the service itself
care closer to home, and within the holistic care free from a commissioners point of view.
than can be provided by the medical home Similarly, use of a private service may mean that
delivered by quality general practice. Examples the pure facility costs are cheaper, allowing a
have been the General Practitioner with a Special lower price to be paid. However, reductions in
Interest (GPSI) [1], the community based diabe- activity and income for hospitals, increases the
tes specialist nurse (DSN) led service [2, 3] and average unit overhead costs and makes hospitals
intermediate services [3, 4]. increasingly unaffordable under the constrained
budgets of the English NHS. It is not too surpris-
ing that 76 % of English hospitals were in deficit
in 2015 [6].
D. Simmons (*)
School of Medicine, Western Sydney University, The evidence that in England, GPSI and inter-
Sydney, NSW, Australia mediate services provide equal or better services
e-mail: [email protected] than a hospital based multidisciplinary service, is
D. Yu limited. Furthermore, there are no externally vali-
Research Institute for Primary Care dated training programmes for GPSIs or diabetes
and Health Sciences, Keele University,
specialist nurses in England (unlike diabetes edu-
Newcastle Under Lyme, UK
e-mail: [email protected] cator programmes in the USA, Canada, Australia
and New Zealand). In England, some DSN led
H. Wenzel
Health Economist, Konstanz, Germany community services have been associated with
e-mail: [email protected] HbA1c reduction among patients under GP care

108 D. Simmons et al.
[3]. However, we now know that reducing the Diabetes Integrated Care Initiative (DICI) was an
HbA1c too far and fast in some patients can be attempt to create such a systematic approach
harmful [7]. This is a particularly important issue, [1012]. We now describe the Initiative, its prog-
where primary care is paid through the Quality ress and the issues that were identified.
Outcomes Framework (QoF), a general practice
pay for performance programme [8]. A major
item within QoF, is the proportion of patients East Cambs and Fenland
below HbA1c criteria: such reporting is not linked
to rates of hypoglycaemia, ambulance call outs or East Cambridgeshire and Fenland (ECF) were
hospitalisation, i.e., a practice could receive a high selected for the intervention, as an area with a
payment through achieving the QoF target, but selection of historically poor diabetes outcomes
with a high hospitalisation/ambulance callout rate. [11] and a number of socioeconomically deprived
Potential consequences of the introduction of communities. The area is largely rural, a 2009 pop-
multiple service providers are fragmentation of ulation of 160,000, with a diabetes population of
care, reductions in continuity of care and propa- 7790. There are 17 general practices and no major
gation of a reluctance to refer on to a more spe- hospital, falling within 5 (four substantial) hospital
cialist service [9]. Within this paradigm, the catchment areas from adjacent areas. These hospi-
creation of a local health system that can inte- tals require at least 30 min travel time, associated
grate primary, secondary and community diabe- with limited public transport and parking chal-
tes care, sharing the work while getting the best lenges. Cambridgeshire includes two other areas:
from each, would seem to be a sensible way for- Huntingdonshire and Cambridge City/South
ward. The East Cambridgeshire and Fenland Cambridgeshire as shown in Fig. 7.1, which also
Fig. 7.1 Locations of the three areas in Cambridgeshire hospital diabetes clinics and the diamonds the location of
(Left). Primary Care Trusts in the East of England are three community hospitals: the black diamond is Ely
shown on the right. The stars show the location of the five hospital
7 English Approaches to Integrated Diabetes Care: The East Cambridgeshire 109
shows the location of the former Cambridgeshire between the community service and
Primary Care Trust (CPCT), the commissioners of Addenbrookes Hospital on how to address ECF
NHS care at the time of the DICI. CPCT paid hos- needs. In April 2008, the two services established
pitals for each patient seen as inpatients or outpa- a monthly, joint community based clinic in Ely,
tients under the national tariff, but paid the in the heart of East Cambridgeshire. The clinic
community services by a block contract, i.e., a included consultant physician, DSN, registrar,
previously agreed single amount for services with HCA and dietitian. This clinic was unfunded
an expectation to achieve key performance indica- work for the hospital, which transferred Ely
tors (KPIs). General practitioners (GPs) were patients, currently being seen in the hospital
largely paid under a GP contract, with a sizable clinic, to a multidisciplinary clinic in the com-
proportion of practice income derived from QoF. munity hospital, i.e., the hospital had a reduction
A separate DSN led community service had in income from tariff. The close working between
been introduced into ECF in 2003 [2] providing Addenbrookes hospital and the community ser-
type 2 diabetes education and seeing patients vice led to the re-establishment of an ECF
within general practices and their homes. The Diabetes Network across GPs and local specialist
service included a health-care assistant (HCA): services in September 2008 [14]. From November
an unregistered practitioner, working under the 2008, Addenbrookes hospital and the community
governance of the two DSNs. Dietetics, podiatry service then began unfunded work with one pilot
and the diabetes specialist nursing service in general practice to develop approaches for a more
one of the neighbouring hospitals were deliv- integrated practice to diabetes care from a pri-
ered by separate community services, within the mary care point of view.
same provider organisation. Most of the general
practices had a diabetes lead GP and practice
nurse who had attended additional health pro- The East Cambridgeshire
fessional diabetes education. The ECF DSN led and Fenland Diabetes Integrated
service provided conferences for health-care Care Initiative
professional diabetes education. Relationships
with hospitals services were limited, including A proposal for the ECF Diabetes Integrated Care
with Addenbrookes Hospital, the local tertiary Initiative (DICI), with an increased specialist
facility. Addenbrookes, and one other hospital, team, was subsequently submitted to CPCT for
run the evidence based and nationally audited funding. Funding (250,000 pa) commenced on
Dose Adjustment For Normal Eating (DAFNE) 1st April 2009, with additional staff members
structured type 1 diabetes education pro- largely appointed by July/August 2009. The con-
grammes [13]. The three other hospitals run tract was additional to the DSN service contract,
self-designed structured type 1 diabetes educa- but was still held and managed by the community
tion programmes. Addenbrookes hospital also service, with the diabetes consultant subcon-
runs the major continuous subcutaneous insulin tracted from Addenbrookes hospital. Registrar
infusion (CSII) service, regional diabetes foot time was provided by Addenbrookes hospital at
service and regional diabetes in pregnancy no charge. The aims of the service were to:
service.
(a) reduce hospitalisation in patients with diabe-
tes, particularly diabetic foot disease
The Pre-intervention Period: (b) improve metabolic control (commencing
The Ely Clinic with HbA1c/hypoglycaemia/weight control
and then moving onto other measures such as
The ECF DSN led service had positive relation- blood pressure, lipids, smoking)
ships with most of the general practices in the (c) improve safety and reduce medication side
area, providing a foundation upon which to build effects where possible (e.g., weight gain,
the DICI. In 2007, discussions commenced hypoglycaemia)
(d) increase access to diabetes education sessions, making travel outside the area unneces-
(e) increase access to diabetes specialist ser- sary for some patients.
vices (e.g., medical, nursing, dietetic and
podiatric services including full multidisci- Primary Care Support and Care Closer
plinary clinics) in ECF to Home
(f) reduce the need for hospital clinic attendance This included three major initiatives:
(g) reduce the risk of complications among those
with diabetes (e.g., end stage renal failure, Community/general practice based DSN and/
acute myocardial infarction and other macro- or new DSD clinics saw patients within the
vascular disease, diabetic eye disease, dia- general practice (locations shown in Fig. 7.2),
betic foot disease) potentially with the GP/practice nurse, other-
wise in a room set aside for the purpose. The
DSN/DSD had access to the within-practice
Interventions clinical records but also wrote up within the
community service electronic record
There were six groups of interventions, discussed (SystmOne). This double recording was
next. required because, generally, the community
service SystmOne did not link to the practice
Community Diabetes Specialist electronic record even if both used SystmOne.
Workforce Expansion The DSD clinics ran in parallel with the com-
This involved an increase in DSNs by 1.8 Whole munity general dietitian clinics. The DSD ses-
Time Equivalent (WTE), diabetes specialist podi- sions referral criteria were:
atrist (DSP) by 1.0WTE, diabetes specialist dieti- T2DM with HbA1c >9 %
tian (DSD) by 1.0WTE, community diabetologist T2DM initiating insulin
(CD) by 0.3WTE and HCA by 1.0WTE. This T2DM on insulin with hypos or erratic
allowed allocation of one DSN to every four to glucose
six practices and the DSD, DSP and CD across T2DM with chronic kidney disease IV+
all practices. There was also more time for home seen outside a multidisciplinary clinic
visits and telephone support to patients. T1DM seen outside a multidisciplinary
clinic
Increased Access to Patient Structured A second closer to home intervention were
Education the multidisciplinary clinics (DSN, DSD, DC,
Access to a locally developed structured, com- registrar with HCA) which included an addi-
munity delivered, type 2 diabetes education was tional monthly Ely clinic and a new monthly
limited due to low staffing levels: DICI brought clinic in Wisbech in the north of the county.
more groups per month. Prior to DICI, ECF type An existing monthly clinic in Doddington
1 diabetes patients had limited access to with a DC from Peterborough Hospital, had a
DAFNE. Although reportedly, there was reluc- DSD added: the existing DSN continued.
tance by some patients to travel out of the area, Criteria for the clinics were generally the
and for those in the north of the area, this was a same as for a hospital based clinic, but gener-
long distance, the main issues appeared more to ally, those on continuous subcutaneous insulin
be due to a patchy reluctance to refer to infusion (CSII) or had undertaken DAFNE (as
Addenbrookes Hospital, where the largest pro- none of the community DSNs were DAFNE
gramme was underway. The integrated approach educators) were excluded.
established communication lines between the The third primary care support/care closer to
DICI staff and DAFNE coordinator. home intervention were shared educational/
Addenbrookes was also able to establish a num- clinical care sessions (virtual clinics) where
ber of ECF community hospital based DAFNE the management of 520 patients with either
Fig. 7.2 Community/

general practice based
DSD/specialist podiatry
clinics and consultant
led community
multidisciplinary (MDT)
diabetes specialist
clinics
an HbA1c 9 %, or considered at high risk of New Diabetes Specialist Podiatrist

hospitalisation, were systematically dis- (DSP) Service
cussed. The time for GP/practice nurse par- This service was created to upskill primary care,
ticipation was initially covered by a ensuring they were aware of the pathways for
countywide payment (known as a Local subacute/acute foot issues, and to be available to
Enhanced Service agreement), between April see patients urgently (including in their homes).
2008 and March 2010. After this time, Clinics were established in 14 practices and at
practices received no payment for their par- Ely and Doddington Hospitals. Criteria for refer-
ticipation and clinic frequency tailed off. ral to the DSP were:
foot lesion at risk of hospitalisation based upon the Barriers to Diabetes Care frame-
foot lesion requiring urgent attention work (Fig. 7.3) developed and used in New
foot lesion requiring higher level of knowl- Zealand and extended to use in the USA and
edge/skill than community podiatrist Australia [1618]. Table 7.1 shows the barriers
recent discharge from hospital and possible strategies to address them [15].
Implementation involved practice nurse training
The DSP also established step-up step-down by the DSN, and use of a variety of materials/
clinics (i.e., clinics for patients with lesser acuity resources including referral pathways. The mate-
and therefore needed lesser input than the high rials were field tested in one general practice. The
risk foot clinics) in the three community hospi- practice nurse reported that the approach changed
tals to allow either earlier discharge from hospital the focus of the Annual Review to being more
clinics/inpatients or to reduce hospital clinic patient led rather than led by recording informa-
attendance frequency. The DSP liaised with the tion with care becoming more individualised
community podiatrists and the lead diabetes spe- with more appropriate use of local resources.
cialist podiatrists at each of the neighbouring However, it was Time consuming need at least
hospitals. The podiatrist worked 1 day/month at 30 min for consultation with majority of baseline
the regional foot clinic at Addenbrookes initially recorded prior to this, when previously would
to maintain and extend her skills. only have blood tests and blood pressure [10].
Addressing Barriers to Care and Self Organisational Redesign

Care in General Practice Organisational redesign was seen as essential to
A DSN (0.5WTE) was allocated to visiting prac- align the hospital, community and GP services
tices and establishing a within-practice system to provided to people with diabetes. This involved
identify and address hurdles faced by both three levels of mechanisms to integrate the differ-
patients and practice staff [15]. The approach was ent services:
Psychosocial Psychological
External Patient Internal

Physical Physical
Educational
Key:
Psychosocial barriers=how patients relate to others including health care professionals, family members
Psychological=how patients behave, feel, perceive, believe
External physical=systems issues such as how the health and social welfare services operate
Internal physical=other health conditions impact on self care, e.g., blindness, obesity, disability
Educational=knowledge of diabetes, health systems and educational attainment
Fig. 7.3 Barriers to Diabetes Care major components (Data from [15, 16])
Table 7.1 Barriers to care, approaches to assessing barriers and interventions that may be beneficial in addressing the
impact of barriers on diabetes care/self care
Barrier and examples Initial assessment External pathway
Psychological issues-normal Detailed assessment Health-care assistant with support
Strictness of regimen (e.g. Goal setting from diabetes specialist nurse
food)
Motivation, laziness, self Motivational interviewing
efficacy
Psychological issues Detailed assessment Diabetes specialist nurse
Unusual e.g. needle phobia Confirm e.g. needle phobia Psychologist
Denial, emotional Counselling
Psychological issues Detailed assessment Finance/work related:
Priority setting Personal time management, budgeting, Citizen Advice Bureau
social support, access to care,
counselling
Time allocation Coping skills, time management Diabetes specialist nurse home visit if
needed
Psychological issues Personalised strategies Discuss with psychologists
Unhelpful health beliefs Education
Cognitive behavioural therapy
Family issues Detailed assessment-identify abuse May need social services involvement
Inadequate support and Make joint family involvement/ Diabetes specialist nurse home visit
obstruction education plan
Bring partner/family member to appt Dietician to see family
Practice nurse home visit District nurses can do annual reviews
if trained up
Unsupportive environment Education e.g. on how to use current Discuss with diabetes specialist nurse
local milieu/products or dietician if needed
Food, activity, insulin Meal/activity plans Dietician to see family
injections, self glucose Dietician should make a plan for each
monitoring area and hand to practise re: food and
activity pack
Past care unsatisfactory Detailed assessment- If due to patient (recurrent behaviour
pattern) then psychologist
Attitudes of health workers Identify ethnicity/gender issues Alternative personnel (difficult in
Personalised care, build trust, takes small practices) diabetes specialist
time nurse
Needs staff training programme
Communication Assessment Discuss with diabetes specialist nurse
Low educational status Identify issues Local education authority
Personal communication plan (e.g.
lower Fleisher score materials, more
time, education)
Physical co-morbidities Assessment Community matron, community
(non-diabetes) services referral-specialist nurse,
Identify management plan and practise May need specialist advice and/or
nurse role referral
Psychiatric co-morbidities Assessment Community mental health team or
Depression Identify management plan and practise psycho geriatric advice/referral
nurse role
Diabetes management side Improved diabetes tools e.g. glucose Discuss with diabetes specialist nurse
effects monitoring and insulin needles,
different medications
(continued)
Table 7.1 (continued)

Barrier and examples Initial assessment External pathway
Educational issues Practise nurse training Discuss with diabetes specialist nurse
Diabetes Group sessions, assessment
Materials, 1:1, group education
Education issues Assessment Home visits, out of hours service
Accessing services Materials Discuss with diabetes specialist nurse
Education Materials
Personal finance Practise nurse to work through Social services
Healthy food, footwear, Citizen Advice Bureau
transport costs, time off work
Physical access Transport Discuss with diabetes specialist
nurse/dietician/podiatry
Care closer to home-special clinics/ Annual review by district nurse
visits (e.g. prison)
Home visits
Poor range of services Evening/weekends, emergencies Discuss with diabetes specialist
Exercise/walking groups, supermarket nurse/dietician home visits
tours
Appointment system Information management, Staff Practise to review systems
management
Staffing numbers/multi-skilling
Used with permission from Harwood et al. [15]
DSN Diabetes Specialist Nurse
Vertical integration, i.e., bringing together the 4. Patient reference group: The patient voice is
different parties to direct the DICI crucial to allow the development within the
Clinical governance, i.e., the oversight of initiative to be grounded in the patient per-
practitioners within the DICI spective. The original plans were reviewed
Horizontal integration, i.e., bringing together and commented upon by the Addenbrookes
the different practitioners delivering care Hospital diabetes care patient advisory com-
mittee (DCPAC) which included some
Mechanisms for Vertical Integration patients from ECF. Attempts to establish an
1. ECF diabetes network: This group helped ECF group were unsuccessful.
develop, and approved, the submission for the 5. Countywide network: The countywide net-
integrated care approach. It was designed to work was created by combining the pre-
serve as a clinical reference group and provide existing networks from the three areas in
governance over the DICI. Cambridgeshire. It provided a forum for dia-
2. CPCT Integrated diabetes care initiative proj- betes health professionals, people with diabe-
ect group: Meetings provided a platform for tes and commissioners to discuss area-specific
the integrated team to report to CPCT and and countywide issues, to monitor diabetes
raise issues and link to other countywide care and to formulate countywide proposals.
initiatives. 6. Cambridgeshire diabetes drugs advisory
3. Secondary provider reference group: This group: This group was established after the
group was established to optimise communi- countywide network advised the CPCT medi-
cation and agreement between the hospital cines management team of the need for pro-
diabetes services in the area. posals relating to diabetes drugs and devices
to be discussed with diabetes health profes- ment of the DC within the community team,
sionals and people with diabetes before CPCT closer working across the services including
policy was made. the multidiscipinary community clinics, tele-
7. Countywide information management work- phone/email advisory and team meetings, pro-
ing group: This working group was set up by fessional support, upskilling and maintenance
the diabetes networks to develop and put into of standards for DSP and DSD. A joint
place a countywide approach to diabetes approach to DAFNE, initial induction of com-
information management. Membership munity staff, opportunities for hospital staff to
included the commissioners, national and work in the community and back up staff
CPCT/local provider IT staff, GP and diabetes members and for community staff to attend
service representatives. hospital clinics were also estasblished.
8. Countywide guidelines and pathways work- 2. Integration between the community trust and
ing group: This working group collated and other diabetes services: DSP worked with all
updated existing diabetes guidelines and path- other local hospital foot services and commu-
ways from across the three areas. nity podiatry, DSD worked with community
9. Countywide diabetes education working dietetics and the DSN worked with
group: This group generated recommenda- Peterborough Hospital staff in Doddington.
tions for diabetes education. Close working with the 17 GP teams with
nine practice-based activities helped integrate
Clinical Governance, i.e., the Oversight care. Cooperation with community matrons
of Practitioners Within the DICI and district nurses was already part of the
Governance is about how local [public] bodies approach used.
and partnerships ensure that they are doing the 3. Integration with other secondary services
right things, in the right way, for the right people including mental health services were planned
in a timely inclusive, open, honest and account- (including psychology) but did not eventuate.
able manner [19]. Clinical governance is the
term used to describe a systematic approach to
maintaining and improving the safety and quality Evaluation
of patient care within a health system. The day to
day activities of the DICI team were under the There were three phases of evaluation:
clinical governance of the team leader (lead
DSN), who reported to the community provider simple satisfaction questionnaires in the Ely
manager. A team meeting was held monthly for clinic before the DICI commenced and evalu-
discussion of policy, key cases and incidents. ation of the initial practice intervention
Technically, the overall clinical governance for service evaluation with some modeling of
the ECF DICI was the responsibility of the DC, hospitalization at the end of year 1 [11]
but this depended upon the monthly team meet- full mixed methods evaluation at the end of
ing, meeting with the lead DSN and any ad hoc year 3 [12]
clinical observations. General practice teams,
including practice nurses, received their clinical Ely Clinic Review
governance through the GPs, and GP clinical Of the 207 ECF patients seen at Addenbrookes
governance came from CPCT. Hospital in the previous 12 months, 70 were allo-
cated to the Ely Clinic in the first 10 months of
Horizontal Integration, i.e., Bringing operation. Fifty were given appointments; others
Together the Different Practitioners either declined, had already been discharged/
Delivering Care died, were DAFNE graduates/awaiting DAFNE,
1. Integration between the community trust and were on insulin pump therapy, or were under the
CUH diabetes services: this included place- high risk foot clinic, Young Adult or nephropathy
clinics. Of the 45 attending patients, 5 were dis- nity service notes. These notes were carried in
charged and 3 returned to Addenbrookes for type the car of the community DSN with its associ-
1 diabetes or joint lipid management. A short ated information governance risk.
patient satisfaction was distributed at each clinic Information management: The electronic link
and 24 were returned (06/clinic). All patients with Addenbrookes worked well, but took the
found the Ely clinic easier to attend, were happy first one to two clinics to implement. Notes
with the advice given and would recommend the within the clinic were paper based.
clinic to others. Only one negative clinical issue
arose: the IT link between the clinic and Evaluation of the Initial Practice Based
Addenbrookes Hospital. Among the 21 patients Intervention Including Virtual Clinics
attending the clinic at least twice in this period, This pilot occurred November 2008August
there were no significant differences in metabolic 2009 in a single handed rural general practice.
results besides diastolic blood pressure which The practice diabetes register included 87 patients
was significantly lower at the final Ely clinic visit in April 2009. The practice had always worked in
(79 9 vs 74 8 mm Hg, p = 0.022, paired t-test). partnership with the community DSN team,
This was felt to be associated with the easier referring patients for assessment for commence-
travel and parking arrangements. ment on insulin, and for advice and management
The Specialist Registrar saw two major areas of patients with a persistently raised HbA1c. The
of benefit from attending the Ely clinic: practice team members had diabetes certification
through a UK University course. A practice recall
Training system was in place. In November 2008 the prac-
close consultant supervision with exem- tice audited its care and commenced closer work
plary teaching with the community team, including the DC. The
benefits from closer contact with DSD and practice nurse also embarked on a further diabe-
DSN tes qualification. Criteria for patient discussion
easier patient follow-up: experience effec- were: HbA1c of 9 %, uncontrolled hypoglycae-
tiveness of treatments, investigations mia, blood pressure or lipids in spite medication,
Continuity of Care triglycerides >8 mmol/l, hospitalisation/CVD
patient rapport: appreciate seeing same events in past 12 months, patients under second-
face, patients appear more engaged in dia- ary care.
betes care At the end of the 8 month pilot, the proportion
able to follow-up results, previous with HbA1c <7.5 % increased from 47 % to 75 %
discussions and those with a HbA1c 9 % dropped from
less time spent looking through old notes 39 % to 12 %. There were various other process
and more time spent in consultation benefits. The barriers tool was piloted.
A second, larger practice (n = 655) adopted the
A range of issues arose during the pilot, as approach once DICI was funded with six virtual
noted next: clinics and increased DSN input. This practice
also had a dramatic reduction in the number of
Organisation of appointments: The appoint- patients with an HbA1c of 9 % from 248/655
ments were made manually by Addenbrookes (37.9 %) to 83 (12.7 %) over 12 months.
Hospital secretarial staff, but the Addenbrookes
Hospital booking system, at times, continued Outcomes Evaluation After 1
to send appointments to patients creating and 3 Years of Intervention [11, 12]
confusion.
Management of medical records: The limited Methodology
time between filling appointments and the De-identified electronic Cambridgeshire
clinic posed problems for generating commu- Secondary Uses Service (SUS) data were
obtained for all patients for any past admission lead DSN allocated the practices blind to the hos-
and hospital clinic attendance between April pitalisation data. Analyses compared the engaged
2007 (i.e., 2 years before the DICI contract com- and less engaged practice hospitalisation rates.
menced) and November 2010 for the 1 year eval- In the 3 year analysis, the Mantel Haenszel
uation and March 2012 for the 3 year evaluation. test was used to compare the proportion of hospi-
Practice, patient age, elective/non-elective status, tal episodes by patients with diabetes in ECF
ICD10 and Health Related Group (HRG) coding with those in the other two areas in Cambridgeshire
were included in the datasets. Diabetes (E10 (Huntingdonshire, Greater Cambridge) in 2007
E14) was considered the primary cause of admis- and 2012. Hospital episodes and outpatient atten-
sion if coded in the first ICD field. Diabetes foot dances were compared before and after the com-
was considered the cause of admission with HRG mencement of the DICI, between those with and
codes of KB03A, KB03B, QZ02A, QZ04Z, without diabetes and between the three
QZ11A, QZ11B, QZ12Z, diabetic ketoacidosis Cambridgeshire areas. Each monthly point shows
with ICD 10 codes of E10.1, E11.1, E12.1, E13.1, the 95 % confidence intervals for the estimate of
and E14.1, and CVD with ICD10 codes of I20, mean change.
I21, I22, I23, I24, and I25. QoF population (2008/9) data (http://www.
Because of the varying underlying patterns in gpcontract.co.uk/download) provided diabetes
admissions in the different areas, regression anal- and overall population by area and the Quality
ysis across the whole time period was not seen as Outcomes Framework metrics (diabetes overall,
a valid approach. A novel approach to assessing DM12 (BP 145/85 or less), DM17 (Cholesterol
the impact of the intervention was required to 5.0 or less), DM23 (HbA1c 7 % or less), DM24
compare changes in hospitalisation before and (HbA1c 8 % or less), DM25 (HbA1c 9 % or less))
after the commencement of ECF DICI. The hos- on an annual basis. Unfortunately, the HbA1c
pital admissions were therefore summed for each thresholds changed during this time, so that only
month for areas/groups of practices. Each month the last 2 years are available with the same thresh-
was taken as a separate time point and the mean olds. Annual referrals for Dose Adjustment for
difference before and after each time point calcu- Normal Eating (DAFNE) were obtained from
lated to provide a real indication of the mean Cambridge University Hospitals Foundation
change in admissions before and after that month. Trust (A Housden, personal communication).
A Poisson regression model was constructed to Emergency Department attendance data for
analyse the percentage change in average monthly hypoglycaemia were not considered adequate
hospital admission rate before and after each due to coding issues. Pharmaceutical data were
monthly cut-off point, e.g., after 1 year, 12 mod- not available. All analyses were conducted in
els were built corresponding to the 12 monthly STATA [STATA/SE 11.0. StataCorp, Texas] and
cut-off points from July 2009 to June 2010. To tests are two tailed with p < 0.05 taken as signifi-
allow for residual seasonality, we incorporated cant. Ethics approval was received from the
terms to describe an annual sinusoidal pattern in National Research Ethics Service Committee-
the numbers of hospital admission. The model- East of England.
ling of potential confounders was completed by
the inclusion of indicator variables for the month. Year 1 Evaluation
A variety of sensitivity analyses were conducted Those adopting the intervention fully had reduced
[11, 12] which showed no change to the results. hospitalisation and hospitalisation costs com-
For the 1 year analysis, practices were defined pared with those defined as less engaged [11].
as either engaged or less engaged based upon Comparable changes in the rate of hospitalisation
the degree of uptake of relevant ECF DICI com- increase were not seen among those without dia-
ponents (referrals, virtual clinics, DSN/DSD/ betes or among those in other parts of
DSP clinics and/or barriers assessments). The Cambridgeshire.
Year 3: Summative Evaluation Table 7.2 Degree of implementation of the mechanisms

for vertical integration/clinical governance in East Cambs
and Fenland (ECF)
Degree of Implementation
The additional staffing, clinics, barriers frame- Governance
component Degree of implementation
work [11, 12, 15], and general practice education
ECFa Diabetes ECF meetings were held
were implemented. Referrals to DAFNE Network approximately quarterly in
increased from 10 to 14 per annum (20032009), year 1. These reduced after the
to 25 in 2010 and 16 in 2011. Consultant virtual introduction of the countywide
clinics occurred in 16 general practices: 23 in network. Governance stopped
after 1218 months
year 1 (when the LES was in place), 14 in year 2
CPCTb integrated Focused on defining key
and 9 in year 3. Conversely, the degree of imple- diabetes care performance indicators, rather
mentation of the vertical integration/governance initiative project than project managing and
arrangements was limited (Table 7.2). A major group facilitating integration
joint effort occurred through 2010 to July 2011, Secondary provider Met infrequently
reference group
including diabetes services, primary care and
ECF patient Never eventuated
patients across the county, built upon the work of reference group
the ECF DICI, to create a proposal for an inte- Countywide network Did not focus on ECF
grated service across Cambridgeshire. A variety Cambridgeshire Collaborated well on
of options for further investment were proposed diabetes drugs countywide drug issues
to CPCT under their Sustainable Health advisory group
Partnership programme, but these were all Countywide Recommendations allowing
rejected. One group of GPs in south information data sharing between general
management practice and other health
Cambridgeshire/Cambridge City (CamHealth) working group services (hospitals and
opted to adopt the ECF approach and invested in community services) were
the additional staff. The rejection of the county- never implemented
wide integrated care proposal, led to uncertainty Countywide The recommendations were
over service commissioning and an increase in guidelines and never monitored
pathways working
competition between the different services. group
Identification of clinical governance issues at this Countywide diabetes The recommendations were
time of increased uncertainty, led to the with- education working never implemented
drawal of Addenbrookes Hospital from the DICI. group
ECF team clinical Clinical incidents were rarely
governance tabled-few were placed in
Impact of the DICI on Hospitalisation,
front of the DC
Metabolic Control and Outpatient Attendance a
East Cambridgeshire and Fenland
by Area b
Cambridgeshire Primary Care Trust
Figure 7.4 shows the weekly hospital episode
rate, and Table 7.3 the monthly hospital episode
rates over the 5 years for those with and without for elective and non-elective hospital admissions
diabetes by area. The proportion of admissions and for those with diabetes as a primary ICD
by those with diabetes increased in all areas over diagnosis code: no reduction in hospitalisation
the 5 years, but the increase was 9.2 (5.513.1)% was seen in ECF. Figure 7.5 shows the monthly
and 54.1 (49.059.4)% greater in ECF than Hunts hospitalisation change overall and for non-
and Greater Cambridge respectively including a elective admissions by area for those with and
7.4 (5.29.2)% and 45.5 (42.548.5)% greater without diabetes: in Huntingdonshire and Greater
increase in ECF over the 3 years of the Cambridge, but not in ECF, the monthly non-
intervention. elective hospitalisation change became similar
Table 7.3 also shows the mean monthly change between patients with and without diabetes. In
rates in hospitalisation across the 5 years, overall, ECF, the difference in monthly change in hospi-
Fig. 7.4 Hospitalisation (number/week) among those with and without diabetes over the 5 years by area. Each figure
is weekly hospital admission series plot with LOWESS fit line
talisation between those with and without diabe- Year 3: Process Evaluation
tes appears to widen after the introduction of the The process evaluation was conducted using the
DICI. The difference between the areas was most Medical Research Council guidelines for evaluat-
marked in those aged 75+ years (Fig. 7.6). The ing complex interventions [20].
DICI was not associated with a reduction in hos-
pitalisation rates for diabetes, diabetic foot, dia- Practice Nurse Telephone Survey
betic ketoacidosis, hypoglycaemia or A telephone survey was undertaken with ECF
cardiovascular disease in people with diabetes practice nurses (14/16 practices one PN was
(Table 7.3). There was no greater achievement of excluded as she became a member of the DICI
diabetes QoF targets in ECF over the 5 years team) including use of a barriers framework [15],
either overall or by individual QoF score. There confidence with insulin management, referral to
was an increase in new outpatient appointments diabetes specialist services, access to specialist
in Huntingdonshire in 2011. podiatry, and how closely they felt they were now
There was no difference between hospitals in working with the community diabetes team/local
hospitalisation rates within ECF over the hospital diabetes services. A reduction in integra-
36 months of the DICI (Fig. 7.7), besides an tion between primary care and hospitals was
improvement in the Huntingdonshire hospital reported by 6/14 PNs across the different hospital
catchment area. Practices associated with DSN catchments with a parallel increase in integration
A, but not DSNs B and C, were associated with with community services reported by 8/14 PNs.
an increase in hospital admission rates over time All PNs reported referring to the community
concurrent with consistent reductions in outpa- DSN. Most PNs reported an increase in their con-
tient referrals (Fig. 7.8). These changes com- fidence in the use of insulin (9/14, although one
menced before the introduction of the DICI under indicated a reduction in confidence) and 7/14
the DSN led team approach. The DSNs covered reported that they referred foot ulcers earlier to
practices across ECF. the community DSP.
Table 7.3 Hospitalisation referral by region
2007 2008 2009 2010 2011
Greater Greater Greater Greater Greater
ECF Cambridge Hunts ECF Cambridge Hunts ECF Cambridge Hunts ECF Cambridge Hunts ECF Cambridge Hunts
Population size Registered 7672 6532 5077 8514 6859 5490 9052 7276 6041 9497 7658 6484 10,051 7977 6855
diabetes
patients
Practice 174,411 244,341 154,947 187,740 245,614 155,845 190,059 249,681 157,537 192,047 253,143 158,342 194,639 254,833 158,933
size
Annual diabetes 2909 2270 2134 4391 3497 3124 4801 4118 3662 5086 4378 3762 5589 4186 4091
hospitalisation (9.07 %) (6.25 %) (8.54 %) (9.74 %) (6.79 %) (8.71 %) (10.09 %) (7.52 %) (9.90 %) (10.23) (7.65 %) (9.82 %) (11.26 %)a (7.50 %)b (10.28 %)b
percentagea
Monthly Diabetes 327 249 (238, 242 361 284 (266, 262 405 (375, 346 (330, 311 422 367 (336, 319 466 (433, 342 (331, 343 (315,
hospitalisation (303, 267) (220, (348, 316) (238, 427) 363) (282, (410, 395) (300, 491) 360) 359)
rate 339) 251) 391) 276) 321) 447) 326)
Non- 3286 3756 (3611, 2554 3393 3999 2702 3597 4287 2782 3670 4413 2877 3650 4265 2996
diabetes (3050, 3965) (2395, (3278, (3890, (2685, (3446, (4046, (2702, (3573, (4254, (2788, (3559, (4071, (2859,
3392) 2704) 3495) 4071) 2764) 3778) 4360) 2864) 3860) 4477) 3007) 3755) 4524) 3057)
Monthly change Diabetes Ref Ref Ref 14.3 15.7 (10.3, 10.1 23.2 32.0 (26.8, 26.0 29.0 38.1 (33.0, 28.7 38.4 (33.9, 33.6 (28.5,
37.1
percentage (9.6, 21.0) % (4.5, (18.6, 37.2) % (20.6, (24.4, 43.3) % (23.3, 43.0) % 38.8) % (31.8,
-overall 19.1) % 15.7) % 27.9) % 31.4) % 33.6) % 34.1) % 42.4) %
Non- Ref Ref Ref 4.5 (3.0, 5.8 (4.4, 7.6 (5.9, 9.6 (8.1, 11.2 (9.8, 9.4 (7.7, 13.9 15.4 (14.1, 13.0 12.5 (11.0, 13.2 (11.8, 16.3
diabetes 6.1) % 7.3) % 9.3) % 11.1)% 12.6) % 11.1) % (12.4, 16.8) % (11.3, 14.0) % 14.5) % (14.6,
15.4) % 14.7) % 18.0) %
Monthly change Diabetes Ref Ref Ref 13.3 11.1 (3.6, 12.4 19.2 24.0 (16.7, 12.7 25.4 26.9 (19.6, 19.0 28.8 (22.0, 23.4 (16.1, 17.1 (8.9,
percentage (6.3, 18.5) % (7.3, (12.3, 31.2) % (4.3, (18.6, 34.1) % (10.7, 35.6) % 30.6) % 25.4) %
-Non-elective 20.3) % 25.3) % 26.1)% 21.0) % 32.2) % 27.2) %
Non- Ref Ref Ref 11.6 1.0 (1.2, 9.9 (7.2, 12.3 (1.2, 5.0 (2.8, 13.3 21.5 9.9 (7.7, 15.8 32.6 (21.8, 8.4 (6.3, 14.1
diabetes (0.04, 3.2) % 12.6) % 23.5) % 7.2) % (10.7, (10.6, 12.0) % (13.2, 43.3) % 10.6) % (11.4,
22.7) % 16.0) % 32.5) % 18.5) % 16.8) %
Monthly change Diabetes Ref Ref Ref 6.6 9.1 (3.8, 2.3 9.4 (2.4, 2.2 (10.8, 24.4 8.5 4.9 (8.1, 19.9 17.6 (5.9, 2.1 30.9
percentage (5.3, 22.0) % (12.3, 21.3) % 15.3) % (10.4, (3.3, 17.9) % (5.7, 29.2) % (15.2, (17.0,
-elective 18.5) % 17.0) % 38.5) % 20.4) % 34.0) % 11.1) % 44.7) %
Non- Ref Ref Ref 1.2 0.2 (4.0, 2.8 0.5 (3.5, 0.9 (4.7, 6.8 3.6 3.1 (6.9, 3.1 1.6 (5.6, 13.5 7.3
diabetes (2.8, 3.6) % (7.2, 4.5) % 2.9) % (11.1, (0.4, 0.7) % (7.5, 2.4) % (17.4, (11.8,
5.2) % 1.6) % 2.3) % 4.5) % 1.3) % 9.6) % 2.9) %
Monthly change Diabetes Ref Ref Ref 2.9 16.9 (3.0, 19.3 20.9 (4.1, 34.6 (15.3, 19.3 14.1 31.7 (12.3, 26.7 29.3 (12.8, 5.2 (15.2, 51.4
percentage (20.5, 36.8) % (2.2, 37.8) % 54.0) % (2.2, (3.0, 51.2) % (5.5, 45.9) % 25.6) % (31.0,
-diabetes as 14.7) % 40.8) % 40.8) % 31.1) % 47.9) % 71.7) %
primary code
Rates are shown as P50 (P25, P75)
Change percentage is estimated by Poisson regression model by using 2007 monthly count as reference, and is present as estimated percentage (95 %CI)
Data coverage: 9 months in 2007 (AprilDecember); 12 months in each year of 20082011 (JanuaryDecember)
a
Annual diabetic hospitalisation percentage = annual diabetic hospitalisation count/annual total hospitalisation count, %
b
Means the results of trend analysis across 5 years were statistically significant (p < 0.0001)
Fig. 7.5 Monthly percentage change in hospital admis- change percentage for those with diabetes. Grey (lower)
sion rate (monthly hospital admission rate before and after plot represents for hospital admission change percentage
the month) across areas among all patients and non- for those without diabetes. Figures above 0 % indicate that
elective admissions. Left plots are for all-caused hospitali- the mean monthly hospitalisation rate after the index
sation; right plots are for non-elective hospitalisation. month is greater than the mean hospitalisation rate before
Black (upper) plot represents for hospital admission the index month
Degree of Integration: Multidimensional working with community and CCS Nurse

Analysis (MDS) Referral indicate increased integration between
Multidimensional Scaling (MDS) used the prac- the community services and general practice,
tice nurse data to create a proximities matrix associated with increased self-reported confi-
(similarities or dissimilarities) between health- dence in managing diabetes. Conversely, hospital
care providers (GPs, community services and referral and working with hospitals are rated
hospitals) [21]. The MDS algorithm minimises quite low, reflecting decreased integrated work-
the standardised residual sum of squares (the ing between general practice and hospitals. This
stress factor): the closer to zero, the better the is in line with our observations during the
representation: a value <0.1 is excellent [22]. project.
Figure 7.9 shows the mapping of health-care pro- Figure 7.10 shows the mapping of the provid-
viders according to their dissimilarities: the stress ers according to their dissimilarities; they form
factor is 0.18, which is acceptable. six clusters of perceived changes and the corre-
The distances between the indicators show sponding rating. The first cluster is best repre-
that they are quite dissimilar. The proximity of sented by provider (EC5) and shows
Fig. 7.6 Monthly percentage change in hospital admis- for those with diabetes. Grey (lower) plot represents for
sion rate (monthly hospital admission rate before and after hospital admission change percentage for those without
the month) across areas among all patients aged <75 years diabetes. Figures above 0 % indicate that the mean
and 75+ years. Left plots are for those aged 75+ years; monthly hospitalisation rate after the index month is
right plots are for those aged <75 years. Black (upper) greater than the mean hospitalisation rate before the index
plot represents for hospital admission change percentage month
improvements in all categories, indicating referral worsened, but all other categories
increased integration. In the second cluster pro- showed improvements. In the fifth cluster CCS
vider (F5) is typical for the cluster and shows Nurse Referral and working with community
only slight improvements, i.e., in the category are rated positively with no changes in own con-
CCS Nurse Referral. This means that integra- fidence. This cluster is best represented by pro-
tion was quite low. In the third cluster, best repre- vider (H3), integration was also very low. For the
sented by provider (F2), CCS Nurse Referral, sixth cluster nothing changed, except a worsening
working with hospital and own confidence in the category CCS Nurse Referral. This indi-
were improved, but hospital referral was evalu- cates no improvement in integration.
ated negatively. This also indicates low integra-
tion, but higher when compared to the fourth Observations and Interviews
cluster. Provider (EC5) is typical for the fourth Ethnographic field notes were collated from
cluster; working with hospital and hospital observations by social scientists. Semi-structured
Fig. 7.7 Monthly percentage change in hospital admis- hospital admission change percentage for those without
sion rate (monthly hospital admission rate before and after diabetes. Figures above 0 % indicate that the mean
the month) by hospital and diabetes status. Black (upper) monthly hospitalisation rate after the index month is
plot represents for hospital admission change percentage greater than the mean hospitalisation rate before the index
for those with diabetes. Grey (lower) plot represents for month
interviews, supplemented by more casual interac- (17/21) were interviewed on two occasions at
tions, were carried out with 21 patients (primary least 6 months apart to obtain insight into how
care alone in two practices (n = 7), DICI service care and perspectives changed. Interview data
(n = 7), hospital multidisciplinary clinic (n = 7)) and ethnographic notes were transcribed and
and 20 health-care professionals (ten DICI staff, entered into NVivo 9 for analytical coding using
six hospital staff (three podiatrists, one consul- a grounded theory approach [23]. The datasets
tant, one DSN, one DSD) across two hospitals were cross-referenced for further refinement and
and four primary care staff (two practice nurse corroboration.
and two GPs) across four practices). Patients
Fig. 7.8 Monthly percentage change in hospital admis- change percentage for those without diabetes. Practice
sion rate and in outpatient referral rate (monthly hospital group A is associated with DSN A. Practice group B is
admission rate before and after the month) by practice associated with DSN B. Practice group C is associated
group. Left plots are for inpatient hospitalisation; right with DSN C. Figures above 0 % indicate that the mean
plots are for outpatient referral rate. Black (upper) plot monthly hospitalisation/attendance rate after the index
represents for hospital admission/outpatient referral month is greater than the mean hospitalisation/attendance
change percentage for those with diabetes. Grey (lower) rate before the index month
plot represents for hospital admission/outpatient referral
Detailed comments from the interviews are Integration

shown elsewhere [12], and indicated three Continuity of care: the Health-care profes-
themes: sional and patient relationship
The need for tailored care
Fig. 7.9 Mapping of the questions according to their dissimilarities based upon the practice nurse survey
Discussion Was the Time to Show Benefits Too

Short?
The expansion of the community diabetes team Lower hospitalisation rates have been seen in
was successfully implemented across ECF, similar interventions in other countries within
patient experiences were positive, PN clinical 12 months within single providers [24, 25], or in
confidence improved, and there were early contexts where multiple primary care providers
reports of clinical benefit [10, 11]. The approach work with a single specialist provider under an
was seen as sufficiently beneficial for one GP integrated insurance scheme [26]. Foot interven-
group in South Cambridgeshire, CamHealth, to tions have also been shown to have benefits on a
invest in a similar joint service. It is therefore sur- population basis within 23 years [27, 28].
prising that the DICI was associated with no Although changes in some adverse outcomes
improvement in QoF measures and no reduction, (e.g., cardiovascular disease, microvascular dis-
and indeed perhaps a worsening, in hospitalisa- ease) would take longer than 3 years, the absence
tion among those with diabetes over the 3 year of any positive impact on foot admissions in spite
period. In view of the size of the investment, the of a major increase in resource indicates that the
acceptability of the approach to patients, primary time should have been sufficient to show an
care and the diabetes specialist services and the improvement in at least this key indicator of
fulfilment of the goals of care closer to home success.
and primary care support, the question has to be
asked why no improvements were shown over Were Patients Worse at Baseline in ECF?
3 years? We wondered if the greater pre-existing morbid-
ity and lower socioeconomic status in ECF was a
major influence, particularly in those practices
Conguraon (Kruskal's stress (1) = 0.181)

1
[6]
H1
0.8
H4
[2]
F5
0.6
EC4 [3]
0.4 F1
0.2 F9
[5]
H3 F2
Dim2
0
-1.7 -1.2 -0.7 -0.2 0.3 0.8 1.3 1.8
F4
F6
-0.2 [1]
EC1
EC6 -0.4
EC3
[4]
F7
H2 -0.6 F8 EC4
EC5 F3
-0.8
-1
Dim1
Fig. 7.10 Mapping of the health-care providers accord- (F1-9), East Cambs practices (EC1-EC6), Trinity
ing to their dissimilarities and clusters based upon the (Fenland) did not participate in the survey and Sutton had
Practice nurse survey. The typical members of the clusters a new practice nurse at the time of the survey
are highlighted. Hospitals (H1-4), Fenland practices
associated with DSN A. However, those practices diabetes coding policies, as the changes occurred
were across ECF, and the trends were not shown independent of hospital.
with DSN B or C. In fact, neither the QoF data The collateral data indicating no significant
nor the hospitalisation data support that patients improvements in metabolic control in ECF also
with diabetes were worse off at baseline in support the evidence that although the DICI was
ECF. There was evidence that the trajectory for implemented, it had no major population impact.
those practices associated with DSN A was
downward. Was the Intervention Fully
Implemented and Maintained?
Were Benefits Obscured by Secular The clinical interventions were indeed imple-
Changes? mented and horizontal integration occurred for
One weakness of the study is that it was not a most of the 3 years. It is clear, however, that the
randomised controlled trial and therefore influ- vertical integration components were not imple-
enced by historical, secular and other changes. mented successfully. The approach appeared to
By comparing data between the three reduce integration between hospitals and primary
Cambridgeshire areas and between those with care, something that is clearly at the core of suc-
and without diabetes, we had hoped to minimise cessful initiatives.
confounding: all patients were under the same The failure to implement integrated informa-
commissioners (CPCT), with no differences in tion management for putative information gover-
policies overall. We excluded possible confound- nance reasons, almost certainly contributed to
ers such as changes in hospital admission and communication and integration difficulties. Most
integrated care initiatives attempt to include data in the integrated care plan. Conversely, the non-
sharing [29] and this was not possible within the medical diabetes specialist staff in the fourth hos-
local information governance arrangements. This pital (Hinchinbrooke in Huntingdonshire), were
was noticed by the patients and was a source of already employed by the community services.
frustration. During the time of the ECF DICI, Circle, a pri-
Several interventions were not maintained. vate provider took over the running of the hospi-
The multidisciplinary clinic in the north of the tal in Huntingdonshire (Hinchinbrooke) (http://
county (Wisbech) occurred only a few times, as a www.circlehealth.co.uk/about-circle/media/
result of a lack of space-although a clinic was circle- named-as-hinchingbrooke-preferred- -
attempted within one large general practice, this bidder-nov-10, accessed 26/6/2015) through a
had to be out of hours which was too difficult to competitive tender process. The Doddington
maintain. After the rejection of the countywide community clinic, run with Peterborough
approach, a more competitive atmosphere Hospital for over 20 years, was taken over by
appeared to emerge. For example, the reducing Hinchinbrooke Hospital with the community ser-
numbers attending DAFNE provided by vices, again highlighting the competitive nature
Addenbrookes Hospital in ECF and Cambridge, of the Cambridgeshire environment. Perhaps in a
coincided with the community services increas- less competitive, more collaborative environ-
ing referrals to the hospital service in ment, ECF DICI would have successfully
Huntingdonshire, run by the community services impacted on diabetes hospitalisation.
themselves. These were also associated with the
recognition of defects in the clinical governance Was There Something Intrinsic
approach. Such clinical changes probably About the Model: Piggy in the Middle?
reflected reductions in the degree of vertical inte- Hints as to why this three component model (GP,
gration as the new GP commissioning era began hospital, community services) might not be effec-
to be introduced. tive come from the observation that the preceding
nurse led service in ECF had reductions in refer-
Was There Something Intrinsic rals to hospital outpatients as one of its targets
About the Model-Lack of Vertical [2], and changes in outpatient appointments were
Integration? included in the DICI variation in contract. Similar
We feel that the low degree of vertical integration emphasis on reducing hospital outpatient refer-
provides a clue as to the reason for the failure of rals (ergo payments) was included in, e.g.,
the ECF DICI to impact on hospitalisation over Birmingham [30]. Within this context, it was per-
the 3 years. In fact, the vertical integration was haps to be expected that attempts at creating
never fully implemented, something the pro- greater integration in information management,
posed countywide diabetes proposal was intended clinical governance, budget and overall manage-
to address. The proposed plan included a pooling ment were agreed but not implemented: actions
of the diabetes service funds and a fully inte- more achievable within a single organisation.
grated countywide clinical/corporate governance Interestingly, integration was perceived as
framework-more akin to a joint venture model. happening when there was one person fronting
The ECF DICI generated a closer relationship up for all those involved [12]. Case management
between the community services and general has been proposed as one approach to integra-
practice, and initially with Addenbrookes tion, and requires the case manager to corral and
Hospital. However, the relationships between the coordinate the services for a given individual
ECF DICI and 3 of the other neighbouring hospi- [31].
tals remained limited and competitive. While the Clinical inertia, a less aggressive approach
community Wisbech clinic was not maintained, to management under non MDT management,
the Wisbech clinic that was an outreach service greater PN confidence and financial disincentives
from the neighbouring hospital was not included to referral to the hospital clinics, could all have
delayed onward referral to the broader associated not yet addressed the need for a competition, and
multidisciplinary team management. We wonder in Derby, a perceived to be successful inte-
if the diabetes foot intervention reflects this most grated (joint venture) care programme was not
clearly. The DSP made major, and successful, recommissioned [33].
efforts to link in with the four neighbouring hos-
pital diabetes foot clinics. However, the inter-
views suggest that this was associated with Conclusion
delayed referral for MDT foot care, i.e., the DICI
service promoted hospital avoidance rather than In conclusion, these data show no improvement
admission avoidance [12]. Many PNs reported in outcomes with investment in a separate com-
that their working with the hospital services was munity diabetes specialist service, and increased
reduced, suggesting that the placement of the fragmentation between primary and secondary
DSNs between the practice and the hospital ser- care. While horizontal integration occurred, ver-
vice created a barrier not a bridge to optimal care. tical integration was negligible.
In fact stepping back, a possible explanation for Studies to demonstrate improvements in dia-
the lack of effectiveness on a population basis, betes outcomes with alternative models of care
may be that the community service were popular are urgently needed.
with general practices by reducing their diabetes
workload (perhaps resulting in similar outcomes), Acknowledgements Elissa Harwood collected some eth-
while obstructing the on-referral to often more nographic data and Chris Bunn/Simon Cohn undertook
the analyses reported elsewhere [12, 15]. Tosin Daniels
skilled, multidisciplinary, one stop shop hospital
undertook the practice nurse interviews. We thank the
services (perhaps resulting in worse outcomes for patients and staff for their interviews, Michael Thoresby
the non-referred patients). This was further for extracting the PCT data, Liz Robin for her useful com-
undermined by the inability of the local leader- ments on the analysis and paper drafts. We thank The ECF
Diabetes Integrated Care Team and local practices for
ship to translate into reality, countywide agree-
their hard work, the Addenbrookes Hospital Diabetes
ments, across all health professionals and patient Care Patient Advisory Committee for guidance.
representatives, on education, information man- This paper presents independent research funded by
agement, clinical governance and a vertically the National Institute for Health Research (NIHR) under
its Research for Patient Benefit (RfPB) Programme (Grant
integrated approach.
Reference Number PB-PG-0808-17303). The views
expressed are those of the authors and not necessarily
Is the ECF Experience Likely those of the NHS, the NIHR or the Department of Health.
to Be Typical?
Whether our findings are due to a unique set of
circumstances or a predictable consequence of a References
three compartment model in a competitive local
health system is unclear, but there are indications 1. Department of Health/Royal College of General
Practitioners. Guidelines for the appointment of gen-
that the circumstances are not special. There are
eral practitioners with special interests in the delivery
calls for more integration and less fragmentation of clinical services: diabetes. London: DOH; 2003.
in health-care [30], yet so far, the major integra- 2. Hollern H, Lunn J. Development of a primary care
tion projects in England have revealed negligible, specialist diabetes nursing service. J Diabetes Nurs.
2004;8:3448.
if any, benefits [25, 32]. The latest changes in
3. Hicks D, James J, Hill J, Vanterpool G. Audit of the
commissioning in the English NHS, with empha- effectiveness of nurse consultant led intermediate dia-
sis on the need to consider Any qualified betes care services in England. Pract Diabetes.
Provider in service delivery, and associated mar- 2012;29:547.
4. Armstrong N, Baines D, Baker R, Crossman R,
ket procurement approaches, could well impair
Davies M, Hardy A, Khunti K, Kumar S, OHare JP,
the quality of diabetes care while increasing Raymond N, Saravanan P, Stallard N, Szczepura A,
overall cost, if the experience here is reproduced Wilson A, ICCD Study Group. A cluster randomized
elsewhere. Calls for more integrated care have controlled trial of the effectiveness and cost-
effectiveness of intermediate care clinics for diabetes improvement of the ABCs of diabetes. Patient Educ
(ICCD): study protocol for a randomized controlled Couns. 2011;82:26670.
trial. Trials. 2012;13:164. 18. Simmons D, Bourke L, Yau E, Hoodless M. Diabetes
5. Simmons D, Wenzel H. Diabetes inpatients: a case of risk factors, diabetes and diabetes care in a rural
lose, lose, lose. Is it time to use a Diabetes Australian community. Aust J Rural Health.
Attributable Hospitalisation Cost to assess the 2007;15:296303.
impact of diabetes? Diabet Med. 2011;28:11238. 19. Langlands A. The good governance standard for
6. Lafond S, Charlesworth A, Roberts A. Hospital public services. London: OPM/CIPFA; 2004.
finances and productivity: in a critical condition? 20. Anderson R. New MRC, guidance on evaluating com-
London: Health Foundation; 2015. plex interventions. BMJ. 2008;337:a1937.
7. The Action to Control Cardiovascular Risk in 21. Rencher AC. Methods of multivariate analysis.
Diabetes Study Group. Effects of intensive glucose New York: J. Wiley; 2002.
lowering in type 2 diabetes. N Engl J Med. 22. Backhaus K, Erichson B, Plinke W, Weiber
2008;358:254559. R. Multivariate Analysemethoden. Eine anwendung-
8. Calvert M, Shankar A, McManus RJ, Lester H, sorientierte Einfhrung. Berlin: Springer; 1996.
Freemantle N. Effect of the quality and outcomes 23. Pope C, Ziebland S, Mays N. Analysing qualitative
framework on diabetes care in the United Kingdom: data. BMJ. 2000;320:1146.
retrospective cohort study. BMJ. 2009;338:b1870. 24. Jha A, Perlin J, Kizer K, Dudley R. Effect of the trans-
9. Rayman G, Kilvert A. The crisis in diabetes care in formation of the veterans health care system on the
England. BMJ. 2012;345:e5446. quality of care. N Engl J Med. 2003;348:221827.
10. Hollern H, Simmons D. Cost saving and improved 25. Ham C, Curry N. Integrated care: what is it? Does it
glycaemic control in an integrated diabetes service. work? What does it mean for the NHS? London:
Prim Care Diabetes. 2011;13:17681. Kings Fund; 2011.
11. Simmons D, Yu D, Wenzel H. Changes in hospital 26. Distiller LA, Brown MA, Joffe BI, Kramer
admissions and inpatient tariff associated with a dia- BD. Striving for the impossible dream: a community
betes integrated care initiative: preliminary findings. based multi-practice collaborative model of diabetes
J Diabetes. 2014;6:819. management. Diab Med. 2010;27:197202.
12. Simmons D, Yu D, Bunn C, Cohn S, Wenzel H, 27. Lavery LA, Wunderlich RP, Tredwell JL. Disease
Prevost T. Hospitalization among patients with diabe- management for the diabetic foot: effectiveness of a
tes associated with a diabetes integrated care initia- diabetic foot prevention program to reduce amputa-
tive: a mixed methods case study. Future Hosp tions and hospitalizations. Diabetes Res Clin Pract.
J. 2015;2(2):928. 2005;70:317.
13. Elliott J, Jacques RM, Kruger J, Campbell MJ, Amiel 28. Rith-Najarian S, Branchaud C, Beaulieu O, Gohdes
SA, Mansell P, Speight J, Brennan A, Heller D, Simonson G, Mazze R. Reducing lower-extremity
SR. Substantial reductions in the number of diabetic amputations due to diabetes. Application of the staged
ketoacidosis and severe hypoglycaemia episodes diabetes management approach in a primary care set-
requiring emergency treatment lead to reduced costs ting. J Fam Pract. 1998;47(2):12732 [serial online].
after structured education in adults with type 1 diabe- 29. Ham C, Walsh N. Making integrated care happen at
tes. Diabet Med. 2014;31:84753. scale and pace. London: Kings Fund; 2013.
14. Simmons D, English P, Robins P, Craig A, Addicott 30. Hill J. Seamless diabetes care: the role of the nurse
R. Should diabetes be commissioned through multi- consultant. J Diabetes Nurs. 2008;12:14954.
disciplinary networks, rather than practice based com- 31. Norris SL, Nichols PJ, Caspersen CJ, Glasgow RE,
missioning? Prim Care Diabetes. 2011;5:3944. Engelgau MM, Jack Jr L, Isham G, Snyder SR,
15. Harwood E, Bunn C, Caton C, Simmons D. Addressing Carande-Kulis VG, Garfield S, Briss P, McCulloch
barriers to diabetes care and self-care in general prac- D. The effectiveness of disease and case management
tice: a new framework for practice nurses. J Diabetes for people with diabetes: a systematic review. Am
Nurs. 2013;17:18691. J Prev Med. 2002;22(4S):1538.
16. Simmons D, Weblemoe T, Voyle J, Prichard A, 32. RAND Europe, Ernst & Young LLP. National evalua-
Leakehe L, Gatland B. Personal barriers to diabetes tion of the department of healths integrated care
care: lessons from a multiethnic community in New pilots: final report:full version. London: RAND
Zealand. Diabet Med. 1998;15:95864. Europe, Ernst & Young LLP; 2012.
17. Piatt GA, Songer TJ, Brooks MM, Anderson RM, 33. Hussain S, Dornhorst A. Integrated care taking spe-
Simmons D, Orchard TJ, Siminerio LM, Korytkowski cialist medical care beyond the hospital walls.
MT, Zgibor JC. Impact of patient level factors on the London: Royal College of Physicians; 2015.
UK Approaches to Integrated
Diabetes Care: DerbyA Joint 8
Venture Model Under the NHS
Paromita King
Why Change? having to tell their story repeatedly. The frustra-

tions are exemplied by Mrs, Smith (Case) and
Derby City has a population of 250,000 with her experience summarised in Fig. 8.1.
18,000 people with diabetes. It has areas of social
deprivation with ethnic minority populations
who have a high prevalence of diabetes with
prevalence rates of 10 % compared with 6.2 % in
the UK in 2014 [1, 2]. Before 2008, Derby had a Case Mrs Smith, age 46, has insulin
traditional acute trust centred model for diabetes treated type 2 diabetes and is overweight
care, with pockets of specialist services in the with poor glucose control. Her GP has
community. sent her to the hospital diabetes team.
Acute trust clinicians felt that there were large Each time she attends she sees a different
numbers being seen for routine review, with fre- doctor who tells her to lose weight. The
quent duplication of appointments compromising last doctor was uncertain how to help her
the capacity of the service. However, variation in as she has forgotten to bring her list of
provision and expertise in primary care limited medication with her but refers her to a
the ability to discharge stable patients to increase nurse for help with insulin management
capacity. and a dietitian to lose weight. In the last 2
Primary care clinicians were overseeing the months she has had four appointments to
care of an increasing number of people with Type see the consultant, nurse and dietitian as
2 diabetes, often with complex co-morbidities. well as attended for retinal screening and
Difculties in communication with the secondary a foot check. She has to take two buses to
care team was not only frustrating, but patients attend hospital and has found that her glu-
struggled to access care in a timely manner cose control has not improved and that
resulting in preventable admissions. she has gained rather than lost weight. At
People with diabetes stated that they valued her last appointment the nurse referred
the expertise of the specialist team; but we pro- her back to the consultant who said he
vided poor continuity of care resulting in them would write to the GP and sends her back
to the nurse. Her GP feels unable to help
as he has not received any communication
from the hospital as the secretary is off
P. King (*) sick. Frustrated at the lack of progress
Department of Diabetes and Endocrinology, Medical and her time consuming journeys she dis-
Specialties, Derby Teaching Hospitals NHS engages and does not attend.
Foundation Trust, Royal Derby Hospital, Derby, UK

132 P. King
Fig. 8.1 Traditional

model-duplication of specialist
appointments, inaccessible
service and poor
communication resulting in a
disengaged patient
All involved in the care of people with diabe- to the patients journey that was integrated at
tes and the users felt that the service was frag- organisational, nancial, technological (IT) as
mented, with variation of provision and expertise, well as clinical levels. The model was under-
and that there was a clear case to redesign the pinned by a single clinical governance structure
service to better meet our user needs. [5]. This involved a monthly multidisciplinary
team meeting of clinicians, managers and users
with the remit to continuously improve the stan-
Our Vision dard of clinical practice. This is the rst diabetes
model in the UK that has successfully integrated
Inspired by the 2008 Royal Colleges publication at all these levels, and these principles are included
Teams without Walls [3], our vision was to move in the ve pillars of integration or key enablers in
away from the silos of the traditional model, where national documents recommending best practice
each professional works diligently in relative isola- for commissioning [6] and clinical practice [7]:
tion delivering their part of the pathway, to design-
ing seamless pathways around the user rather than 1. A single IT system or integrated IT
location, where they were seen by the right person 2. Alignment of nances
at the right time. In addition, we wanted to reduce 3. Care Planning
variation and duplication, improve outcomes across 4. Clinical Engagement
the whole health community and improve commu- 5. Single Clinical Governance Structure.
nication as well as the user experience. Our aim
was to achieve this through an integrated approach.
While there are many denitions of integrated The Derby Model
care, our vision is succinctly summarised by The
Kings Fund/Nufeld Trust [4] as: Like evidenced based medicine, integration is a
An approach that seeks to improve the quality of term currently much in use, but not always with a
care for individual patients, service users and car- full understanding of what it means in practical
ers by ensuring that services are well co-ordinated terms, and so an appreciation of the different
around their needs. aspects of integration is important to gain an
understanding of the Derby model.
How Did We Approach the Problem?

Organisational Integration
Acute trust clinicians and managers, general prac-
titioners and the primary care trust worked col- It was felt that an organisation equally owned by
laboratively to develop a whole system approach primary and secondary care was fundamental to
8 UK Approaches to Integrated Diabetes Care: DerbyA Joint Venture Model Under the NHS 133
fostering the shared responsibility and clinical two senior managers and a consultant. These
ownership that is essential when working across shareholders comprised the Board of the com-
organisational boundaries. In addition, it would pany which had contractual, governance and
allow services that were needed to deliver seam- nancial responsibility of the InterCare diabetes
less pathways centred around the user to be service. The services provided by the two compa-
brought together under one umbrella. There are a nies were similar, but further description of the
number of legal frameworks available in the UK model relates specically to the InterCare diabe-
that could have potentially achieved these objec- tes service.
tives [8]. The options considered included Social
Enterprises and companies limited by guarantee.
A Social Enterprise is an organisation whose Financial Integration
members share the same core values and where
the business has primarily social objectives All diabetes care was delivered using a single
whose surpluses are principally reinvested for pooled budget with the exception of complex ser-
that purpose in the business and community [8]. vices requiring cross speciality working. These
In a company limited by guarantee, members exceptions included: inpatients, multidisciplinary
who run the company are bound by a guarantee foot care, antenatal (not preconception), joint
rather than shares which require them to pay renal clinics and transition.
for debts up to an agreed maximum amount. A single budget allowed us to deliver seamless
Ultimately, a joint venture company limited by pathways across organisational boundaries with-
shares was chosen as the best vehicle to obtain a out the competition imposed by Payment by
Specialist Provider Medical Services (SPMS) Results, the payment system in England by which
contract while retaining the exibility to diver- commissioners pay providers a nationally agreed
sify within and beyond diabetes services [9]. tariff for each patient seen depending on the type
Two joint venture organisations, First Diabetes and complexity of the case [10]. It also allowed
and InterCare Health, were commissioned by the clinical team to own problems as well as solu-
Derby City Primary Care Trust in 2009 to deliver tions and invest prots from efciencies into new
NHS services for the patients with diabetes in the areas of care. For example, InterCare funded the
two practice based commissioning groups that innovative community based integrated precon-
covered Derby City. These were both not for ception care model PROCEED [11] after the end
prot National Health Service Companies lim- of its funded pilot until its business case was
ited by shares. accepted by Southern Derbyshire Clinical
First Diabetes was established to support the Commissioning Group.
First Provider group of practices which com-
prised ve practices with 2500 people with dia-
betes. The company had only two shares; one Information Technology Integration
owned by the First Provider group and the other
by the acute trust, and was overseen by a Board In Derby City 85 % of practices use The Phoenix
comprising two Directors one from First Provider Project (TPP), SystmOne, which was used by
group and one from the acute trust. InterCare and First Diabetes to provide a single
The second practice based commissioning shared record. As part of the referral process to
group comprised 29 GP practices with 15,000 specialist services, the patient consented to their
people with diabetes and was supported by record being shared. Access to the record at the
InterCare Health Limited. InterCare was briey point of referral ensured that they saw the most
piloted in seven practices before rolling out to the appropriate clinician to meet their needs avoiding
29. InterCare was also jointly owned by primary duplication as well as being able to provide holis-
and secondary care, but there were six shares tic and safe care. The electronic communication
held by three GPs and three by the acute trust, system using Tasks (similar to messaging) enabled
134 P. King
instant communication between clinicians and invaluable for building relationships and partner-
organisations. For example, a GP and nurse spe- ship working.
cialist were able to view each others consulta-
tions and clinicians were able to ask for advice or Education for People with Diabetes
agree a plan of care. In addition, the specialist Fragmentation of commissioning of educational
team could make recommendations to changes in initiatives meant delivering equitable education
the treatment which could be implemented with- for type 2 diabetes was challenging. Through
out waiting for letters or the need for the user to subcontracting arrangements, the different edu-
make an additional appointment with their GP. cational resources were brought together to pro-
vide equitable access to structured education for
Type 1 and 2 diabetes at diagnosis and other
Clinical Integration points of the users journey. These included
DAFNE (Dose Adjustment for Normal Eating)
Core Care for Type 1 diabetes, and a range of locally devel-
For the majority who have Type 2 diabetes, core oped initiatives to address the needs of those with
care was undertaken in the primary care setting, prediabetes, newly diagnosed Type 2 diabetes,
with care seamlessly escalated to and de- people on maximum oral hypoglycaemic agents
escalated from the specialist team across organ- and those on a basal bolus insulin regimen.
isational boundaries as needed. The terms
escalation and de-escalation reect the fact Specialist Care
that the whole team of clinicians were responsi- When targets were not being met or where there
ble for the care of people with diabetes either was a need for specialist team support, care was
directly or indirectly through raising standards of escalated to the multidisciplinary team where the
care through the support and training outlined user could be seen in one of the four community
below. In addition, the terms reects the lack of based clinics of their choice. As stated above, a
nancial (payment by results) or technological single record meant the user saw the right clini-
(multiple patient records) thresholds that have to cians to meet their needs or access the most
be crossed, as might be understood by the tradi- appropriate educational initiative. They were also
tional terms refer and discharge. able to have other aspects of their care such as
retinal screening at the same time as seeing the
Supporting Primary Care diabetes clinicians. To promote shared decision-
Practice support was designed to improve stan- making, mutually agreed targets such as HbA1c
dards of care and reduced variation. A nurse spe- or weight were set in consultations and recorded
cialist and consultant provided satellite support on templates that were visible to the primary care
to the practices tailored in accordance to their team. When targets were achieved, care could be
needs. This ranged from supporting clinics in de-escalated back to the primary care team. The
surgeries to case based discussion. In addition, user was given a contact number for the service
educational courses were provided for all prac- or clinician so that if support or advice was
tices, which included sessions on improving the needed in the future, they were able to directly
psychological literacy of consultations through access the specialist team and retain continuity of
the promotion of motivational interviewing tech- care. The de-escalation allowed the service to
niques, the promotion of shared decision-making maintain sufcient capacity to see users urgently
and care planning as well as sessions on compli- if needed but did not leave patients and col-
cations of diabetes and glucose management. leagues feeling isolated and unsupported.
Visits to practices were mutually benecial as it Services requiring horizontal or cross special-
also allowed members of the specialist team to ity integration in the acute trust were outside the
gain an understanding of some of the pressures of integrated diabetes contract, and nancially sup-
working in primary care and the visits were also ported by Payment by Results rather than the
single integrated care budget, and included tran- tionally hard to reach. For example, by working
sition, multidisciplinary foot care, antenatal and with the community matrons (advanced nurse
renal services. Clinical pathways in integrated practitioners with clinical and a case management
care were extended beyond these contractual lim- role coordinating care in the community for those
its so that the user journey could be continued with complex needs [12]), who also use
seamlessly in and out of secondary care services. SystmOne, housebound users could be supported
For example when a woman attending the inte- at a distance.
grated care preconception service became preg- Derby has a large South Asian Community
nant, her care would be transferred to the and we have worked across organisational bound-
antenatal clinic. While this service was outside aries to bring services together to meet their
integrated care, the same clinicians worked in needs through a Community Engagement Project.
integrated care and the antenatal clinic, maintain- For example, bilingual educators and interpreters
ing continuity of care for the user. Postnatally, supported the delivery of structured education,
she would be transferred back to integrated care and public health colleagues attended these ses-
and if appropriate her care could be de-escalated sions to promote their lifestyle initiatives, in par-
to Primary care. ticular local walking groups. The dietitian leading
In addition to these core principles, there were the project established links in religious and
a number of initiatives to meet the needs of spe- community venues to deliver healthy eating and
cic groups. cooking sessions in Hindu temples, gurdwaras
and Pakistani community centres. In addition,
Type 1 Diabetes she supported the local community pharmacists
Users with Type 1 diabetes described the impor- who were undertaking diabetes awareness events
tance of seeing clinicians who understood the where a risk calculator was used to discuss indi-
principles of intensive glucose management such vidual risk. The dietitian and pharmacist could
as carbohydrate counting and basal rate testing then discuss lifestyle changes at an individual
for those on pumps. Most felt that the primary level.
care team did not have these skills and that it was
important to maintain contact with the specialist
team, not so much for routine review, but to sup- Clinical Governance
port them at times of crisis. As the skills of the
primary care team improved, many were com- All these measures were supported by a single
fortable to attend primary care for their annual clinical governance structure. A multidisciplinary
review but to have a telephone consultation from clinical board of managers, administrators, pri-
the specialist team to support any care planning mary and secondary care clinicians and users met
issues that arose from their review. With the tech- monthly to address issues of clinical governance
nological advances in the management of Type 1 such as:
diabetes, these users were offered a Consultant
appointment every 23 years as an opportunity to Staff competencies and mandatory training
review their medication, and discuss new tech- (supported by annual appraisal).
nologies such as advances in glucose meters and Plan Do Study Act cycles, case reviews and
pumps. When a preference was expressed for tra- audits to drive efciencies to continuously
ditional care, the annual reviews were undertaken improve quality and look for new opportuni-
by the specialist team. ties for service development.
Discuss safety issues.
Hard to Reach Groups Review pathways.
Practising in community based settings gave Review agreed outcomes including user
clinicians the opportunity to develop other part- experience.
nerships to improve care for those who are tradi-
136 P. King
The chair of the clinical board reported monthly

to the InterCare Health Limited board of directors, The administrator contacts Mrs Smith
who carried ultimate responsibility for the gover- who chooses to attend at a community
nance of the service. based venue within walking distance of her
home. Her retinal screening is arranged on
the same day. When she attends, she sees
Summary the dietitian and the nurse specialist
together to ascertain her priorities and set
In summary, the ve pillars of integration have mutually agreed targets in terms of glucose
allowed people with diabetes in Derby City to control and weight loss and the practice is
access a model of care that has pathways centred sent a Task to prescribe the GLP-1 ana-
around them rather than location and is delivered logue. Further appointments with the nurse
by a team working together across organisational specialist are face to face and by telephone
boundaries committed to delivering the best care. where her progress and targets can be
reviewed. As she begins to lose weight, and
glucose control improves, she is empow-
Mrs. Smith in Integrated Care ered to adjust her own insulin and is able to
reduce the dose herself, thus reducing the
Returning to Mrs Smith, the case account below number of appointments needed. When she
and Fig. 8.2 describes a different experience in a achieves her targets, her care is de-escalated
system where diabetes care is integrated. to the practice, but she has the nurses
mobile number and telephone as well as
e-mail and telephone contacts for the ser-
vice so she can call or request an appoint-
Case The Practice is struggling with Mrs ment if she has problems in the future.
Smiths diabetes and has referred her to the
specialist team for help with glucose control
and weight. The practice has been trained
and the nurse is aware that GLP-1 analogues Service Evaluation
may be the best way forward for her, but she
has not yet had much experience in the use Outcome data for First Diabetes are published
of GLP-1 analogues with insulin. At the elsewhere [9]. InterCare Health was commis-
time of referral, Mrs Smith consents to shar- sioned for 3 years, so a comprehensive service
ing her record. The GP lead who is review- evaluation was presented to commissioners as we
ing referrals is able to see the reason for approached the end of the contract in 2012/2013,
referral and the whole GP record. It is clear although aspects of service evaluation were regu-
that she has been on insulin for a number of larly evaluated as part of Plan Do Study Act
years, but is struggling with weight and her cycles to continuously improve quality.
glucose control remains poor. It is also clear The Donabedian principles [13] of improve-
that there are no other medical issues and ments in structure, process and outcome were
that Mrs Smith is not at risk of pancreatitis, core to the service design. The changes in struc-
has normal liver and renal function. He can ture are described above and the service evalua-
see that the practice regularly undertake her tion concentrated on process and outcome
review but she is due her retinal screening. measures. In addition, indicators were chosen to
He recommends that Mrs Smith has a single reect all six domains of quality improvement as
appointment with the nurse specialist and described in the 2001 National Institute of
dietitian and then follow up with the nurse. Medicine/National Academy of Science publica-
tion Crossing the Quality Chasm [14]:
(continued)
Fig. 8.2 Integrated care with care centred around the user, with routine review at the practice but additional care at one
venue using a shared record and free of nancial restraints allowing ow of care across organisational boundaries
Effectiveness was assessed though evaluating was evaluated using a whole group of endpoints
changes in individual and service level glu- rather than individual items. The approach is
cose and blood pressure control, and also described in 10 High Impact Changes for
looking at the impact on admissions with a Service Improvement and Delivery Change 6,
primary diagnosis of diabetes. NHS Institute for Innovation and Improvement,
Efficiency and Timeliness was evaluated by 2004 [15]. This approach, particularly when used
auditing the referral and assessment system. in conjunction with Plan Do Study Act cycles,
Equity was improved by implementing the mea- has been shown to reduce variation and is appro-
sures to improve access and provision for hard to priate for supporting a care pathway crossing
reach groups, and evaluated through user feedback. organisational boundaries.
Safety was evaluated in plan do study act Data from the National Diabetes information
cycles, changes in glucose, blood pressure Service [9] were used to evaluate quality out-
control and ascertaining whether there was a come framework targets in Derby City.
reduction in admissions due to diabetes.
The provision of a Person Centred Service
was assessed through the implementation of Process
target setting reecting shared decision-
making as well as staff and user experience. Data were collected prospectively using
SystmOne and reviewed monthly.
Methodology
Activity
A Care Bundled approach was used to systemati-
cally appraise the quality of the service in 2012 Activity in 20122013 was in line with our plan
2013. This means that the entire care pathway and there was a progressive reduction in patients
138 P. King
attending acute trust clinics and an increase in turnaround was achieved in all but one of the
those attending integrated care. Initially 50 % of referrals. The next step was contacting the user to
all activity was undertaken in the acute trust, but arrange a convenient appointment and our admin-
an increasing number of users were transferred to istrative team consistently achieved this within
the new service and by the end of 2012, 70 % of 72 h. The last step was the time to the rst
people with diabetes were being seen in inte- appointment. In the rst audit, it appears that
grated care. only 84 % achieved this target, but when the rea-
In 2012, fewer patients failed to attend their sons behind this were explored, the majority of
appointments (DNA) in integrated care specialist cases the delay was due to user choice. When this
clinics compared with the acute trust with rates was considered, the 90 % target was met as 94 %
reducing from 20 % to 11 % for nurse specialists achieved the target.
and 1611 % for consultant sessions. In 2013 our These targets continued to be met. Although
overall DNA rate was reduced to 7 %. InterCare was not commissioned to deliver an
urgent service, unlike the traditional model it was
able to generate sufcient capacity to be able to
Referral and Assessment see urgent referrals and promptly respond to que-
ries, and so provide a safer as well as a more ef-
The referral and assessment process is unique cient service than the traditional model.
and was audited in two time periods during 2012.
The components of the process were broken These outcomes demonstrated the efficiency,
down and time frames allocated to each (Fig. safety and timeliness of the service.
8.3). The aim was to process 90 % of referrals in
the time period. The outcome of the audit is
shown in Fig. 8.4. Clinical Outcomes
Referrals were sent to the service electroni-
cally through SystmOne. The user consented to Target Setting
their record being released to the specialist team Target setting was introduced in 2012 to encour-
as part of the referral process. The GP lead triage shared decision-making and to work towards
aged these referrals, and as he had access to the care planning. Uptake was initially slow with
whole patient record was able to allocate the only two clinicians routinely using target setting
most appropriate clinician or clinicians for the in rst quarter of 2012. Shared decision-making
user to see or to give advice. The target of 48 h was discussed at the monthly team meetings and
Fig. 8.3 Referral and

assessment process (triage)
Fig. 8.4 Triage process audit
as clinicians started to understand its benets in 1. Glucose control as a Target:

particular how it could be used to support care In March 2012 an audit of 50 patients seen
planning and review progress uptake increased, in InterCare for suboptimal control demon-
and in the last quarter of 2012, target setting and strated a reduction in mean HbA1c from
review was evident in 95 % of consultations. The 11.1 % (98 mmol/mol) to 8.9 % (74 mmol/l)
majority of targets were to improve glucose con- in 6 months. In 2013 changes in HbA1c
trol. Other targets were: were evaluated in 50 people who had a target
set for glucose and their progress reviewed
Hypoglycaemia reduction after 6 months. Mean HbA1c at the time of
Improvement of hypoglycaemia awareness referral was 9.7 % (82 mmol/mol) compared
Weight reduction with a 10 % (86 mmol/mol) 12 months pre-
Better blood pressure control viously. This level of poor control was seen
Improved understanding of carbohydrate in patients attending acute trust as well as
counting primary care clinics. Six months after their
To undertake DAFNE (Dose Adjustment for referral to InterCare mean HbA1c reduced
Normal Eating, Type 1 diabetes structured to 7.95 % (63 mmol/mol); P < 0.01 (Fig.
education) 8.5).
Establishing the cause of proteinuria 2. Quality Outcome framework (QOF) targets in
patients attending InterCare Health specialist
team clinics:
The improvement in target setting reflects a more Changes in QOF were assessed by auditing
person centred approach to consultations. SystmOne looking at all patients who had
attended InterCare specialist team clinics.
Glucose and Blood Pressure Control Compared to 2009/2010, in 2011/2012 there
Improvements in glucose and blood pressure con- was a:
trol were demonstrated at service, i.e., those referred 62 % increase in the number achieving an
to the specialist multidisciplinary team, and city Hba1c 7.5 % (58 mmol/mol)
level, i.e., outcomes for the total practice diabetes 42 % increase in those with and HbA1c 8
population, the real numerator of the service: (64 mmol/mol) or 9 % (75 mmol/mol)
140 P. King
Fig. 8.5 Glucose control in

patients attending InterCare
(ICH) specialist clinics
15 % increase in the number with a blood with data from Derbyshire County (Fig. 8.6,
pressure 150/90 yellow bar). Derbyshire County has a tradi-
30 % increase in those with a blood pres- tional model, but the majority of areas are
sure <140/80 supported by the same specialist team. NDIS
3. Changes in glucose and blood pressure targets took demography in to consideration, and
in Derby City, i.e., the practice wide popula- Derbyshire County has a different demo-
tion of people with diabetes: graphic mix to the City, and would be expected
The National Diabetes Information Service/ to score better HbA1c targets than Derby City.
Yorkshire Public Health Observatory (NDIS), They fact the two areas were equivalent in
now part of the National Cardiovascular 2011/2012 was indirect evidence for the ben-
Intelligence Network [16], collects diabetes ets of integrated care.
outcome data at primary care trust and more All these measures reflected the effective-
recently clinical commissioning group level. ness of the service.
As all Derby City practices were part of inte- 4. Cost savings:
grated care, comparing data serially gave an NDIS compares outcomes with expenditure
indication as to whether there was an impact (DOVE tool). Expenditure across different
on outcomes within primary care as opposed areas is shown as a ranking based on 152 pri-
to improving outcomes in the patients seen by mary care trusts. Figure 8.7 demonstrates that
the specialist team as discussed above. Data in 20092010 Derby City ranked 16th least
from 2009/2010, 2010/2011 and 2011/2012 expensive for total spend on diabetes, but once
were examined as it was not possible to use integrated care became established, the
this tool beyond 2012 as the data collection ranking fell to the 2nd lowest, while the equiv-
changed to clinical commissioning group alent ranking for Derbyshire County in 2010
level, which for Southern Derbyshire included 2011 was 14th. Data for Derby City and
areas that did not have access to integrated Derbyshire County are shown in Table 8.1.
care. There was a reduction in costs for total pro-
Both blood pressure and glucose outcomes gramme spend and prescribing in Derby City as
improved progressively between 2009 and a result of implementing integrated care. Costs
2012. The data for the percentage achieving a were consistently less than those for Derbyshire
HbA1c <7.5 % (58 mmol/mol) are shown in County in 2010/2011. In 2010/2011, Derby
Fig. 8.6. Clearly factors other than clinical City spent 54 per patient per year less on dia-
improvement such as improvements in data betes care than Derbyshire County, as well as
capture could have inuenced these results. improving outcomes as described above. Given
The 2011/2012 data were therefore compared the diabetes prevalence for Derby City was
Fig. 8.6 Glucose control in

Derby City
Fig. 8.7 Total programme

spend/person with diabetes;
Derby City and Derbyshire
County Primary Care Trust
(PCT) ranking (the lower the
spend, the lower the rank)
14,791, this equates to cost savings of compare admissions with a primary diagnosis of
800,000 annually with improved outcomes. diabetes from practices in Derby City that did
If the model had been extended to Southern and did not have access to integrated care. In the
Derbyshire, the area covered by Southern rst 6 months, there was a reduction of 292 bed
Derbyshire Clinical Commissioning Group, in days and a reduction of mean length of stay from
2012 with 28,000 people with diabetes, the 11 to 6.5 days in patients admitted from the seven
potential savings would be 1.5 million. As InterCare practices compared with the same time
these gures do not include complication period 12 months previously (Fig. 8.8).
management, it would be expected that further Compared with non- InterCare practices,
savings would be seen through the reduction InterCare practice patients spent 1.8 days less in
of complications such as sight threatening hospital (Fig. 8.9). Using a bed day cost for
retinopathy, the number of patients on dialysis Diabetes and Endocrinology of 260, this equates
and amputation rate. to savings of 76,000 over 6 months.
Figure 8.10 shows unplanned admissions with
Inpatient Activity a primary diagnosis of diabetes from all 29
As the InterCare service was initially provided InterCare practices. A progressive reduction in
for 7 practices before rolling out to 29, the rst 6 admissions and bed days is shown between
months of the service gave the opportunity to 20102011 and 20122013. Using the bed day
142 P. King
Table 8.1 Expenditure for Derby City and Derbyshire County (/person with diabetes)
20092010 20102011
Spend Rank/152 PCTs Spend Rank/151PCTs
Total spend
City 418 16 422 2
County 411 11 476 14
Prescribing (all)
City 234 10 240 4
County 244 17 258 14
Non-insulin prescribing
City 50 1 61 16
County 55 3 69
Insulin prescribing
City 124 65 120 58
County 130 95 129 98
Glucose testing
City 58 50 56 44
County 56 39 56 43
Fig. 8.8 Unplanned

admissions with a primary
diagnosis of diabetes; rst 6
months of InterCare compared
with 12 months previously
cost of 260 as above, the savings from the reduc- They received a questionnaire with seven ques-
tion in inpatient activity between 20102011 and tions. Where appropriate they were asked to
20122013 was 270,000. rate the service from 1- very poor to 5- excel-
lent. Twenty-one questionnaires were returned.
These outcomes reflected effectiveness and A high level of clinician satisfaction was
improvements in safety. reported with:
91 % rating their experience referring patients

User and Referring Staff Experience to InterCare as excellent or very good;
73 % describing the experience as excellent or
Staff Experience very good compared with referrals into
In 2012, the 29 referring practices were sur- hospital;
veyed and asked about their experience of 100 % rating their experience of tasking clini-
being part of the InterCare Diabetes Service. cians as very good or excellent.
Fig. 8.9 Unplanned

diagnosis of diabetes; mean
length of stay (LOS) in the
rst 6 months of InterCare
Health, compared with
practices that did not have
access to integrated care, also
compared with 12 months
previously
Fig. 8.10 Unplanned

diagnosis of diabetes all
InterCare practices
All clinicians were aware of the referral process. A patient participation group helped develop
Twenty-eight percent received training through aspects of the service and a focus group of users
an accredited course, 76 % have attended courses with Type 1 diabetes helped to rene the service
but many requested ongoing education and train- to meet their needs as described above.
ing. Practice nurse comments included: We used video to capture some user views. The
Joys visits are fantastic. video can be accessed using the following link:
Karen is very helpful and always sorts out any https://dl.dropboxusercontent.
problems. Quick response to queries. com/u/75004548/intercareviewsvideo.wmv
InterCare is fantastic! Formal feedback was obtained annually using
a questionnaire which explored their experiences
GP comments included: in the service using a 5 point score with 5 being
InterCare was able to help with patients who are excellent and 1 poor. The questionnaire was
difcult and we need another opinion. administered to 50 users in 2012. The process
Whilst there is merit in e-mail dialogues. It helps was repeated in 2013, but this time the survey
to meet face to face. (Consultant practice visit) was administered by users instead of InterCare
administrators. The results were similar. Overall:
User Experience
A variety of methods were used to capture user 85 % rating the service as excellent or very good;
experience. Users were encouraged to feedback 70 % felt that InterCare was excellent or very
and make suggestions for service improvement. good compared with their previous care;
144 P. King
80 % felt the waiting time between referral Gino DiStefano, Garry Tan, Mark Browne, Stuart
Holloway, and Musaddaq Iqbal who had a part in the
and rst appointment was excellent or very
development and running of the InterCare Diabetes
good; Service.
90 % felt that InterCare was able to help them
with the problem they were referred with;
95 % stated they would recommend the ser- References
vice to another person with diabetes.
1. Diabetes UK Diabetes Facts and Figures. https://
Free text comments included: www.diabetes.org.uk/Documents/Position%20state-
ments/Facts%20and%20stats%20June%202015.pdf
like the services altogether, i.e., nurse, dietician (2015). Accessed 24 Mar 2016.
and eye screening. 2. Health and Social Care Information Centre, Quality
prefer the continuity in ICHmakes me feel Outcomes Framework. http://qof.hscic.gov.uk/search/
condent. index.asp (2015). Accessed 24 Mar 2016.
ICH is better because I have better control and 3. Royal College of Physicians, the Royal College of
seen quickly. General Practitioners and the Royal College of
Paediatrics and Child Health. Teams without walls.
The value of medical innovation and leadership.
User and staff feedback provided evidence of https://www.rcplondon.ac.uk/sites/default/files/
teams-without-walls-1_0.pdf (2008). Accessed 7 May
delivering a person centred service.
2015.
4. Goodwin N, Smith J, Davies A et al. Integrated care
for patients and populations: improving outcomes by
Summary working together: a report to the Department of
Health and Future Health Forum. http://www.kings-
fund.org.uk/sites/files/kf/integrated-care-patients-
InterCare was an integrated model for diabetes populations-paper-nuffield-trust-kings-fund-
care, which was the rst diabetes model in the january-2012.pdf (2012). Accessed 9 May 2015.
UK that integrated not just clinically, but also 5. Starey N. What is clinical governance? http://www.
medicine.ox.ac.uk/bandolier/painres/download/wha-
achieved integration at a nancial, organisational
tis/WhatisClinGov.pdf (2001). Accessed 24 Mar
and information technological levels too. The 2016.
whole service was supported by a single clinical 6. Joint Societies. Best practice for commissioning dia-
governance structure. betes services. An integrated care framework. http://
www.diabetes.org.uk/Documents/Position%20state-
The service was evaluated using the care bun-
ments/best-practice-commissioning-diabetes-
dled approach described, and showed improve- services-integrated-framework-0313.pdf (2013).
ments in all six domains of quality improvement: Accessed 7 May 2015.
effectiveness, safety, efciency, timeliness, 7. Diabetes UK. Improving the delivery of adult diabe-
tes care through integration. Shared experience and
equity and provision of a person centred service.
learning. http://www.diabetes.org.uk/Documents/
InterCare resulted in 800,000 savings in About%20Us/What%20we%20say/Integrated%20
20102011 with 270,000 saving as a result of diabetes%20care%20(PDF,%20648KB).pdf (2014).
reducing admissions between 2010 and Accessed 7 May 2015.
8. British Medical Association. Collaborative GP net-
20122013.
works basic legal structures. http://www.bma.org.
This model is currently being further devel- uk/support-at-work/gp-practices/gp-networks/basic-
oped and extended to provide a single integrated legal-structures (2015). Accessed 26 Mar 2016.
service for all people in Derby City and Southern 9. Rea RD, Gregory S, Browne M et al. Integrated dia-
betes care in Derby new NHS organisations for new
Derbyshire, the area covered by the current com-
NHS challenges. http://www.practicaldiabetes.com/
missioners, Southern Derbyshire Clinical S p r i n g b o a r d We b A p p / u s e r f i l e s / e s p d i / f i l e /
Commissioning Group. September%202011/Rea%20online%20+%20charts.
pdf (2012). Accessed 16 May 2015.
10. Department of Health. A simple guide to payment
Acknowledgement I am grateful to Dr Kyran Farrell, a
by results. https://www.gov.uk/government/
GP with a special interest in diabetes, for his helpful com-
uploads/system/uploads/attachment_data/
ments. I would also like to acknowledge Dianne Prescott,
le/213150/PbR-Simple-Guide-FINAL.pdf (2012). 15. NHS Modernisation Agency. Change no 6. Increasing

Accessed 24 Mar 2016. the reliability of performing therapeutic interventions
11. King P. Proceed. www.derbyproceed.co.uk (2014). through a care bundled approach in 10 high impact
Accessed 9 May 2015. changes for service improvement and delivery. http://
12. Lillyman S, Saxon A, Treml H. An evaluation of the www.nursingleadership.org.uk/publications/HIC.pdf
role of the community matron a literature review. (2004). Accessed 14 May 2015. (Please note that this
https://www.wlv.ac.uk/media/wlv/pdf/nov2009_ is a NHS Institute for Innovations and improvement
SLillyman.pdf (2009). Accessed 10 Oct 2015. publication their website no longer exists but the
13. Donabedian A. The quality of medical care. Science. publication can still be accessed through the link
1978;200:85664. provided.)
14. Institute of Medicine. Cross the quality chasm: a new 16. The National Diabetes Information Service. http://
health system for the 21st century. https://www.iom. www.yhpho.org.uk/resource/view.aspx?RID=102082
edu/~/media/Files/Report%20Files/2001/Crossing-the- (2015). Accessed 14 May 2014. (Data after 2013 are
Quality-Chasm/Quality%20Chasm%202001 %20%20 part of National Cardiovascular Intelligence Network,
report%20brief.pdf (2001). Accessed 14 May 2015. accessible through the same link.)
Integrated Diabetes Care:
Coventry andWarwickshire 9
Approach
PonnusamySaravanan, VinodPatel,
JosephPaulOHare, andSudheshKumar
Introduction emphasis on reducing this cost and improving

(or at least without compromising) the quality of
The clinical and economic burden of Type 2 care, over the past two decades [3]. Case-
Diabetes (T2D) is very high across the world as finding approaches [4], care closer to home
the number of people suffering from diabetes is and pay for performance [5] are such exam-
increasing rapidly, even in low-income countries. ples. In the UK, prior to the introduction of pay
A total of 380 million people live with diabetes for performance as a part of the Quality and
and health expenditures of 548 billion US dollars Outcomes Framework (QoF) in 2004, there was
(USD) which is 581 (ID), were spent on the con- wide primary care variation in the care and
dition in 2013 across the world [1]. The vast prompt referral of patients with T2D [6].
majority of this is spent in high-income countries. Although the QoF initiative improved overall
The healthcare cost for managing chronic dis- process and intermediate outcome measures, sig-
eases such as Diabetes is increasing even in these nificant disparities in diabetes outcomes still
high-income countries [2]. There has been much exist between general practices and regions,
especially among patients from ethnic minorities
[7]. These intermediate improvements appear to
P. Saravanan (*)
George Eliot Hospital NHS Trust, Coventry, UK be levelling off, which may partly be due to less
challenging targets to secure the QoF points, as
Warwick Medical School, The University
of Warwick, Coventry, UK well as increasing use of exception reporting
(proportion of patients who can be excluded by
University Hospitals Coventry and Warwickshire
NHS Trust, Coventry, UK the practice team for a variety of reasons) [8].
e-mail: [email protected] There is now evidence that these improvements
V. Patel in individual risk factors have not translated into
George Eliot Hospital NHS Trust, Coventry, UK improvements in hard outcomes [9].
Warwick Medical School, The University of The majority of the practices across the UK
Warwick, Coventry, UK achieved the QoF targets within a few years of
e-mail: [email protected] the introduction of the incentive scheme.
J.P. OHare S. Kumar However, nationwide audit data for England
Warwick Medical School, The University of 20092010 showed that more stringent targets
Warwick, Coventry, UK for HbA1c (7.5%/58.5 mmol/mol), blood pres-
University Hospitals Coventry and Warwickshire sure (BP) (<140/80 mmHg) and total cholesterol
NHS Trust, Coventry, UK (<4.0 mmol/l) were achieved in only 67%, 69%
e-mail: [email protected]; Sudhesh.kumar@
warwick.ac.uk and 41% of people with T2D.Poor glycaemic

D. Simmons et al. (eds.), Integrated Diabetes Care, DOI10.1007/978-3-319-13389-8_9
148 P. Saravanan et al.
control was associated with younger age and ally when there is e vidence of complications.
social deprivation [5]. Significant variations There is no structured educational support for
existed with practices that are less likely to the primary care team that provides regular
achieve adequate levels of control, in areas of update on diabetes knowledge. In addition,
high deprivation and serving populations with Payment by results (PBR) has also actively,
higher proportions of Black and Minority Ethnic albeit indirectly, disincentivised primary care
(BME) groups [7, 10] as well as those practices to seek opinion from specialist services [13].
with lower levels of organisation [11]. West GPs with a special interest in Diabetes (GPwSI)
Midlands South (theCoventry and Warwickshire were thought to offer a solution to this problem
area) and East Midlands (particularly Leicester and were rapidly adopted across the UK
city) are such areas with high BME population between 2004 and 2006. The GPwSIs either on
density coupled with lower socioeconomic status their own, or in collaboration with hospital dia-
among White Caucasians. In Coventry and North betes specialists, set up alternative services to
Warwickshire there were also significantly higher hospital-based services across the UK.These
numbers of small general practices. Although clinics are called Intermediate Care Clinics for
geographically close and densely populated Diabetes (ICCD). The very first ICCD set-up
(~800 mi2 and ~850,000 population), the Coventry was in Bradford in 1998, based in 19 clinics
and Warwickshire region has three different NHS across the city, and led by GPwSIs, supported
providers (hospital trusts) and initially six differ- by a community diabetologist, diabetes special-
ent commissioning groups (which became three ist nurses, dieticians and podiatrists. An evalua-
primary care trusts and then three clinical com- tion conducted between 1999 and 2001 found
missioning groups). This added further challenges high levels of uptake, with over 2000 referrals
in providing cohesive, integrated care within the [14]. During this period, hospital attendance
region for the 39,000 patients with diabetes, fell by 25%, but overall attendance (hospital
catered for by 147 general practices. and ICCD) increased by 35%. Intermediate
care clinics were popular with patients and
practitioners and their cost was similar to a hos-
lternate Services toTraditional
A pital outpatient clinic. Problems identified
Hospital Based Clinics included inequity in referrals between practices
in which the clinics were based and other prac-
Increasingly, patients with diabetes have been tices, and lack of communication and strategic
managed in primary care across the world. This planning [14]. Similarly, several other regions
trend was partly due to the increasing number (Southampton, Lambeth in London, and Wales)
of patients with T2D and the inability of exist- had their own model of ICCD clinics with vary-
ing specialists to cope with the demand. Our ing success [1517]. Although recent evidence
local audit data shows that approximately has suggested that the most effective interven-
8590% of patients with diabetes are managed tions in Diabetes include the trial of team
by GPs and practice nurses in Coventry and changes, case-finding and management and
Warwickshire. Only a small proportion of patient education and empowerment [18], none
newly diagnosed patients with T2D (typically of these models had a unifying methodology. In
around 510%) who attend the DESMOND addition, none of the ICCD services that have
(Diabetes Education and Self-Management for been set up were evaluated in randomised clini-
Ongoing and Newly Diagnosed) education pro- cal trials or had their cost-effectiveness assessed
gramme come into contact with some aspect of with appropriate health economic evaluations.
the specialist services [12]. Apart from this This has raised serious questions about the
group, patients with T2D come into contact validity of the clinical and cost-effectiveness of
with a member of the specialist services, usu- ICCD clinics.
9 Integrated Diabetes Care: Coventry andWarwickshire Approach 149
I nefficiency DuetoLack improved glycaemic control, lower blood pres-

ofIntegrated Services sure and better management of lipids [2123].
Patients with improved glycaemic control also
Large volumes of data are collected by various consistently report better functional status and
services ranging between primary care, local lab- wellbeing [2426]. Studies showed that intensive
oratory facilities, ambulance services, hospital control of risk factors can be achieved in the UK
clinics (of varying specialties), retinal screening at comparable cost. However, detailed cost-
services and several allied healthcare profession- effectiveness of this strategy was not discussed
als. However, the majority of these systems are [27]. The challenge for diabetes services across
not unified and therefore result in duplication of the UK and the world is how to translate these
data collection and lack of data utilisation beyond interventions into individualised patient care.
the purpose of collection. This can result in Whilst the QoF has raised the standard of cardio-
missed opportunities, delayed communication, vascular risk factor data collection (HbA1c, BP
inability to use electronic solutions (prompts, and Cholesterol), these have been primarily used
alerts, algorithms etc.), inefficient use of as cross-sectional data in patient management,
resources and patient fatigue (repeated testing but which is probably one of the reasons for the lack
no apparent benefit). Thus, in the majority of the of improvement in hard outcomes [9]. Although
regions in England, the delivery of diabetes care available data have been more complete since the
is disjointed and lacks integration. Each service introduction of QoF in 2004, they have not been
collects and utilises data for their own narrow utilised for aggressive case-finding or indi-
purpose, which could be used in a holistic way vidualised cardiovascular risk management.
with the patient at the centre of the care (Fig. Trends in deterioration could potentially be iden-
9.1a, b). tified by electronic solutions at individual patient
A meta-analysis that assessed the effective- level for each risk factor and used for proactive
ness of 11 strategies to improve HbA1c in patients risk management.
with T2D managed in primary care showed that To improve the local services in diabetes care
only three strategies resulted in improvement in Coventry and Warwickshire, two studies were
(team changes, case-finding and management, set up. First was a cluster randomised controlled
and patient education and empowerment) [18, trial to evaluate the effectiveness and cost-
19]. Intensive management can also reduce com- effectiveness of the ICCDs based in three pri-
plications such as retinopathy, nephropathy and mary care trusts (PCT) in Coventry, North
neuropathy, as well as reducing the risk of cardio- Warwickshire and Leicester [28, 29]. The second
vascular disease [20]. Benefits are gained from was an innovation project funded by Department
Fig. 9.1(a) Current disjointed services. (b) An integrated service model

of Health (Health Innovation Education Cluster), recruited to the study were randomised to either
with a view to digitally integrate diabetes data (i) usual care or (ii) intervention, with the latter
across 12 GP practices. These two projects are having access to the new ICCD clinics. Practices
discussed in this chapter. randomised to usual care were issued with local
guidelines and were advised to manage their
patients in the usual way.
I ntermediate Care Clinics
inDiabetes Study I nclusion andExclusion Criteria
Inclusion criteria were kept broad. All T2D
Detailed protocol, study design and results have patients aged 18 years or over were eligible to
been published [28, 29]. participate. Patients with following characteris-
tics were excluded:
Aims andObjectives 1 . Severe cognitive impairment

2. Severe mental illness
The overall aim was to evaluate the effectiveness 3. Receiving terminal care
and cost-effectiveness of community based inter-
mediate care clinics in the management of T2D. Outcome Measures
Specific objectives were to:
Primary Outcome
1. Compare the following in patients with T2D Proportion achieving ALL the targets HbA1c
registered with practices that have access to (<7.0%), blood pressure (<140/80), cholesterol
ICCD with those that have access only to (<4 mmol/l)
usual care:
Glycaemic control, Secondary Outcome
Control of blood pressure and 1. Proportion achieving individual risk factor

dyslipidaemia, targets:
Quality of life, 2. Mean HbA1c
Satisfaction with services and continuity of 3. Mean Systolic and Diastolic Blood Pressure
care, 4. Mean Total Cholesterol,
Referral patterns, 5. Ten year CVD risk score by UKPDS risk

Annual cost per patient with diabetes. engine [30]
2. Estimate the difference in the cost of the

resources used by patients in each arm of the Other Assessments
trial, and the cost-effectiveness of the ICCD Health status assessment EQ-5D [31]
intervention.
3. Explore the views of users and service provid- 1 . Problem Areas in Diabetes (PAID) [32]
ers participating in the study. 2. Continuity of Care questionnaire [33]
3. Economic outcome measure data on NHS
resource use and personal costs [34]
Methods
ample Size Calculations
S
Study Design Percentage of patients achieving control in the
It was a pragmatic two-arm cluster randomised usual care group in a UK survey prior to funding
controlled trial in three different PCTs application were: median practice performance
(Coventry PCT1, North Warwickshire PCT2 in achieving HbA1c <7.548%; blood pressure
and Leicester city PCT3). Randomisation was <145/8559%; and cholesterol <560% [35].
stratified by PCT and GP practice size. Practices We used HbA1c for our primary sample size
calculation as this is the outcome variable for cholesterol targets, there was a substantial differ-
which there is the most robust information on ence between groups with respect to the com-
intra-class correlation (ICC) which has been esti- bined control (primary outcome), with 11.2% in
mated as 0.047 [36]. To detect a difference the intervention and 8.7% in the control arm.
between percentage well controlled of 50% in Table 9.2 shows there were large differences
the control group and 60% in the intervention between PCTs in levels of control at baseline, in
group (alpha = 0.05 Power = 0.8), not allowing particular, in the rates of achievement in the
for clustering, required a sample size of 408 sub- combined control and cholesterol control
jects in each arm. Using an ICC of 0.047, and between PCTs 1 and 3.
with 72 patients in each cluster, the necessary In PCT 1 and 2, the ICCDs that existed at the
sample size in each arm was 1770, a total of 3540 time of funding application were closed due to
patients. This number is also adequate to detect a lack of adequate referrals. Therefore for the pur-
10% difference in total cholesterol control (from pose of the study, at study initiation two clinics
60% to 70%) and blood pressure control (from were set-up. These were agreed between the
60% to 70%). Estimates of ICC for blood pres- local consultants, GPs and PCT.The consultants
sure and cholesterol were taken from UKADS, a provided their time for the ICCDs at no charge,
study of care provision for people of South Asian community Diabetes Specialist nurses provided
ethnicity with diabetes [37]. Assuming the ICC their time as a part of their community role and
for our combined primary outcome (adequate one GP (PCT 1) provided her time at no charge
control of HbA1c, blood pressure and choles- for the purpose of the study in order to work
terol) was 0.05 and achievement was at 20% alongside an experienced consultant. All these
(from a baseline of 1520% as suggested by clinics (all three PCTs) ran for 18 months. In
local audit data) in the control arm and 30% in PCTs 1 and 2, only trial participants were
the intervention arm, we would need a total of referred to ICCD.These were 145 and 35
2848 patients. patients, respectively. In PCT 3, all patients of
The study was successful in recruiting 49 practices randomised to the intervention group
practices (11, 13 and 25in PCT 1, 2 and 3 respec- were eligible for referral this was due to the
tively). A total of 1997 patients were enrolled local PCT guidelines at that time. The overall
with an average of 42 per practice. This sample recruitment rate of patients to the trial from the
size had 80% power to detect a 12% difference intervention arm was 19%, and so we estimate
in the combined outcome measure, if 75% fol- that of the 101 patients who attended clinics in
low-up was achieved. PCT 3, 19 (19%) were trial participants. Thus a
total of 199 (145+35+19) trial participants
attended ICCD, representing 18.8% of trial par-
Results ticipants. The proportion of participants referred
to ICCD varied across the three PCTs: 145/431
Of the 49 practices, 24 were randomised to inter- (34%) in PCT 1, 35/240 (15%) in PCT 2 and
vention and 25 to the control arm. Of the 1997 19/386 (5%) in PCT 3.
patients recruited 64% were followed up. Two Follow-up data was available for 1280
practices declined to allow the research team to patients. Last observation carried forward method
undertake follow-up assessments. The consort (LOCF) was applied for missing values: primary
diagram (Fig. 9.2) shows the recruitment, follow- outcome variable (combined control), individual
up rates and numbers analysed. secondary outcome variables, and each of the
Table 9.1 shows the baseline characteristics by covariates used in the primary analysis. Thirty-
trial arm. It shows although the intervention and three patients were excluded fromthe final analy-
control groups were similar at baseline with sis due to missing data, with no possible
respect to gender, smoking status, co-morbidities LOCF.The baseline characteristics of the final
and achievement of blood pressure, HbA1c and 1247 patients are shown in Table 9.3.
Assessed (51 practices)
Refused (2 practices)
Randomised (49 practices)
Intervention (24 practices) Control (25 practices)

1 cluster dropped out prior to baseline 2 clusters dropped out prior to baseline
data collection data collection
Received intervention Control group
23 Practices, mean (sd) size 48.0 23 Practices, mean (sd) size = 40.9
(27.8), range 11-126. (22.7), range 16-92.
Total 1057 participants Total 940 participants
Lost to follow-up Lost to follow-up

2 Practices, mean (sd) size 15.9 (26.0), 0 Practices, mean (sd) size 8.85 (8.94),
range 0-90. range 0-26.
Total 324 patients Total 204 patients
Analysed Analysed
Intervention: 21 practices, Control: 23 practices,
mean (sd) size 28.0 (19.3), mean (sd) size = 27.7 (16.3),
range 0-72 range 9-66,
Excluded from analysis: Excluded from analysis:
89 patients excluded**why 100 patients excluded**why
Participants Participants
Total 644 analysed Total 636 analysed
Fig. 9.2 GP Practice and patient recruitment and progress through the study
Primary Outcome is normally distributed. Equivalently, a separate

linear regression model is calculated for each
A series of analyses were performed for the pri- practice, but the coefficient for each covariate is
mary outcome using logistic mixed effect mod- the same in each of those models. This approach
els. Such models compensate for variations at the allows consideration of the effect of group-level
GP practice level by modelling the contribution variance separately to the effect of individual-
of the practices as a random variable with a nor- level variance between groups.
mal distribution. In effect, this allows the model In the first, main analysis, general practice
to have multiple intercepts, one for each practice, was included as a random effect, with the follow-
where the intercepts variation from a fixed point ing covariates included as fixed effects:
Table 9.1 Baseline characteristics by trial arm

Control (n=940) Intervention (n=1057)
Variable n (%) n (%)
Achieved combined control 81 (8.7) 116 (11.2) *
Controlled HbA1c (<=7.0%) 497 (53.9) 536 (51.7)
Controlled blood pressure (<140/80) 304 (32.8) 398 (38.3)
Controlled cholesterol (<4 mmol/l) 442 (48.2) 519 (50.2)
PCT 1 242 (25.7) 431 (40.8)
PCT 2 225 (23.9) 240 (22.7)
PCT 3 473 (50.3) 386 (36.5)
Male 543 (58.1) 613 (58.4)
Smoker 118 (12.7) 116 (11.1)
Hypertension 505 (55.6) 612 (59.1)
IHD 161 (17.7) 149 (14.4)
CVD 35 (3.85) 28 (2.72)
Heart failure 25 (2.75) 35 (3.38)
PVD 10 (1.10) 15 (1.45)
Renal failure 24 (2.63) 24 (2.31)
Ethnicity: White 614 (65.3) 554 (52.4)
Asian 257 (27.3) 405 (38.3)
Black 32 (3.40) 55 (5.20)
Other 37 (3.94) 43 (4.07)
Table 9.2 Baseline characteristics by PCT

Coventry (n=673) Nuneaton (n=465) Leicester (n=859)
Variable (PCT 1) n (%) (PCT 2) n (%) (PCT 3) n (%)
Achieved combined control 48 (7.2) 42 (9.4) 107 (12.6)
HbA1c (<=7.0%) 323 (48.5) 256 (58.3) 454 (53.2)
Blood pressure (<140/80) 228 (34.2) 170 (38.0) 304 (35.6)
Cholesterol (<4 mmol/l) 270 (40.5) 190 (44.1) 501 (58.7)
Male 383 (57.2) 290 (63.2) 483 (56.6)
Smoker 74 (11.1) 53 (11.8) 107 (12.5)
Hypertension 393 (59.2) 351 (78.0) 373 (44.9)
IHD 92 (13.8) 88 (19.6) 130 (15.7)
CVD 20 (3.00) 35 (7.78) 8 (0.97)
Heart failure 19 (2.86) 12 (2.67) 29 (3.50)
PVD 6 (0.90) 14 (3.12) 5 (0.60)
Renal failure 15 (2.25) 24 (5.33) 9 (1.08)
Table 9.3 Baseline characteristics of patients used in the final analysis

Control (n=636) Intervention (n=591)
Variable N (%) N (%) p-value
PCT 1 164 (25.8%) 166 (28.1%) 0.399
PCT 2 339 (53.3%) 237 (40.1%) <0.001
PCT 3 133 (20.9%) 152 (25.7%) 0.054
Male 370 (58.2%) 347 (58.7%) 0.894
Smoking 77 (12.1%) 66 (11.2%) 0.672
Co-morbidity
Hypertension 341 (53.6%) 335 (56.7%) 0.307
IHD 115 (14.9%) 95 (16.1%) 0.392
CVD 22 (3.46%) 15 (2.54%) 0.438
Heart failure 15 (2.36%) 17 (2.88%) 0.697
PVD 7 (1.10%) 9 (1.52%) 0.690
Renal failure 13 (2.04%) 12 (2.03%) 1.000
Ethnicity
White 365 (57.4%) 271 (45.9%) <0.001
Asian 98 (15.4%) 202 (34.2%) <0.001
Black 20 (3.14%) 33 (5.58%) 0.050
Other 20 (3.14%) 22 (3.72%) 0.690
Baseline assessment of outcome measures
Primary outcome (combined control) 61 (9.59%) 76 (12.9%) 0.084
HbA1c (<=7.0%/53 mmol/mol) 347 (54.6%) 326 (55.2%) 0.878
Blood pressure (<140/80 mmHg) 354 (55.7%) 324 (54.8%) 0.812
Cholesterol (<154 mg/dl/4 mmol/l) 305 (48.0%) 308 (52.1%) 0.162
Individual factors (mean/sd)
HbA1c (%) 7.22 (1.24) 7.18 (1.23) 0.470
Systolic BP (mmHg) 137.5 (17.3) 137.0 (18.0) 0.528
Diastolic BP (mmHg) 80.6 (10.0) 79.3 (10.7) 0.005
Total cholesterol 4.05 (1.04) 3.99 (1.18) 0.231
Intervention arm, combined control and age at 1.59 (95% CI: 0.983, 2.49). Not surprisingly,
baseline, gender, ethnicity, smoking status at fol- only the baseline levels of combined controlled
low- up, PCT, deprivation and presence of was a significant contributor to this OR but not
co-morbid conditions such as hypertension, isch- the other covariates (as highlighted earlier),
aemic heart disease, cerebrovascular disease, including deprivation index. Similar effects were
heart failure, peripheral vascular disease and seen when intention to treat analysis was carried
renal failure. Intervention arm was considered out (n=1910, aOR 1.60 (95% CI: 0.984, 2.60).
the main exposure variable, estimated as a fixed
effect.
The achievement of combined and individual Secondary Outcomes
risk factor control by trial arm is shown in Table
9.4. The odds ratio (OR) of achieving the primary The mean values of individual risk factors
outcome (combined control) was significantly achieved at the baseline and follow-up by inter-
better in the intervention arm at 1.62 (95% CI: vention and control group are shown in Table 9.5.
1.04, 2.43) though when adjusted for covariates, The OR of the primary outcome and these key
it lost significance. The adjusted OR (aOR) was secondary outcomes are shown in Table 9.6.
Table 9.4 Proportion of patients achieving risk factor Table 9.6Odds ratios for primary and secondary
control at the end of the study outcomes
Intervention (644) Control (636) Odds 95% confidence
Baseline (B) or ratio interval
follow-up (F) B F B F Primary outcome 1.56 (0.983, 2.49)
% Achieved 12.9 14.3 9.7 9.3 (composite)
combined (82) (92) (61) (59) Secondary outcome measures
control (n) HbA1c control (<=7.0%) 1.45 (1.07, 1.96)
% Controlled 54.9 57.5 55.3 51.1 Blood pressure control 1.23 (0.88, 1.73)
HbA1c (n) (349) (370) (347) (325) (<140/80)
% Controlled 40.0 39.8 33.2 32.0 Total cholesterol (<4 1.48 (1.08, 2.03)
blood pressure (254) (256) (209) (203) mmol/l)
(n)
% Controlled 52.9 61.8 48.8 55.2
cholesterol (n) (335) (397) (305) (351)
based on 2010/2011 figures. Average costs per
consultation and patient attending at the three
Table 9.5Mean risk factor control at baseline and sites are shown in Table 9.7.
follow-up Questionnaire responses at follow-up were
Intervention Control used to derive patient costs in terms of NHS
Baseline (B) or resources used. This is referred to as the resource
follow-up (F) B F B F use element of the study questionnaire.
HbA1c % 7.18 7.17 7.22 7.28
(n=1249) nalysis ofDirect Costs
A
Systolic BP mmHg 137.0 136.9 137.5 138.0 For calculating the direct healthcare cost clean
(n=1251)
dataset of n=1322 is used. Patients in the inter-
Diastolic BP mmHg 79.3 79.1 80.6 80.5
(n=1251) vention group did not have a statistically signifi-
Cholesterol 3.99 3.79 4.05 3.90 cant difference in total resource use costs
(n=1245) mmol/l (p=0.101). However, breaking the data down
into the cost headings, in the intervention group,
the cost of primary care and community clinic
There were no differences seen in the UKPDS consultations were higher than for the control
risk score between the groups. Similarly, there group (184.98 vs. 76.82; p<0.001). This is likely
were no differences in any of the four domains of to be because some of the ICCD attendances are
the Continuity questionnaire or the PAID included in this category and/or because ICCD
questionnaire. input triggered more primary care consultations.
The detailed breakdown of the direct costs is
shown in Table 9.8.
Economic Evaluation
Indirect Costs
ost ofConsultations inICCD
C Only around 50% of the participants responded
Separate data were collected through direct con- to the sickness rates. This is probably partly the
tacts with the ICCD sites independent of the trial. reflection of the age group studied (retired).
These data were used to estimate the cost of see- Therefore, sickness was not included in the indi-
ing a patient in each of the IC clinics. As described rect cost in the economic analysis.
earlier, in PCTs 1 and 2, clinics were available
only to trial participants, and so the total cost of ost Utility Analysis
C
these services has been included. In PCT 3, only The QALY gain during the baseline to follow-up
a minority of clinic attenders were trial patients period was calculated by using the following
(19%). In this PCT the cost per consultation was formula:
Table 9.7 Average cost per ICCD consultation/patient attending at each site
Total consultations/patients Average cost per consultation/
PCT Total costs attending ICCD patient attending ICCD
1 43,553 442/145 98.54/300.37
2 8881 120/35 74.01/253.74
3 14,701 95/19 154.75/773.74
Total 67,135 657/199 102.18/337.36
Table 9.8 Direct cost spent according to resources utilised

Intervention (n=665) Control (n=657)
Resource item Mean (SE) Mean (SE) P value
ICCD
Intermediate care clinic for 60.18 N/A
diabetes
Cost of consultations
Primary care doctor and nurse 37.25 2.335 31.19 2.044 0.051
costs
Community clinic staff 1.46 0.381 0.49 0.201 0.025
Hospital doctor and nurse costs 26.13 3.876 32.03 5.272 0.366
AE staff 1.02 0.525 0.59 0.295 0.476
Optometrist, podiatrist and 11.65 1.047 12.51 0.907 0.534
dietician
Sub total 137.70 5.53 76.82 6.25 <0.001
Cost of care
Diabetes tests 58.27 2.27 62.74 2.63 0.199
Hospital inpatient costs 155.71 75.60 98.96 58.67 0.554
Total costs 351.68 76.51 238.52 60.70 0.247
formula, the following calculations were

0.5 ( BL Quality + FU Quality ) performed:
QALY = /
* No. of months between BL and FU
12 ( BL = Baseline; FU = Follow up ) Cost utility = ( Control cost Intervention cost ) /
( Control QALY Intervention QALY )
[38]. Individual scores for patients were summed
to give a total QALY gain for each trial partici- The incremental cost-utility ratio focused on
pant, and also summed at the level of the two the costs per QALY gained. Bootstrap resam-
cohorts. As there was no significant difference in pling with 1000 replications was performed.
baseline Quality of Life scores between two Scatter plots of 1000 bootstrapped ICERs on the
groups (0.69 vs. 0.70, p=575), the QALY analy- cost-effectiveness plane were generated. The
sis was not adjusted for baseline QALYs [39]. cost-effectiveness results were presented as Cost-
The incremental cost-effectiveness ratio (ICER) Effectiveness Acceptability Curves (CECA).
was calculated using the usual formula These show the probability that the intervention
( C1 C0 ) / ( E1 E 0 ) , with C being costs and E group is cost-effective relative to the control,
the clinical or QALY outcome. Using this given varying values of ceiling ratios, i.e., the
willingness to pay (WTP) for one quality- Cost-Effectiveness

adjusted life-year. In this analysis we used the Due to the uncertainty around ICER estimates,
threshold amount of 30,000 based upon the cost-effectiveness acceptability curves (CEACs)
threshold figures usually employed by NICE are often used within theeconomic evaluation of
[40]. For the QALY data and ICER analysis, as clinical trials [41]. The CEAC curve below indi-
per NICE recommendation these values were dis- cates the probability that the intervention group
counted at 3.5% [40]. Table 9.9 show the sum- is cost-effective relative to the comparator group
mary data used in ICER analysis. for a range of possible societal valuations of a
This suggests that the intervention group had QALY.If, for any given valuation of a QALY, the
ahigher average cost per patient, primarily CEAC reaches or exceeds a 95% probability,
because of the higher hospital cost utilised by a then it is possible to conclude that this interven-
small number of patients. Therefore, sensitivity tion is cost-effective relative to the control group.
analysis was carried out by removing these outli- Sensitivity analysis was carried out by removing
ers. The cost was estimated by matching the sam- outliers who had hospital admission stay of more
ple to those completed the EQ-5D.By this than 10 days. CEAC were recalculated and
method the median cost for QALY gained was shown in Fig. 9.3. This demonstrated that there
marginally higher in the intervention group is a high probability (>95%) that the interven-
(7912 vs. 7778) after applying 1000 replicated tion group was more cost-effective than the con-
bootstraps. trol group above the cost per QALY threshold
(20,00030,000). However, these data should
Table 9.9 Summary data for ICER

QALY Cost
Intervention Control Intervention Control
Mean 1.274 1.219 1041.14 418.58
Standard error 0.001 0.001 6.92 1.54
Fig. 9.3 Cost-effectiveness acceptability curve

be interpreted with caution as the data were still take advantage of these opportunities, but in
skewed despite the sensitivity analysis. In addi- practice the time available for them to do this was
tion, although only a small number of patients often limited (this particularly seemed to be the
were excluded from the cost analysis (those not case for GPs). Intermediate care professionals
completing the Resources Use questionnaire), a temporarily taking over patient care seemed to
much larger number were excluded from the work well, and was regarded as offering impor-
QALY analysis because they had not completed tant benefits for patients such as the opportunity
their EQ-5D questionnaires at both baseline and to have a fresh look at an ongoing problem and/or
follow-
up. As economic evaluations do not for them to consult a professional with a higher
attempt to test a specific hypothesis, sample size level of expertise.
is less important. However, a reduction in the Whilst these two key forms of support were
sample may affect the reliability of the results. valued, there was, however, general agreement
This should be kept in mind in interpreting these that they should lead to genuine capacity building
economic analyses. within primary care and that practices need to
rely on intermediate care to take over patient care
should reduce accordingly. Participants reported
Qualitative Study Summary that, if an intermediate care service is to work
well, then two main factors are important: close
Previous studies that focused on community- coordination with primary care (both when the
based care in chronic diseases such as diabetes service is being set-up and on an ongoing basis);
have shown that patients prefer care closer to and making the intermediate care service acces-
home. This has resulted in several initiatives by sible and appropriate for patients.
the Department of Health in the UK to drive
chronic disease management to be carried out
in the community, although many did not show Discussion
any clear clinical benefit. Therefore a qualita-
tive study was carried out alongside this ran- The ICCD study was an ambitious attempt to
domised trial to focus on the stakeholders assess the effectiveness of community-based
perspectives. The study results are briefly sum- intermediate clinics in a randomised control trial
marised below. setting. The results show that provision of ICCDs
There was general agreement among partici- did not significantly increase the proportion of
pants across both primary and intermediate care patients achieving good control of hyperglycae-
that most T2D care should take place within pri- mia, blood pressure and cholesterol. When PCT
mary care, but that this was currently not always was removed as a covariate, the primary outcome
possible as general practices varied in terms of measure reached statistical significance
their capacity and levels of expertise. Intermediate (p=0.048); further highlighting the baseline vari-
care was seen as providing useful support to pri- ations between PCTs have probably contributed
mary care, especially for those practices not cur- to the overall outcome. Any improvement, how-
rently able to provide comprehensive diabetes ever, in clinical terms was small reflecting the
support in-house. Two main types of support global experience of difficulty in ensuring
were identified: that ICCDs can help to upskill improved healthcare in this group [18].
primary care professionals; and that patient care Although the trial was not powered to detect
may be temporarily taken over by intermediate the differences between the PCTs, there were
care professionals. Both formal and more infor- important differences between them both at base-
mal opportunities for upskilling were provided line and follow-up. In addition, there were also
across the three sites, and include formal training differences in the proportion of patients attending
sessions, telephone support, and case discussions. the ICCDs in PCTs. This was probably due to the
Primary care staff appeared keen in principle to differences in the case-finding approach used in
the PCTs. In PCT 1 and 2, active case-finding primary care. Primary care appreciated the case
approach was used and GPs were prompted and based educational opportunities on offer. It was
encouraged to use the ICCD services whilst in also perceived that digital integration of data
PCT 3 patients were referred by the GPs as they could provide further benefits for both ease of
saw fit. The active case-finding approach was case-finding as well as individualised risk fac-
carried out by one of the ICCD nurses by looking tor target setting and management. One of the
through the GP records for patients who were not regional leads (Ponnusamy Saravanan) of the
adequately controlled according to their risk fac- ICCD study team secured an innovation grant to
tors. The small changes in risk factor control may implement such integrated service with digital
reflect low referral to ICCDs, which could have integration of data in 12 practices in PCT 1
been improved by more active case management. (Coventry) and PCT 2 (North Warwickshire).
ICCDs are one way to provide such an enhanced This innovation project was funded by
case-finding and more aggressive target manage- Department of Health (Health Innovation
ment service in the community with specialist Education Cluster funding stream) and led by
input. Integrated case management through two of the authors (Vinod Patel and Ponnusamy
case-finding coupled with intensive interven- Saravanan). This was considered the follow-on
tion within existing primary care services and project to the ICCD study, albeit as a pilot study
settings might be equally effective. Though the in a subgroup of GP practices. The rationale,
ICCDs are not explicitly constructed around the aims and objectives, methodology and results of
proven principles of case management and work this study called, Community Based Integrated
force changes [18], PCT 1 and 2 used the proac- service and Education for Diabetes (COMBINE
tive case-finding approach which were led by for Diabetes) will be discussed in the next
new team members. This in part may have been section.
the reason for better uptake for the ICCDs and
improvement in the control arms in these PCTs
(concomitant primary care education and upskill- OMBINE forDiabetes: APilot
C
ing), whilst the uptake was poor in PCT 3, with Study ofaDigitally Integrated
deterioration in the risk factor control in the con- Diabetes Care Model
trol arm.
The economic analysis suggested that ICCD Introduction
is cost neutral and potentially beneficial gain at
QALYs (20,00030,000) though this needs to Although the ICCDs were considered a success
be interpreted with caution due to the skewed dis- with some aspects of the local diabetes care
tribution of hospital cost by a small number of model, one of the key stakeholders feedback was
patients. The findings also raise the notion that the lack of utilisation of the risk factor data that
stratified management of diabetes according to are being collected for the purpose of QoF.In
their phenotype (for example morbidly obese addition, the primary care teams (physicians and
patients may need bariatric surgery) or co- nurses) feel that they do not have opportunities
morbidity (for example patients with mental for high quality, ongoing education in the field of
health problems would need treatment strategies diabetes and cardiovascular disorders, despite the
that target their risk factors as well as their mental increasing emphasis on managing these chronic
health). Such stratified management of diabetes diseases in primary care. Patients are also increas-
might be more cost-effective. ingly frustrated about the lack of ability in assess-
From the qualitative data, all the stakeholders, ing their data by the diabetes specialists in the
including the patients, felt the usefulness of such region for improving their individualised care.
ICCDs, in particular the aspect of care closer to As highlighted earlier in this chapter,
home and the benefits of working closely with improvement in the process measures since the
introduction of QoF in 2004 enabled cardiovas- Objectives

cular risk factors data (e.g., HbA1c, BP, Lipids,
Smoking data, presence of CKD) are collected 1. Improving cardiovascular risk factors of

regularly and available electronically at individ- patients with diabetes Individualised patient
ual patient level. Pulling together the longitudinal targets for various risk factors by integrated
data at individual patient level at real-time that collaborative approach between primary and
can be seen electronically could offer several secondary care team.
advantages: 2. Education and upskilling of GPs and prac-
tice nurses specifically targeting the prac-
(a) Can be accessed and monitored remotely by tices that deliver less than satisfactory quality
specialist services (GPwSI, practice nurse of care (will be identified using Quality out-
with special interest and expertise in diabe- come Framework QoF data).
tes, diabetes specialist nurses, consultant
diabetologists);
(b) Identify trends in deterioration of a particular Methods
risk factor for an individual patient prior to
development of complications; These 12 practices were chosen from the pool of
(c) Proactive case-finding (identifying patients GP practices that participated in the ICCD study.
with worse control) before the patients pre- The Diabetes Manager system utilised the QoF
senting with symptoms; template for remotely extracting the read-codes
(d) Effective use of limited specialist resources (each risk factor, e.g., HbA1c, BP). To avoid man-
(more patients can be reviewed electronically ual input, initially the system was only able to
than face-face); extract the data from GP practices that did not use
(e) Setting individualised risk factor targets for web-based systems. The practices were chosen
individual patients, if appropriate and based on their systems. The lead specialist met all
(f) Developing machine learning techniques to the 12 practices individually to demonstrate the
spot the deterioration as well as alerts for system and firm commitment was obtained that
inappropriate prescribing (e.g., metformin in they will implement the plans provided by the
patients with deteriorating renal function, specialist for the individual patients. This agree-
prompting discontinuation of hypoglycaemic ment included a named GP and practice nurse for
agents when there is no improvement in individual practices. Once the agreement was in
HbA1c, etc.). place, it took an average of 2 weeks for the system
to be ready and up- and running. Each user was
This innovation project was funded based on given a unique user name and password. The risk
the above rationale and the potential advantages factor data were automatically uploaded every
of digital integration and remote monitoring of Sunday (so that at any given time, the data were
primary care cardiovascular risk factor data in only 6-days old). The system was fully compliant
patients with diabetes. This pilot study was with Data Protection Act 1998, as this was anony-
implemented in six general practices in Coventry mous and handled by the same company that
and six in North Warwickshire. The study utilised extracted the data for the QoF.Only patients age
a system called Diabetes Manager [42], which and sex was visible along with a unique id that can
was developed by a GPwSI, based in Norfolk. only be linked by the named GP and Practice
Details of this system, and its subsequent version, nurse in a given GP practice. A simple, user-
Eclipse, are discussed below. The study was led friendly interface was developed using these indi-
by a local diabetologist and the intervention ran vidualised risk factor data. For ease of monitoring
for a period of 3 months. this interface utilised traffic light system. Figure
Fig. 9.4(a) Diabetes Manager interphase. (b) Summary page, tabular form. (c) Single patient summary. (d) Single
patient summary, graphical
Fig. 9.4(continued)
9.4ad shows the screen shots of various graphs Individualised target setting and (c) Patient con-
that can be generated. They can be seen either as sultation as and when required and appropriate.
longitudinal trend graphs (for assessing deteriora-
tion or improvement at individual level) or as Education Workshops
tables with traffic light system (for selecting a Alongside the intervention, two half-day work-
group of patients above or below a particular tar- shops were conducted for all practice nurses and
get/threshold). Thus, for both the primary care GPs. This was tailor made for the practice nurses
team and the specialist team, it was possible to and GPs one on basic education and the second
identify the patients who require attention within on advanced diabetes care. Each workshop had a
three to four clicks after logging to the system. maximum capacity of 20 places and was con-
Although the system allows real time referral that ducted by an experienced diabetes specialist
can be sent to the specialist (and guidance back to nurse (DSN) and a clinician specialised in diabe-
the primary care team), this facility was not uti- tes. A total of ten workshops were conducted dur-
lised during this pilot study to standardise the pro- ing the 12-month period, to aid self-sustainability
cess (some GP practices do not want this to be for developing individualised cardiovascular risk
done). factor management plans (care plans) in patients
with diabetes. Each participant therefore had to
I ndividualised Risk Factor attend two half-days on two different days (a total
Management Strategy of 78 h of education).
For an individual patient, the relative deteriora-
tion of a risk factor is more important than blan- valuation oftheIntervention
E
ket threshold set by NICE.However, this is not (a) Diabetes knowledge questionnaire pre and
done in routine practice. In this proposed project, post workshops
patients were set individualised targets for their (b) Individualised risk factor modification

risk factors by careful analysis of the longitudinal change in risk factors (3 months)
values. The specialists did this in collaboration
with the patients and the GPs. Such a comprehen-
sive approach was seen as more likely to get the Results
buy in from patients and therefore result in
improvement. ardiovascular Risk Factor Data
C
The specialist proactively identified the The total number of patients with diabetes in
patients with poor cardiovascular risk factor these 12 practices was 3400. Of these 408 patients
management control in each of the 12 practices were identified as having poor control: 353 for
and devised individualised management plans for HbA1c; 222 for BP and 105 for Cholesterol. For
the patients. These patients were then put on the the purpose of this pilot study QoF targets were
watch list, which enabled the follow-up was chosen as this itself identified nearly 15% of the
done easily. Each practice had this management patients (HbA1c: >7.5%, BP: >145/85 mmHg
plan input once during the study. In addition, GPs and Cholesterol >5.0 mmol/l). Typically three-
could and did contact the specialist for asking steps were given for individualised management
doubts about specific patients. If the specialist plan for each of these patients, by the specialist to
felt the decision for the management plan the primary care team. All patients had a follow-
couldnt be implemented remotely, the primary up period of minimum 3 months (some had up to
care team was recommended to send referral to 6 months).
the relevant specialist team (including commu- The baseline characteristics of patients who
nity diabetes specialist nurses). Thus the individ- were identified as having HbA1c, BP and total
ualised risk factor management was carried out in cholesterol above the QoF threshold in these 12
the following steps: (a) Risk factor screening, (b) practices were shown in Table 9.10. Three-month
Table 9.10The baseline characteristics of patients valuable; n=176). Examples of the feedback

identified
from the GPs were: Make sense, Can we be
N Mean SD part of the programme, I will now manage
Characteristics of patients with HbA1c >7.5% patients aggressively early in the disease, and
Age years 353 66.17 10.1 We want to tap in the expertise and not be penal-
BMI kg/m2 347 31.7 6.5 ized at the same time. Example feedback from
HbA1c % 353 8.64 0.89 the practice nurses were: We never get hands on
Systolic BP mmHg 352 138.4 19.5 practice like this, Gives us confidence to man-
Diastolic BP mmHg 352 79.8 11.1 age patients I mean the practice with real
Total cholesterol mmol/L 344 4.3 1.0 patient case histories, Learnt why metformin
Characteristics of patients with BP >145/85 mmHg should be prescribed, and We can ask opinions
Age years 222 66.06 9.0
quickly especially when want to be reassured
BMI kg/m2 217 32.1 6.3
what you are thinking is right.
HbA1c % 220 8.32 0.94
Systolic BP mmHg 222 152.0 16.3
Diastolic BP mmHg 222 86.4 9.7
Discussion
Total Cholesterol mmol/L 217 4.3 0.9
Characteristics of patients with total cholesterol
>5.0 mmol/L This was one of the few projects in the UK that
Age years 105 64.45 10.7 attempted to digitally integrate the individual
BMI kg/m2 104 31.2 6.2 patient data from multiple GP practices and used
HbA1c % 105 8.55 1.27 case-finding approach to improve the cardiovas-
Systolic BP mmHg 105 136.9 18.6 cular risk factors in patients with diabetes. This
Diastolic BP mmHg 105 80.6 10.8 simple, targeted case-finding approach improved
Total cholesterol mmol/L 105 5.62 0.75 the cardiovascular risk factors significantly.
Although this innovation project did not have a
follow-up data were available for only 225/353 control group and HbA1c data were not repeated
patients with HbA1c, whilst the data for follow- at 3 months for all the patients who were identi-
up BP (n
=
222/222) and total cholesterol fied, the marked improvement observed at 3
(n=102/105) were near complete. There were months supports this model of care. This project
marked improvements in all of these risk factors: clearly showed that effective utilisation of the
HbA1c: 0.77% (0.75); Systolic BP: 13.43 routinely collected longitudinal data is possible
mmHg (20.12); Diastolic BP (13.50); and with little effort from the HCPs involved. It also
Total cholesterol: 0.56 mmol/L (0.72); all showed that such data can be monitored remotely
p<0.001. and proactively identify patients whose cardio-
vascular risk factors are not controlled (case-
Education Workshops finding). Whilst the individualised management
A total of 182 healthcare professionals (HCPs) plan by the specialist may have partly helped to
attended the ten workshops. After the first set of improve the cardiovascular risk factors, the spe-
workshop, this was opened to the whole region cialist himself believes that the proactive case-
due to overwhelming request from the partici- finding approach played a major role and must
pants. At the end of the year, 91 different HCPs become part of integrated diabetes care models.
attended these workshops twice. Of these, 57% Finally, health economic analysis was not part of
were GPs, 40% were practice nurses and 3% this pilot study. However, only a day per GP prac-
were others. tice was spent on identifying and devising the
The feedbacks on workshops were highly pos- individualist management plan (total of 12 days)
itive (4.75 out of 5 where 5 was extremely by the specialist. The cost of Diabetes Manager
was provided free of cost for the purpose of the d iseases such as diabetes and the need for proac-
pilot study and the updated version (Eclipse) tive case-finding approach. The infrastructure
approximately costs 600 per practice per year and links developed during this RCT were uti-
[43]. Therefore, this digitally integrated diabetes lised to conduct the digital integrated diabetes
care model provides an opportunity for commis- care innovation project (COMBINE for Diabetes)
sioning organisations across the UK to provide in a subset of these practices in North
high quality individualised diabetes care at Warwickshire and Coventry.
reduced cost. Indeed, subsequent to the comple- The innovation project targeted two key issues
tion of this innovation project, more than 30 in diabetes care models: digital integration of
practices in Nuneaton and Coventry (2 different individual longitudinal data and proactive moni-
clinical commissioning groups CCGs) have toring, case-finding and care-plan development
incorporated the Eclipse system as a way to for all of the patients managed by primary care.
improve the diabetes services. This project showed that such case-finding and
proactive management might be feasible with
existing resources. Such innovation, combined
Conclusion andFuture Directions with locally relevant educational workshops for
upskilling the primary care HCPs could also pro-
In this chapter, two sequential projects that were vide a sustainable long-term strategy for chronic
conducted in the Coventry and North disease such as diabetes. Success of these inte-
Warwickshire region which has a high proportion grated diabetes care models will require strategic
of small and single-handed GP practices catering leadership, adoption by the commissioners as
for multi-ethnic UK population. This region also well as the specialist care providers in a given
had one of the poorest wards in the country, with region. The innovation pilot has provided a strong
severe deprivation and higher than national aver- case for working towards such an integrated dia-
age prevalence of T2D [44]. Coventry also has a betes care service and breaks the arbitrary bound-
very high prevalence of obesity and has one of aries not only between the primary and specialist
the highest incidences of stroke rates in the UK services but also other services that cater to
[44]. Similarly, city of Leicester (part of the clus- patients with diabetes.
ter RCT) has the highest proportion of South Finally, to provide high quality care and
Asians living in a city in the UK [45]. The first reduce the cost burden of diabetes, any integrated
project was an ambitious attempt to assess the diabetes care models must prioritise prevention
clinical and cost-effectiveness of ICCDs in an and early aggressive intervention over down-
RCT setting in the UK.Although the study only stream interventions (secondary and tertiary pre-
had 64% follow-up rates, highlighting the diffi- vention). This can allow commissioning of
culties of conducting large-scale cluster RCTs, it services that extend from community prevention,
enabled objective assessment of ICCDs. Prior to screening, early diagnosis, strong foundations of
this RCT there were no such RCTs that compre- care, and education to the treatment of complica-
hensively assessed the performance of the tions and end of life care. We name this model as
ICCDs. The study highlighted the importance of Diabetes Matrix which is shown in Table 9.11.
team-change (new personnel), need for local This table summarises an integrated approach to
champions to drive improvement in chronic prevention, care, and clinical commissioning that
Table 9.11 Diabetes matrix

Level Target group Recommendations
1: Community prevention Entire local population GP, local authority, employers,
community to promote healthy
lifestyle choices
2: Prediabetic screening At risk groups within the local population GP screening for at risk individuals:
questionnaire, HbA1c%, etc.
3: Early diagnosis Prediabetic population, known impaired GP: monitors and manages those
glucose tolerance, newly diagnosed DM with IGT, IFG and newly diagnosed
diabetes
4: Forging foundations Newly diagnosed GP/specialist: individuals care-
planning, patient education and
excellent clinical care according to
current best practice and NICE
5: Rolling review 5A: Well controlled with few risk factors to GP: year of care or all main
manage clinical needs embedded within an
annual review [46]
5B: Complicated, higher risk, poorer GP or integrated/specialist: proactive
quality care case-finding and management
6: Early escalation Uncontrolled clinical and social factors at GP or integrated/specialist:
high risk of complications and admission aggressive management of difficult
to control risk factors, consider
referral or seek advice
7: Curbing complications 7A: Patients with known complications/ GP and shared care: advice sought
conditions from best local advice, consider
7B: Patients with unpredictable specialist referral
complications
8: Avoidable admissions Hypoglycaemia, DKA, foot ulceration and Specialist acute care with diabetes
infection input: Think Glucose management
to reduce length of stay [47].
Discharge to GP or shared care to
continue
9: Unavoidable admissions Patients with advanced disease and Specialist acute care with diabetes
complications input: Think Glucose management
to reduce length of stay. Usually
shared care with GP/specialist to
continue
10. Rationalised long-term Patient with co-morbidities not amenable to GP or specialist or both to rationalise
care treatment care: review clinical targets,
outcomes and medication.
Coordinate care acceptable to
patients/care. Aim for symptom free,
high quality end of life care
can be followed and implemented in a healthcare technology that can aid individualised care is
economy. It describes ten steps of integrated dia- implemented across the world for people with
betes care from community prevention to end and at risk of diabetes. We should also move
of life care with the target population at each away from short-term goals (for example: pre-
level and recommendations for both clinicians scribing cost of drugs) to long-term goals and
and commissioners. It is time that proactive strategies that can reduce the clinical and cost
preventive management incorporating latest burden of diabetes.
Acknowledgements Authors of this invited book chapter English primary care practices under a new system of
fully acknowledge all the ICCD study group members: incentives. Diabet Med. 2007;24(5):50511.
Natalie Armstrong, Darrin Baines, Richard Baker, 12. Davies MJ, Heller S, Skinner TC, Campbell MJ,

Richard Crossman, Melanie Davies, Ainsley Hardy, Carey ME, Cradock S, etal. Effectiveness of the dia-
Kamlesh Khunti, Sudhesh Kumar, Joseph Paul OHare, betes education and self management for ongoing and
Neil Raymond, Ponnusamy Saravanan, Nigel Stallard, newly diagnosed (DESMOND) programme for peo-
Ala Szczepura and Andrew Wilson. ple with newly diagnosed type 2 diabetes: cluster ran-
Dr Saravanan thanks Dr Julian Brown, Founder of domised controlled trial. BMJ.
Diabetes Manager & Eclipse systems for providing the 2008;336(7642):4915.
system free during the COMBINE for Diabetes innova- 13. Khunti K, Davies M.Glycaemic goals in patients
tion programme. with type 2 diabetes: current status, challenges and
recent advances. Diabetes Obes Metab.
2010;12:47484.
Conflict of Interest Authors declare no conflict of inter-
14. Nocon A, Rhodes PJ, Wright JP, Eastham J, Williams
est. Although Dr Ponnusamy Saravanan provided feed-
DR, Harrison SR, etal. Specialist general practitio-
back regarding the Diabetes Manager system, he was not
ners and diabetes clinics in primary care: a qualitative
paid for this service.
and descriptive evaluation. Diabet Med: JBr Diabet
Assoc. 2004;21(1):328.
15.
Halfyard C, McGowan D, Whyte M, Gayle
References C.Diabetes rapid access clinic: a bridge between pri-
mary and secondary care (Diabetes UK poster presen-
tation). Diabet Med. 2010;27 Suppl 1:137.
1. Federation ID.IDF diabetes atlas. 6th ed. Brussels:
16. Sharp PS.Glycaemic outcomes following discharge
International Diabetes Federation; 2013.
from an intermediate care diabetes clinic. Pract
2. Guariguata L.Tackling the global epidemic new
Diabetes Int. 2010;27(2):534.
estimates from the IDF diabetes atlas update for 2012.
17. Health in Wales. Cardiff CCM demonstrator model
Diabetes Voice. 2012;57(3):125.
for diabetes intermediate care clinics-progress under-
3. Ham C, York N, Sutch S, Shaw R.Hospital bed utili-
way 2010 [cited 2011 29 August]. Available from:
sation in the NHS, Kaiser Permanente, and the US
http://www.wales.nhs.uk/news/16399.
Medicare programme: analysis of routine data. BMJ.
18. Tricco AC, Ivers NM, Grimshaw JM, Moher D,

2003;327(7426):1257.
Turner L, Galipeau J, etal. Effectiveness of quality
4. Murphy E.Case management and community
improvement strategies on the management of diabe-
matrons for long term conditions. BMJ.
tes: a systematic review and meta-analysis. Lancet.
2004;329(7477):12512.
2012;379(9833):225261.
5. The NHS Information Centre. National diabetes audit
19. Shojania KG, Ranji SR, McDonald KM, Grimshaw
executive summary 20092010.
JM, Sundaram V, Rushakoff RJ, etal. Effects of qual-
6. Khunti K, Ganguli S.Who looks after people with
ity improvement strategies for type 2 diabetes on gly-
diabetes: primary or secondary care? JR Soc Med.
cemic control. JAMA JAm Med Assoc.
2000;93:1836.
2006;296(4):42740.
7. Millett C, Gray J, Saxena S, Netuveli G, Khunti K,
20. Gaede P, Vedel P, Larsen N, Jensen GV, Parving HH,
Majeed A.Ethnic disparities in diabetes management
Pedersen O.Multifactorial intervention and cardio-
and pay-for-performance in the UK: the Wandsworth
vascular disease in patients with type 2 diabetes. N
Prospective Diabetes Study. PLoS Med.
Engl JMed. 2003;348(5):38393.
2007;4(6):e191.
21. Holman RR, Paul SK, Bethel MA, Matthews DR,
8. Calvert M, Shankar A, McManus RJ, Lester H,
Neil HA. 10-year follow-up of intensive glucose con-
Freemantle N.Effect of the quality and outcomes
trol in type 2 diabetes. N Engl JMed.
framework on diabetes care in the United Kingdom:
2008;359(15):157789.
retrospective cohort study. BMJ. 2009;338:b1870.
22. Holman RR, Paul SK, Bethel MA, Neil HA, Matthews
9. Serumaga B, Ross-Degnan D, Avery AJ, Elliott RA,
DR.Long-term follow-up after tight control of blood
Majumdar SR, Zhang F, etal. Effect of pay for perfor-
pressure in type 2 diabetes. N Engl JMed.
mance on the management and outcomes of hyperten-
2008;359(15):156576.
sion in the United Kingdom: interrupted time series
23. Ray KK, Seshasai SR, Wijesuriya S, Sivakumaran R,
study. BMJ. 2011;342:d108.
Nethercott S, Preiss D, etal. Effect of intensive con-
10. Millett C, Saxena S, Ng A, Mainous 3rd A, Majeed
trol of glucose on cardiovascular outcomes and death
A.Socio-economic status, ethnicity and diabetes
in patients with diabetes mellitus: a meta-analysis of
management: an analysis of time trends using the
randomised controlled trials. Lancet.
health survey for England. JPubl Health (Oxf).
2009;373(9677):176572.
2007;29(4):4139.
24. Ahroni JH, Boyko EJ, Davignon DR, Pecoraro

11. Gulliford MC, Ashworth M, Robotham D, Mohiddin
RE.The health and functional status of veterans with
A.Achievement of metabolic targets for diabetes by
diabetes. Diabetes Care. 1994;17(4):31821.
25. Van der Does FE, De Neeling JN, Snoek FJ, Kostense quality indicators for diabetes: population based sur-
PJ, Grootenhuis PA, Bouter LM, etal. Symptoms and vey of 53,000 patients in primary care. BMJ.
well-being in relation to glycemic control in type II 2004;329(7477):12679.
diabetes. Diabetes Care. 1996;19(3):20410. 36. Kinmonth AL, Woodcock A, Griffin S, Spiegal N,
26. Papanas N, Tsapas A, Papatheodorou K, Papazoglou Campbell MJ.Randomised controlled trial of patient
D, Bekiari E, Sariganni M, etal. Glycaemic control is centred care of diabetes in general practice: impact on
correlated with well-being index (WHO-5) in subjects current wellbeing and future disease risk. BMJ.
with type 2 diabetes. Exp Clin Endocrinol Diabetes. 1998;317:12028.
2010;118(6):3647. 37. Bellary S, OHare JP, Raymond NT, Gumber A,

27. Gray A, Clarke P, Farmer A, Holman R.Implementing Mughal S, Szczepura A, etal. Enhanced diabetes care
intensive control of blood glucose concentration and to patients of south Asian ethnic origin (the
blood pressure in type 2 diabetes in England: cost UnitedKingdom Asian Diabetes Study): a
analysis (UKPDS 63). BMJ. 2002;325(7369):860. cluster
randomised controlled trial. Lancet.
28. Armstrong N, Baines D, Baker R, Crossman R,
2008;371(9626):176976.
Davies M, Hardy A, etal. A cluster randomised con- 38. Miyamoto JM, Eraker SA.Parameter estimates for a
trolled trial of the effectiveness and cost-effectiveness QALY utility model. Med Decis Making Int JSoc
of Intermediate Care Clinics for Diabetes (ICCD): Med Decis Making. 1985;5(2):191213.
study protocol for a randomised controlled trial. 39. Richardson G, Manca A.Calculation of quality

Trials. 2012;13:164. adjusted life years in the published literature: a review
29. Wilson A, OHare JP, Hardy A, Raymond N,
of methodology and transparency. Health Econ.
Szczepura A, Crossman R, etal. Evaluation of the 2004;13(12):120310.
clinical and cost effectiveness of intermediate care 40. Baker R, Chilton S, Donaldson C, Jones-Lee M,

clinics for diabetes (ICCD): a multicentre cluster ran- Lancsar E, Mason H, etal. Searchers vs surveyors in
domised controlled trial. PLoS One. 2014;9(4):e93964. estimating the monetary value of a QALY: resolving a
30. The Diabetes Trials Unit Oxford. UKPDS risk engine: nasty dilemma for NICE.Health Econ Policy Law.
home page. 2010 [cited 2010 3 February]. Available 2011;6(4):43547.
from: http://www.dtu.ox.ac.uk/riskengine/index.php. 41. Petrou S, Gray A.Economic evaluation alongside ran-
31. The Euroqol Group. EQ-5D user guide: a measure of domised controlled trials: design, conduct, analysis,
health-related quality of life developed by the EuroQol and reporting. BMJ. 2011;342:d1548.
Group. The Netherlands: EuroQol Business 42. Brown J.Diabetes manager. (http://www.diabetes-

Management; 1995. manager.org.uk). 2010.
32. Welch GW, Jacobson AM, Polonsky WH.The prob- 43. Brown J.Eclipse. (http://www.eclipsecolutions.org).
lem areas in diabetes scale. An evaluation of its clini- 2013.
cal utility. Diabetes Care. 1997;20(5):7606. 44. Health and Social Care Information Centre. (2015).
33. Gulliford MC, Naithani S, Morgan M.Continuity of Accessed at http://www.hscic.gov.uk on 19 Dec 2015.
care and intermediate outcomes of type 2 diabetes 45.
Geographies of diversity in Leicestershire;
mellitus. Fam Pract. 2007;24(3):24551. Leicestershire City Council. (2013). Accessed at
34. Drummond MF, Sculpher MJ, Torrance GW, etal. http://www.ethnicity.ac.uk/.
Methods for the economic evaluation of health care 46. Getting to Grips with the year of care: a practical
programmes. 3rd ed. Oxford: Oxford University guide. (2008). Accessed at http://www.yearofcare.
Press; 2005. co.uk on 19 Dec 2015.
35.
Hippisley-Cox J, OHanlon S, Coupland 47. Think glucose. (2011). Accessed at http://www.insti-
C.Association of deprivation, ethnicity, and sex with tute.nhs.uk/ on 19 Dec 2015.
in Germany: Triple Aim 10
in Gesundes Kinzigtal
Caroline Lang, Elisa A.M. Kern, Timo Schulte,

and Helmut Hildebrandt
Introduction required to contribute 8.2 % of their gross

income; employers pay 7.3 % [1];
The German health system is based upon a com- 2. Tax financed subsidies: in 2012 this was
pulsory, statutory health insurance (SHI). The about 4.8 % [1, pp 115] to 7.38 % [2] of the
contribution fee is based on income, although SHI income;
employees with incomes above a certain thresh- 3. Additional contribution fees, on average
old, and the self-employed, can opt out of the about 0.9 % of gross income [1]. This effec-
SHI and insure themselves privately. Contrary to tively means that parity financing has to date
the terminology, the SHI is not insurance been given up.
where premiums are risk-based but rather a
fund, into which members have to pay. This health care funding faces an enormous
Consequently, health fund would be the more increase in healthcare system expenditure as a
appropriate term. Three pillars make up the SHI result of a range of demographic changes. An
budget: analysis of the Robert Koch-Institute (RKI) from
2012 shows that more than 50 % of German peo-
1. Contributions based exclusively on income ple over 65 years suffer from at least one chronic
from gainful employment, pensions or unem- disease, approximately 50 % suffer from two to
ployment benefits, but currently not savings, four chronic diseases, and over a quarter suffer
capital gains or other forms of unearned from five or more diseases [3]. This growing rate
income. Since 2005, employees have been of chronic diseases and multi-morbidity in the
ageing population coupled with the compara-
tively high life expectancy in Germany, in a set-
ting of high-quality care standards and the
C. Lang (*) universal provision (regardless of income), a
Department of Internal Medicine, Prevention and
Care of Diabetes, University Hospital Carl Gustav broad range of medical services, medicines and
Carus at the Technische Universitt, Dresden, medical aids have contributed significantly to an
Dresden, Germany increase in Germanys public health sector
e-mail: [email protected] expenditure [46]. Currently the public sector
E.A.M. Kern covers the majority (77 %) of health expenditures
Gesundes Kinzigtal GmbH, Haslach, Germany in Germany [7]. An important driver of expendi-
ture is the provision of hospital services which is
T. Schulte H. Hildebrandt about 77.0 billion (26.04 %) in 2011 rising to
OptiMedis AG, Hamburg, Germany
e-mail: [email protected]; [email protected] 82.4 billion (26.1 %) in 2013 [8]. If we take all

170 C. Lang et al.
forms of inpatient care into account it is 36.6 % reform of the German statutory health insurance
of all expenditures, increasing to 37.6 % in 2013. (GKV) law in 2000, the establishment of medical
These recent changes in healthcare in Germany care centres ( 95 SGB V), GP-supporting care
are leading to new challenges requiring new concepts and GP-centred care ( 73b SGB V),
approaches to healthcare. Amelung summarizes implementation of Disease-Management
the following aspects, which should be consid- Programmes (DMP; 137f-g SGB V), the enact-
ered in combination for the development of new ment for Integrated Care Solution ( 140a-d SGB
forms of care [9]: V) and the care structure law
(Versorgungsstrukturgesetz; 87b SGB V). The
1. Competition regarding quality and financial traditional model of collective contracts between
contribution as a regulatory policy objective; health insurance companies and healthcare pro-
2. Opening of the healthcare market for new viders were superseded by allowing selective
providers; contracting between particular institutions [11,
3. Breaking rigid structures in healthcare sectors 12].
through new forms of care; The first approaches towards integrated
4. Meeting new challenges in healthcare caused healthcare were introduced in Germany in 2002
by chronic diseases and multi-morbidity; through Disease-Management Programmes.
5. Developing strategies against underutiliza- These structured treatment programmes were
tion, especially in regions difficult to supply; designed to ensure integrated, cross-sectoral and
6. More flexible work models for female evidence-based treatment and care for chroni-
physicians; cally ill people diagnosed with asthma, breast
7. Restructuring of care processes through tar- cancer, chronic obstructive pulmonary disease
geted incentives to promote and reward health (COPD), coronary heart disease, Type-1-diabetes
maintenance. mellitus (T1DM) and Type-2-diabetes mellitus
(T2DM). The DMPs aimed to avoid chronic dis-
According to a survey of insured patients in ease complications and associated excess hospi-
Germany, which was performed by the talization, to reduce over- and underutilization of
Commonwealth Fund in 2013, 58 % of the care, and thereby ensuring efficient care nation-
respondents perceived that the German health ally [13]. Currently there are 7,566,191 patients
system fundamentally needed to be reformed or registered in a DMP in Germany, of which
completely rebuild [10]. The current organization 3,969,019 patients are enrolled in the DMP for
of the health system is characterized by a strict T2DM. This programme was introduced in July
sectorial segregation. This makes it difficult to 2002. Participation in a DMP is voluntary and at
implement interdisciplinary, cooperative and no personal cost [14]. Although perceived to have
cross-sectoral network, efficient communication positive effects on DMP patient survival, evalua-
and information provision between inpatient-, tions has been limited and divergent [13].
outpatient-, rehabilitation care and adequate pub- Integrated care solutions have been regulated by
lic health services. This is particularly the case in law in Germany since 2004 and resulted from
the care of the increasing number of patients with reforms of the Statutory Health Insurance (SHI)
chronic diseases. To provide an efficient, effec- Modernization Act (GMG). Financial support
tive and high-quality health system in Germany, a was promoted from 2004 onwards by the stan-
redesign seems to be inevitable, but depends on dard care budget and by governmental regulated
how to reorganize the care of patients with com- start-up funding for integrated care projects up to
plex needs and diseases [5]. 2009 [15]. Despite the introduction of these mea-
In recent years, various models were initiated sures and a relatively good level of healthcare
to guarantee cross-sectoral and integrated care provision in Germany, the treatment and care of
and to facilitate more competitive health insur- chronically ill people is faced with historically
ance in Germany. Reorganization started with the evolved, system resistance that hinders optimal
10 Integrated Diabetes Care in Germany: Triple Aim in Gesundes Kinzigtal 171
integrated care. This is due to strict cross-sectoral Strategic Framework and Objectives
boundaries between outpatient and inpatient of Gesundes Kinzigtal
care, public health services and insufficient coop-
eration in care processes [5]. The main strategic framework of GK is based on
T2DM and its complications have become a the Triple Aim Approach, developed in 2008 by
growing health, social and economic burden in Berwick et al. in cooperation with the Institute
Europe and worldwide. An estimated number of for Healthcare Improvement (IHI) [28]. Berwick
56.3 million people are living with diabetes in et al. take a United States perspective, that a
Europe [16]. According to a recent RKI study responsible, sustainable and high-quality health-
(Studie zur Gesundheit Erwachsener in care system has to address [28]:
Deutschland; DEGS), 4.6 million adults
between 18 and 79 years old (7.2 %) in Germany 1. Improvement in individual experience of care;
are estimated to have been diagnosed with either 2. Improvement in population health;
T1DM, T2DM or gestational diabetes [17]. The 3. Reduction in the per capita costs of care for
lifetime-prevalence of diabetes has increased populations.
noticeably due to ageing. The total healthcare
expenditures of diabetes are currently estimated These three dimensions are displayed below (Fig.
by 30 billion per year [18, 19]. In light of these 10.2) in accordance of the Triple Aim Model of
changes and challenges, there is an urgent need Berwick et al. [28, 29].
for action, especially in chronic disease such as Derived from this approach, there are several
T2DM. It seems inevitable that meeting the objectives of GK [30]:
increasing needs of this health burden will require
optimized integration and coordination of chronic 1. Financial success with an innovative shared
care [20]. Comprehensive healthcare reforms health gain approach (see the section on
should initiate the development of integrated and Financing Model in this chapter);
coordinated care solutions, ensure good coopera- 2. Development of better organized healthcare
tion of healthcare providers and facilitate a more for the population in Kinzigtal, in cooperation
efficient approach to healthcare provision. One with the patient, the other local health partners
step towards a better healthcare system in and health insurance companies;
Germany was the integrated care initiative 3. Increasing the attractiveness of the Kinzigtal
Gesundes Kinzigtal (GK) which was launched region for the regional population through
on 1st November 2005 in Haslach, Germany. development of additional services and ensur-
ing local long-term healthcare;
4. Securing an appropriate number of providers
Gesundes Kinzigtal: A German in the area;
Approach for a Fully Integrated 5. Increasing the attractiveness of the Kinzigtal
Care System region for young health professionals in medi-
cine and increasing job satisfaction of
The healthcare network and management com- physicians;
pany Gesundes Kinzigtal Ltd described as a 6. Use of latest scientific findings for prevention
flagship-project among integrated care and treatment created in close association
approaches in Germany, is located in the affluent with all those involved in GK;
rural Kinzigtal region (population 69,000) that 7. Introducing innovations in the organization
lies in the southwest of the federal state Baden- and delivery of healthcare.
Wuerttemberg, close to Freiburg in Germany
(Fig. 10.1) [21]. The integrated healthcare system These objectives were expected to be achieved
GK was introduced here in 2005 [2227]. by improved cross-sectoral management, more
172 C. Lang et al.
Fig. 10.1 Region of Kinzigtal in Baden-Wuerttemberg, Germany
insured patients, the health insurance companies

and society. On the other hand, they offer opti-
mized leadership for healthcare in the region
[32]. Combining the knowledge and experience
of medical healthcare providers with those of
Health Sciences and Management was expected
to lead to improved cooperation and achieving
the GK objectives.
Stakeholder Involvement
in Gesundes Kinzigtal
GK consists of several organizations that cooper-

ate with each other. The shareholders of this com-
Fig. 10.2 The Triple Aim Model pany are local physicians Physicians Network
in Kinzigtal Region (MQNK) and the health sci-
ences based OptiMedis AG in Hamburg. Two
efficient cooperation by healthcare providers health insurance companies (sickness funds)
between different care sectors, a reduction in partnered with the project in 2006: AOK Baden-
morbidity, especially in chronically ill people, Wuerttemberg (AOK BW) in 2005 and the Social
and favourable conditions for purchasing exter- Security of Agriculture, Forestry and Horticulture
nal products such as medication [22, 31]. On the (SVLFG; previously LKK Baden-Wuerttemberg).
one hand, the Triple Aim dimensions form the AOK BW and SVLFG cover the less educated
basis for the actions and decision-making pro- part of the population who also experience higher
cesses of GK with the political authorities, the morbidity and are more vulnerable. A 10-year
contract was in place 20052015, in accordance Instead, the cooperation agreements within GK
with 140a-d SGB V. The contract is currently are based on a new and innovative shared health
under renegotiation for unlimited extension. gain approach for healthcare financing: shared-
About 33,000 patients, nearly half of the Kinzigtal savings contracts similar to Accountable Care
population, are insured either by the AOK BW Organization (ACO)-models in the US healthcare
(about 31,600 members) or SVLFG (about 1,400 system [2527, 32, 33]. The intention is to create
members). Since July 2006, these patients have greater efficiency (balance between expenditure
been invited to enrol in GK generally, or in spe- and health benefits) through optimizing the
cific healthcare programmes, to take advantage Kinzigtal health system. The financing of GK is
of the additional healthcare services of achieved by the two health insurance companies
GK. Approximately 30 % of insured individuals making advance payments of expected savings,
(mainly the elderly and those with greater mor- which are then invested into improvement activi-
bidity) under the two participating health insur- ties. Savings are calculated yearly in retrospect as
ance companies have enrolled in GK until the financial difference between (i) the actual
December 2014. With the exception of dental total costs of care of all patients in the region of
care, all healthcare sectors are covered by the GK Kinzigtal compared with (ii) their expected mean
company [13, 22, 23, 2527, 32, 33]. costs, derived from the German morbidity-
GK is also supported by several other partners oriented risk structure compensation system
covering many services including psychothera- (morbiditaetsorientierter Risikostrukturausgleich)
pists, physiotherapists, general physicians and and income to the health insurance companies.
specialists, paediatricians, hospitals, rehabilita- The contribution margin (, delta) of the insured
tion centres, nursing homes, nursing services, patients in Kinzigtal is the defining characteristic
pharmacies and welfare centres. Sports and cul- of the financial model, and is used as the indica-
tural clubs, gyms, podiatrists and wellness facili- tor of financial success (Fig. 10.3) [26, 27, 32,
ties provide further services for the wellbeing of 33]. Remuneration for collaborating parties and
the patient and to increase and support healthy for GK comes from lower healthcare costs for the
lifestyles [34]. The insured patients enrolled in regional population. Sustainable and increasing
GK also have additional medical time to achieve health benefits for patients is expected to be
treatment objectives and ensure intensive and achieved by GK, through patient-centred activi-
patient centred outpatient care [23]. ties, support of self-management and targeted
A Patients Advisory Committee (PAC) was prevention [27].
established to strengthen cooperation in the com- Figure 10.3 illustrates the development of the
munity and to mediate between enrolled patients, risk-adjusted expected costs in Germany (blue
the company and service providers should prob- line indexed in 2005), actual costs in GK (green
lems arise with GK or practices. PAC supports line), the surplus contribution margin (, delta)
patients with chronic disease to develop their and the number of AOK BW-insured patients
own vision for personal health, striving for goals enrolled in GK (light grey bars) from 2005 to
with the agreement of their family doctor. The 2013.
Committee consists of five patients who are par-
ticipating in GK and one ombudswoman who
supports the PAC if complaints arise [34]. Care Management of Gesundes
Kinzigtal
Financing Model of Gesundes As with other countries, there have been ques-
Kinzigtal tions over the sustainability of the financing of
the German healthcare system with the ageing
There are no incentives for risk-selection by population. The predominant type of financing
healthcare providers under the GK contract. currently is for the number of health services ren-
174 C. Lang et al.
Fig. 10.3 Development of expected costs in Germany, actual costs in GK, contribution margin and number of insured
individuals of the participating health insurance companies
dered (retrospective fee-for-service payments), Sociology at the University of Freiburg in

rather than for preventive aspects of healthcare Germany. EKIV is accountable for the provision
[23]. Associated partners of GK cooperate to of an evaluation plan, currently with four mod-
close this gap by initiating goal-setting agree- ules, which have been agreed upon from GK,
ments between physicians and patients, develop- AOK BW and SVLFG [25, 26]. The internal
ing individual treatment plans on the basis of a evaluation aims to show the effects of integrated
shared decision-making process and supporting care (among GK enrolled patients) on the dimen-
self-management, through coaching and individ- sions of the Triple Aim Approach, and to assess
ual care (especially for those with a chronic dis- whether the objectives of GK have been achieved
ease). In addition, communication and patient through a range of parameters and quality indica-
information flow is assured through a system- tors relating to, e.g., diabetes, heart failure and
wide electronic patient record. This enables all dementia [13, 32]. Central evaluation-relevant
participating partners to provide effective, effi- parameters and quality criteria for the external
cient and cross-sectoral healthcare [25]. evaluation, which are compared with conven-
tional care, include, e.g., improved patient
empowerment, patient- and care giver satisfac-
Evaluation of Gesundes Kinzigtal tion, development of patients health status, indi-
cations for over-, under- or misutilization of
Since its inception in 2005, the GK has continu- health services, interdisciplinary cooperation and
ously been evaluated: externally through an inde- economic, high-quality healthcare [25, 35].
pendent scientific research institution, and Both, the health insurance partners AOK BW
internally by the shareholding OptiMedis AG. and SVLFG and the shareholding OptiMedis AG
The external evaluation of GK is led by the in Hamburg, evaluate the financial impact of the
coordinating institution for evaluation of inte- approach. OptiMedis also provides feedback
grated care (EKIV; www.ekiv.org) which has reports for providers and performs potential anal-
been implemented by the Department of Medical yses to assess the impact of planned healthcare
Fig. 10.4 Interventions of GK in the Context of Diabetes
programmes, health-economic evaluations of programme called Healthy Weight to reduce

implemented interventions and several risk anal- the development of risk factors related to the met-
yses [32]. abolic syndrome, the development of T2DM and
cardiovascular diseases. The International
Diabetes Federation (IDF) criteria are used by
Gesundes Kinzigtal Interventions the programme to define central obesity (waist
in the Context of Diabetes circumference: men 94 cm; women 80 cm)
with at least two of the following measures [36]
GK follows a holistic approach to optimize care (Table 10.1).
for chronically ill patients with T2DM (Fig. 10.4).
A whole cascade of interventions is being offered,
depending on the risk level and the needs and co- Goals of the Healthcare Programme
morbidities of the patients. In this chapter, the Healthy Weight
development and implementation of a Kinzigtal-
specific healthcare programme Healthy Weight GK supports and motivates members to change
for at-risk individuals is described. their lifestyle through specialized comprehensive
medical care, nutrition counselling and sports
activities, with a special focus on obese patients.
Background of the Programme Reaching these targets is based on the biopsycho-
Healthy Weight: The Deadly social model, developed by Engel during the
Quartet 1970s [37]. The approach involves considering
the biological, psychological and social condi-
The interaction of different factors visceral tions involved during the development and pro-
obesity, hypertension, hyperglycaemia and gression of a (chronic) disease. The
dyslipidaemia-constitutes the metabolic syn- biopsychosocial model emphasizes the active
drome (also known as Deadly Quartet). Since role of the individual in the protection and pro-
2007, GK has offered a secondary prevention motion of their own health [37]. Supporting an
176 C. Lang et al.
Table 10.1 The IDF consensus worldwide definition of excluded: such patients can participate in the spe-
the metabolic syndrome
cialized DMP for T2DM (see Chap. 1).
Raised triglycerides 150 mg/dL (1.7 mmol/L)
Or specific treatment for this
lipid abnormality
Risk Status and Goal-Setting
Reduced high density <40 mg/dL (1.03 mmol/L) in
lipoprotein males
Agreement
(HDL-cholesterol) <50 mg/dL (1.29 mmol/L) in
females The nominated doctor fills in a risk status ques-
Or specific treatment for this tionnaire for GK members during the registration
lipid abnormality process, estimates the individual health status of
Raised blood pressure Systolic BP 130 or the patient and then invites patients to select differ-
(BP) Diastolic BP 85 mmHg ent healthcare programmes. At the same time, the
Or treatment of previously nominated doctor develops a goal-setting agree-
diagnosed hypertension ment together with the patient. This agreement is
Raised fasting plasma (FPG) 100 mg/dL (5.6 an essential tool for shared decision-making and
glucose (FPG) mmol/L)
motivation of the patient, with the intent to pro-
Or previously diagnosed
T2DM mote lifestyle changes and enhance self-manage-
If above 5.6 mmol/L or 100 ment [32]. The agreement includes definition and
mg/dL, steps to achieve individual goals along with a pre-
OGTT is strongly viously agreed date to ensure sustainability.
recommended but is not
necessary to define presence
of the syndrome
Three Standardized Programme
Modules
active patient role is a key part of the Healthy
Weight programme. The programme is based on close guidance of the
patient and a combination of three standardized
programme modules: medical care, nutrition
Enrolment in the Programme counselling and sporting activities. Actors in
Healthy Weight the programme are the patient, the nominated
doctor, the medical assistant, specialists, psy-
Healthy Weight covers a period of 15 months chologists, dietitians, sports clubs, GK and the
and can be extended for further 15 months if the participating health insurances companies.
patients fulfil predefined criteria (vide infra).
Inclusion and Exclusion Criteria Medical Care in Context of Healthy

The two main criteria to include patients in the Weight
healthcare programme Healthy Weight are reg-
istration in GK, and either a BMI of 30 kg/m2 or The 15-month-intervention Healthy Weight
waist circumference of 88 cm in women and includes regular contact with the nominated doc-
102 cm in men. A positive risk status is tor and the medical assistant through six medical
another important requirement for the participa- examinations: one during the enrolment, one
tion in this programme. The following flowchart every quarter and a final examination. At each
(Fig. 10.5) illustrates the enrolment procedure for visit, the following patient-related parameters are
Healthy Weight. Insulin treated patients are asked, measured and documented:
Fig. 10.5 Enrolment

procedure for Healthy BMI 30
Weight OR waist 88 /
102
AND
Fasting blood
glucose
between 100 -
< 100 mg/dl
125 mg/dl
AND
Familial related
diabetes risk
Measurement
HbA1c
Enrolment healthy
weight
Size and height the role of a coach and supports patients in

BMI and waist circumference achieving their individual goals (see section in
HbA1c, FPG, BP, Cholesterol, Triglycerides this chapter on Risk Status and Goal Setting
Diabetes mellitus (yes/no) Agreement). Moreover, conversations between
Insulin treatment (yes/no) physicians and patients are valuable for checking
Oral antidiabetics (yes/no) the current status and the development of the
Frequent hypoglycaemia (yes/no) programme.
Family-related diabetes risk (yes/no)
Nutrition Counselling
These parameters are evaluated regularly by GK Dietary change is an important component of
(see section in this chapter on Evaluation) and Healthy Weight. The nominated doctor strongly
used to improve and revise healthcare pro- advises patients to participate in nutritional
grammes when deemed necessary. courses and dietary consulting. GK offers, in
The medical care is also important for goal- cooperation with AOK BW, different courses and
setting and motivation of the patient. In the con- consulting services for their members and, in par-
text of Healthy Weight the physician takes over ticular, for Healthy Weight participants.
178 C. Lang et al.
One-to-One Consultation tion centre and qualified trainers. Health lectures

In cooperation with GK, AOK BW offers consul- are integrated within the sporting activities to
tations for different subjects, e.g., dietary coun- combine theory and practice. Collaboration with
selling. The GP or specialist prescribes a self-help groups is supported by GK. The exer-
prevention recommendation for a one-to-one cise programme is adjusted to the individual
consultation with a dietician. The consultation is needs of the participants. The nominated physi-
oriented towards the standards of the German cian observes the evolving patient health status
Nutrition Society (Deutsche Gesellschaft fuer during medical examinations.
Ernaehrung e.V., DGE) and contains a case his-
tory, coaching and a nutrition protocol. The ther-
apy is individually adjusted and takes four Results of Internal Data Analysis
sessions on average. On top of the one-to-one Concerning Diabetes Care
consultation, AOK BW offers online programmes in Gesundes Kinzigtal
to support their members [38].
The results presented in this chapter are mostly part
Nutritional Education in Groups of the internal evaluation. All GK healthcare pro-
Group nutritional education consists of eight units, grammes are broadly supported by different datas-
which are presented by three lecturers with differ- ets, including evaluations and feedback reports. The
ent backgrounds. Nutrition training forms the two participating SHI (AOK BW and SVLFG) pro-
basis of the course and covers four units. This part vide their regional claims data to GK, which then
is taught by a nutritionist who is a trained diabetes tasks the shareholding OptiMedis AG with data
advisor. Core learning includes food ingredients, analysis. These data are held within data warehouse
different diets, causes for overweight and purchas- architecture and used for different kinds of analyses.
ing training. Additionally, a unit with a qualified The whole GK integrated care system and most of
psychologist takes place to discuss the psychologi- its disease-specific interventions are also evaluated
cal components of obesity and poor nutrition. scientifically using the same data and supplementary
Learning is enhanced through practice sessions data from another comparable region (see section on
during two cooking evenings. To connect nutrition Evaluation in this chapter or www.ekiv.org).
and exercise, the course includes an introductory In the following section some results from the
session on Nordic walking or gymnastics exer- analyses are presented, including the prevalence
cises. This session is used to introduce the partici- of T2DM in the region of Kinzigtal, the mean
pants to the topics of sports and exercise. healthcare costs of this population and their most
common co-morbidities. Some preliminary
Sporting Activities results of the evaluation of the Healthy Weight
Another significant module of the healthcare pro- programme are then described, using a controlled
gramme involves encouraging patients to join cohort study design with matched pairs.
sports activities. The aim of Healthy Weight is
to provide ongoing courses and to integrate their
members into sports clubs and societies, where Potential Analysis of People
they become part of a social network. GK refunds with Diabetes in the Region
up to 150 of the costs to enable everyone to of Kinzigtal
participate in sports activities.
The built environment in rural areas like These analyses include inpatient and outpatient
Kinzigtal, including their lack of sports activities, data from patients with T2DM from the region of
is a problem, especially for obese patients with Kinzigtal. Patients with ZIP-codes of the region
severe diseases like T2DM. For this reason, GK and the ICD-10-GM diagnosis E11.*: Type 2
established sports courses for the target group in diabetes were selected. In 2013 (the latest year
cooperation with some sports clubs, a rehabilita- with complete claims data) the diabetes preva-
lence in the region of Kinzigtal was 9.2 % based and programme specific inclusion diagnosis
upon healthcare provider consultation coding for (T2DM, metabolic syndrome and obesity) in
2860 patients who are members of the two par- eight cost classes. People, who were not consis-
ticipating SHI. This reflects a significant growth tently insured at the time of the evaluation,
since the first year in 2006 when the prevalence including those who had died, were excluded.
was 7.0 %. The mean age of the T2DM-cohort in Each programme participant was matched in a
2013 was 71.2 years and 53.5 % were women. In ratio of 1:1 because of the limited data set. The
2013 the top 5 co-morbidities of patients with non-attenders are insured persons from the two
T2DM were essential hypertension (78.3 %), participating SHI who also live in the region of
dyslipidaemia (50.5 %), disorders of refraction Kinzigtal, but who mainly visit providers that are
and accommodation (38.2 %), back pain (33.8 %) not part of the integrated care system GK.
and obesity (33.3 %). Their top 5 hospital dis- The evaluation was not performed per calen-
charge diagnoses were heart failure (3.6 %), cere- dar year, but by number of years from enrolment.
bral infarction (1.9 %), T2DM (1.8 %), angina Controls had the same starting date as the index
pectoris (1.6 %) and atherosclerosis (1.0 %). case in the Healthy Weight programme. A time
T2DM in Kinzigtal was associated with mean period of 3 years follow-up was examined. To
expenditure of 5,935.70 per person in 2013 avoid bias, the latest date for enrolment was 31st
(not necessarily only for diabetes care) including of December 2010 because 2013 was the latest
40 % from inpatient stays, 24 % from drug pre- year with complete data. Analyses involved com-
scriptions, 19 % from physician remuneration in paring the case-control difference before and
ambulatory care and the rest from remedies and after the intervention.
adjuvants (e.g., insulin pen systems, wheelchairs, Of 149 individuals enrolled up to 31st of
physiotherapy, etc.), work incapacity or rehabili- December 2010, 136 (91 %) had a matched con-
tation. In 2013, the net mean loss for the two SHI trol. The small numbers allow only preliminary
from all diabetes patients in the Kinzigtal region insights into the achievements of the programme
was 172.00 per patient; however, it already to date. The mean ages of cases and controls were
improved by +299.20 per patient per year com- 56.3 and 56.4 years respectively. In both groups
pared to the initial year of the integrated care 106 individuals (78 %) were female. One year
project in 2006. The normal improvement rate for before the start (baseline) of the programme
all insured persons in the same period of time Healthy Weight about 80 % of cases and con-
was +21.40 per year. To put these results into trols had diabetes, about 93 % obesity and about
perspective it has to be considered that in 2009 60 % had dyslipidaemia.
the German risk adjustment scheme has been Figure 10.6 shows the hospitalization rates in
changed for all patients in any SHI, allocating the two groups. While 24 patients (17.7 %) from
more money to patients having specific diseases the Healthy Weight participants had been hos-
including (among others) T2DM, so that part of pitalized (from any cause) in the year before
the improvement of the contribution margin of enrolment, this was the case for only 12 control
the T2DM population is system-based, which is individuals (8.8 %). All-cause hospitalization
why more detailed evaluations are performed in rates were similar in the first year, but continued
GK concerning the intervention participants. to decrease to 16 patients (11.8 %) in the third
year of follow-up in cases, while rates remaining
increased in controls. In a difference-in-difference
First Results from the Internal analysis these reverse trends led to 16 less cases
Evaluation of the Healthy Weight in the Healthy Weight group compared to the
Programme control group in the third year.
A comparison of the total annual difference in
The GK- Healthy Weight programme is con- expenditure reveals a slower growth in the
tinuously evaluated using a controlled cohort Healthy Weight cohort over the 3 years of fol-
study design with an exact matching of age, sex low-up resulting in 149.4 less expenditure per
180 C. Lang et al.
Fig. 10.6 Development of patients with hospital stay in relation to the enrolment date
capita and a difference-in-difference of 659.7 address all dimensions of the Triple Aim (care,
per capita in the third year (Fig. 10.7). health and cost) and to commit all stakeholders to
Furthermore, the mean number of days off a process of healthcare delivery that target these
work for sickness were over 50 % lower among dimensions as a whole. After 10 years of innova-
participants than controls, particularly in the third tive healthcare practice, the management com-
year of follow-up (Fig. 10.8). pany GK now receives more in income than it
Results of the external evaluation supplement spends. Up to 2011, all three dimensions of the
the results of the internal evaluation, although Triple Aim Approach have developed positively
they have a longer time lag and take longer to within GK including its complex sub-
perform due to database size and methodological interventions [41]. However, further studies are
issues. The most recent evaluation in May 2015 necessary before evidence of sustained success
supported the internal evaluation and demon- by GK can be described as proven [13, 20]. A
strated another improvement in diabetes care. In critical success factor, already identified by the
2011 only 2.3 % of the GK-enrolled diabetes holistic, public health-related approach of GK, is
patients were admitted with diabetes compared the long lasting integrated care contract that is
to 4.0 % among not-GK-enrolled patients. based on mutual trust between GK and the two
Statistical significance is barely missed, which is health insurance companies. This facilitates
shown in Table 10.2 (in accordance with [39]). investments in sustainable interventions with
their long-term benefits, the support of a cultural
change among physicians and patients, and the
Outlook use of instruments like patient empowerment,
shared-decision-making and coaching for the
Against the background of rapidly increasing self-management of chronic conditions.
chronic diseases and a growing burden for Meanwhile, an ongoing sharing and analysing of
patients to manage their disease, innovative data helps to identify strengths and weaknesses
approaches and holistic, patient-centred interven- of interventions much quicker than in usual prac-
tions are needed that fit into the realities of the tice, and enables timely refinements of existing
daily lives of patients [40]. It is important to programmes.
Fig. 10.7 Total cost

difference of Healthy
Weight-participants: matched
pairs
Fig. 10.8 Mean number of days off work for sickness since enrolment
Table 10.2 Comparison of enrolled (GK) and not- GK has now made a step abroad. In 2015,
enrolled (Not GK) insured individuals of the AOK BW in the Dutch subsidiary OptiMedis Nederland
the Kinzigtal region with diabetes and hospitalization due
B.V. based in Leiden was founded. Following
to diabetes (E10E14)
the successful approach of their colleagues in
GK Not GK Odds
Haslach in Germany, the next challenge is to
Year % % Overall % ratio 95 %-CI
improve the healthcare of the local population of
2006 2.8 2.8 2.8 1.01 0.581.75
40,000 inhabitants of Nijkerk in the Netherlands
2007 2.3 3.5 3.1 0.61 0.351.08
2008 3.1 4.8 4.2 0.65 0.421.03
[42]. Discussions with other health insurance
2009 2.0 3.4 2.9 0.59 0.351.01
companies are also currently taking place, in
2010 2.7 3.5 3.2 0.76 0.471.23
order to give other patients the opportunity to use
2011 2.3 4.0 3.3 0.62 0.381.02 the GK services for their healthcare and wellbe-
ing [34]. A GK study revealed that almost all
182 C. Lang et al.
respondents (92.1 %) would be willing to recom- GP General practitioner

mend the GK healthcare programme [32]. We HDL High density lipoprotein
feel that an integrated care system like GK can be ICD-10-GM International classification of
beneficial within the current healthcare system in diseases 10th revision, German
Germany, especially for people with chronic con- modification
ditions, through its systematic use of its popula- IDF International diabetes
tion health management approaches to optimize federation
the quality of care. The current deficiencies in the IHI Institute for healthcare
regular healthcare system to address the Triple improvement
LKK Agricultural Health Insurance
Aim goals adequately should strengthen move-
Company [Landwirtschaftliche
ment towards more intelligent solutions such as
Krankenkasse]
the GK programme.
LTD Limited Company
MQNK Physicians Network
in Kinzigtal Region
Abbreviations [Medizinisches QualitaetsNetz
Kinzigtal e.V.]
AG Incorporated Company OECD Organization for economic
[Aktiengesellschaft] cooperation and development
ACO Accountable Care Organization OGTT Oral glucose tolerance test
AOK BW General Local Health PAC Patients advisory committee
Insurance Company in RKI Robert-Koch-Institute
Baden-Wuerttemberg SGB V Book five of Germanys social
[Allgemeine Ortskrankenkasse security code
Baden-Wuerttemberg] SHI Statutory health insurance
BMI Body mass index SVLFG Social security of agricul-
BP Blood pressure ture forestry and horticul-
CI Confidence interval ture [Sozialversicherung fuer
COPD Chronic obstructive pulmonary Landwirtschaft, Forsten und
disease Gartenbau]
DEGS German Health Interview T1DM Type-1-Diabetes Mellitus
and Examination Survey for T2DM Type-2-Diabetes Mellitus
Adults [Studie zur Gesundheit US United States
Erwachsener in Deutschland] WHO World Health Organization
DGE German Nutrition Society ZIP-code Zone improvement plan-code
[Deutsche Gesellschaft fuer [Postleitzahl]
Ernaehrung]
DMP Disease Management
Programme
References
EKIV Coordinating Institution for
Evaluation of Integrated 1. Busse R, Blmel M. Germany: health system review.
Care [Evaluations- Health Syst Transit. 2014;16(2):1296. xxi.
Koordinierungsstelle Integrierte 2. Rosenbrock R, Gerlinger T. Gesundheitspolitik eine
Versorgung] systematische Einfhrung. 3rd ed. Bern: Huber; 2014.
p. 569. vollst. berarb. Aufl. ed.
FPG Fasting plasma glucose
3. Nowossadeck E. Demografische Alterung und Folgen
GK Gesundes Kinzigtal Ltd. fr das Gesundheitswesen. Berlin: Robert-Koch-
GKV German statutory health insurance Institut; 2012.
[Gesetzliche Krankenversicherung] 4. Global Health Observatory. Life expectancy
GMG Health modernization act [Internet]. [cited 08/05/2015]. Available from: http://
www.who.int/gho/mortality_burden_disease/life_
[Gesundheits-Modernisierungs-
tables/situation_trends/en/.
Gesetz]
5. Amelung V, Hildebrandt H, Wolf S. Integrated care in 17. Kurth B. Erste Ergebnisse aus der Studie zur
Germany-a stony but necessary road! Int J Integr Gesundheit Erwachsener in Deutschland (DEGS).
Care. 2012;12, e16. Bundesgesundheitsbl. 2012;55(8):98090.
6. Mladovsky P, Allin S, Masseria C, Hernndez- 18. Hall M, Felton A, Henrichs H, Barnett A. Diabetes
Quevedo C, McDaid D, Mossialos E. Health in the the policy puzzle: towards benchmaking in the EU 25.
European Union: trends and analysis. 2010: WHO International Diabetes Federation European Region
Regional Office Europe; 2010. and Federation of European Nurses in Diabetes.
7. OECD. How does Germany compare? OECD Health 2005; 103.
Statistics 2014 [Internet]. [cited 10/04/2015]. 19. Heidemann C, Du Y, Schubert I, Rathmann W,
Available from http://www.oecd.org/els/health- Scheidt-Nave C. Prvalenz und zeitliche Entwicklung
systems/Briefing-Note-GERMANY-2014.pdf. des bekannten Diabetes mellitus. Bundesgesundheitsbl.
8. Gesundheitsausgaben nach Einrichtungen [Internet]. 2013;56(56):66877.
[cited 08/05/2015]. Available from: https://www.des- 20. Hildebrandt H. Crossing the boundaries from indi-
tatis.de/DE/ZahlenFakten/GesellschaftStaat/ vidual medical care to regional public health out-
G e s u n d h e i t / G e s u n d h e i t s a u s g a b e n / Ta b e l l e n / comes: the triple aim of Gesundes Kinzigtal- better
Einrichtungen.html. health+ improved care+ affordable costs. Int J Integr
9. Amelung V. Neue Versorgungsformen auf dem Care. 2014; 14(5).
Prfstand. In: Amelung V, Eble S, Hildebrandt H, edi- 21. Seiler R. Die Schwarzwaldformel-das Modell
tors. Innovatives Versorgungsmanagement: Neue Gesundes Kinzigtal. In: Weatherly J, Seiler R,
Versorgungsformen auf dem Prfstand. Berlin: Meyer-Lutterloh K, Schmid E, Lgel R, Amelung V,
Medizinisch Wissenschaftliche Verlagsgesellschaft; editors. Leuchtturmprojekte Integrierter Versorgung
2011. p. 316. und Medizinische Versorgungszentren. Belrin:
10. Osborn R, Schoen C. The Commonwealth Fund 2013 Medizinisch Wissenschaftliche Verlagsgesellschaft;
international health policy survey in eleven countries. 2007. p. 13947.
Consult le [Internet]. [cited 21/05/2015]. Available 22. Siegel A, Stel U, Gener D, Knittel R. Kooperation
from: http://www.commonwealthfund.org/~/media/ und Wettbewerb im integrierten Versorgungssystem
files/publications/in-the-literature/2013/nov/pdf_ Gesundes Kinzigtal. In: Amelung V, Sydow J,
schoen_2013_ihp_survey_chartpack_final.pdf. Windeler A, editors. Vernetzung im Gesundheitswesen-
11. Sachverstndigenrat fr die Konzertierte Aktion im Wettbewerb und Kooperation. Stuttgart:
Gesundheitswesen. Bedarfsgerechtigkeit und W. Kohlhammer; 2009. p. 14555.
Wirtschaftlichkeit: Band III: ber-, Unter-und 23. Hildebrandt H, Michalek H, Roth M. Integriertes
Fehlversorgung. Ausfhrliche Zusammenfassung. Management auf Augenhhe Anforderungen an
Gutachten 2000/2001 [Internet]. [cited 25/04/2015]. eine konsistente Fhrungsphilosophie in IV-Systemen
Available from: http://www.svr-gesundheit.de/ am Beispiel von Gesundes Kinzigtal. In: Amelung V,
fileadmin/user_upload/Gutachten/2000-2001/ Eble S, Hildebrandt H, editors. Innovatives
Kurzf-de-01.pdf. Versorgungsmanagement: Neue Versorgungsformen
12. Sachverstndigenrat zur Begutachtung der auf dem Prfstand. Berlin: Medizinisch
Entwicklung im Gesundheitswesen. Koordination Wissenschaftliche Verlagsgesellschaft; 2011.
und Integration. Gesundheitsversorgung in einer p. 21526.
Gesellschaft des lngeren Lebens. Sondergutachten 24. Siegel A, Kster I, Schubert I, Stel U. Integrierte
2009 [Internet].[cited 25/04/2015]. Available from: Versorgung Gesundes Kinzigtal: Ein Modell fr
http://www.svr-gesundheit.de/index/php?id=196. regionale Prvention und Schnittstellenoptimierung.
13. Fuchs S, Henschke C, Blmel M, Busse R. Disease 2012. p. 14864.
management programs for type 2 diabetes in Germany. 25. Hildebrandt H, Schulte T, Stunder B. Triple aim in
A systematic literature review evaluating effective- Kinzigtal, Germany. J Integr Care.
ness. Dtsch Arztebl Int. 2014;111:45363. 2012;20(4):20522.
14. Bundesversicherungsamt Zulassungsstand. 2014 26. Hlzel LP, Vollmer M, Kriston L, Siegel A, Hrter
[Internet]. [cited 16/04/2015]. Available from: http:// M. Patientenbeteiligung bei medizinischen
www.bundesversicherungsamt.de/weitere-themen/ Entscheidungen in der Integrierten Versorgung
disease-management-programme/zulassung-disease- Gesundes Kinzigtal: Ergebnisse einer kontrollierten
management-programme-dmp.html#c207. Kohortenstudie. Bundesgesundheitsbl.
15. Rschmann H, Roth A, Krauss C. Vernetzte Praxen 2012;55(1112):152433.
auf dem Weg zu managed care? Aufbau-Ergebnisse- 27. Hildebrandt H, Schmitt G, Roth M, Stunder
Zukunftsvision. Springer; 2013. B. Integrierte regionale Versorgung in der Praxis: Ein
16. Aguiree F, Brown A, Cho NH, Dahlquist G, Dodd S, Werkstattbericht aus dem Gesunden Kinzigtal. Z
Dunning T. IDF diabetes atlas: International Diabetes Evidenz Fortbild Qualitt Gesundheitswesen.
Federation; 2013. 2011;105(8):5859.
184 C. Lang et al.
28. Berwick DM, Nolan TW, Whittington J. The triple Konzeption, Herausforderungen,
aim: care, health, and cost. Health Affair. Lsungsmglichkeiten. In: Amelung V, Eble S,
2008;27(3):75969. Hildebrandt H, editors. Innovatives
29. Institute for Healthcare Improvement. The IHI triple Versorgungsmanagement: Neue Versorgungsformen
aim. Available from: http://www.ihi.org/Engage/ auf dem Prfstand. Berlin: Medizinisch
Initiatives/TripleAim/pages/default.aspx. Wissenschaftliche Verlagsgesellschaft; 2011.
30. Hermann C, Hildebrandt H, Richter-Reichhelm M, p. 14555.
Schwartz F, Witzenrath W. Managementgesellschaft 36. International Diabetes Federation. The IDF consensus
organisiert Integrierte Versorgung einer definierten worldwide definition of the metabolic syndrome.
Population auf Basis eines Einsparcontractings. Available from: http://www.idf.org/webdata/docs/
Gesundheits- und Sozialpolitik. 2006. p. 1129. MetS_def_update2006.pdf.
31. Siegel A, Stel U, Gaiser K, Hildebrandt 37. Pauls H. Das biopsychosoziale Modell: Herkunft und
H. Integrierte Vollversorgungssysteme und soziale Aktualitt. 2013;1(1): 17.
Ungleichheit-das Beispiel Gesundes Kinzigtal. 38. Allgemeine Ortskrankenkasse. Einzelgesprche
Public Health Forum. 2008;16(2):e14. (One-to-one consultations). In: Kern E, editor.
32. Hildebrandt H, Pimperl A, Schulte T, Hermann C, Healthy Weight; 2015.
Riedel H, Schubert I. Triple aim Evaluation in der 39. Kster I, Ihle P, Schubert I. Evaluationsbericht 2004
Integrierten Versorgung Gesundes Kinzigtal 2011 fr Gesundes Kinzigtal GmbH. AOK-Daten.
Gesundheitszustand, Versorgungserleben und Kln: 2015. p. 244.
Wirtschaftlichkeit. Bundesgesundheitsbl. 2014. 40. Leppin AL, Montori VM, Gionfriddo MR, editors.
p. 110. Minimally disruptive medicine: a pragmatically com-
33. Hildebrandt H, Hermann C, Knittel R, Richter- prehensive model for delivering care to patients with
Reichhelm M, Siegel A, Witzenrath W. Gesundes multiple chronic conditions. Healthcare.
Kinzigtal Integrated Care: improving population Multidisciplinary Digital Publishing Institute; 2015.
health by a shared health gain approach and a shared 41. Busse R, Stahl J. Integrated care experiences and out-
savings contract. Int J Integr Care. 2010;10:e046. comes in Germany, the Netherlands, and England.
34. Gesundes Kinzigtal im berblick. Available from: Health Aff (Project Hope). 2014;33(9):154958.
http://www.gesundes-kinzigtal.de/die-idee-von- 42. Ansatz von Gesundes Kinzigtal in Holland hoch im
gesundes-kinzigtal/. Kurs: Gesundes Kinzigtal. Available from: http://
35. Siegel A, Kster I, Schubert I, Stel U. Evaluation www.gesundes-kinzigtal.de/news/een-kopie-van-
der Integrierten Versorgung Gesundes Kinzigtal: gesundes-kinzigtal-en-nederland-2014.html.
Diabetes Care in the Netherlands 11
Harold W. de Valk and Helmut Wenzel
Background to the Dutch Health tal care, varying quality of care between provid-
Care ers, value for the money, with rising
expenditures and an ageing society. Health
According to the 2014 Euro Health Consumer expenditures reportedly grew by 67 % per year
Index (ECHI) the Netherlands is the best country 20072009, with data from the Organisation for
in Europe to live for health care. In a ranking of Economic Co-operation and Development
37 countries the Netherlands was top with a score (OECD) demonstrating that the growth in health
of 898 out of 1000. Switzerland was second, fol- expenditure was above the OECD average. Their
lowed by Norway, Finland and Denmark. The analysis indicated that this growth was due to
UK, excluding Scotland, landed in 14th place increasing volumes of care. Whereas prices
(718 points) with Spain 19th, Italy 22nd, and increased on average by 1.6 % per year, the vol-
Germany in 9th place [1]. By 2014, the ume of services from Dutch hospitals grew by
Commonwealth Fund placed Netherlands 5th 4.2 % per year, inpatient admissions by 3 %
(tied with Germany) out of 11 countries: ranked annually and day-patient admissions by 10 %. In
second in timeliness of care, but ranked 7th8th order to get a more complete picture they stated
in safety, efficiency and equity and 10th (before that the volumes for outpatient care grew sub-
the USA) in per capita cost [2]. stantially (5.5 % on average per year), while the
A survey from 2010 from the Dutch Ministry price of medicines had fallen significantly even
of Health comparing 125 performance indica- though the number of prescriptions had increased
tors across several countries, drew a more precise (about 15 % in 2008) [3].
picture of getting access, varying quality, and The Netherlands has the highest per capita
rising costs [3]. Challenges that were revealed, spending in Europe [1]. In order to evaluate the
dealt with timely access to ambulatory and hospi- affordability of health-care financing, expendi-
tures are set in relation to the economic perfor-
mance of the national economy the Gross
Domestic Product (GDP). Health expenditure, as
H.W. de Valk (*) a percentage of GDP, increased from 7.4 % in
Internist-endocrinologist, Department of Internal 1980 to 11.9 % in 2011 [4, 5]: a share which is
Medicine, University Medical Center Utrecht, higher only in the US health-care system (18 % of
Utrecht, The Netherlands GDP). In the Netherlands, 1.7 % of the expen-
ditures are privately, and 10.2 % publicly, financed
H. Wenzel in 2011. In the US the corresponding numbers
Health Economist, Konstanz, Germany
e-mail: [email protected] are 89.1 % [4]. Experts expect a dramatic

186 H.W. de Valk and H. Wenzel
increase in expenditure so that by 2040, one quar- the public programme (known as the fund for
ter of GDP will be needed to provide care [6]. the sick) and stronger expenditure control was
Ageing of the population, medical and therapeutic required [12]. The reform that was introduced
progress and global economic distortions have was in response to a number of problems: a two-
made it necessary to assess whether the Dutch tier system of private health insurance for the
health system organisational principles will be people with a good income (approximately one
able to meet future challenges. third of the population at that time) and state cov-
erage for the rest; an inefficient and complex
bureaucracy; lengthy waiting lists and a lack of
Growing Burden of Diabetes patient-focus [13]. The Dutch health-care system
was, and still is, made up of three branches, so-
Diabetes prevalence in 2014 was about 7.24 % [7] called compartments. The first compartment of
and it is expected to increase to 9.5 % by the year care emphasises care and support for those peo-
2035 [8]. Diabetes is an expensive disease, for ple who have to cope with irreversible damage to
example, the American Diabetes Association physical or mental integrity [12]; the second
observed that the medical expenditures of people compartment of care focuses on recovery and
with diabetes, on average, are approximately 2.3 includes hospital care and visits to a primary care
times higher than what expenditures would be in physician. The third compartment of care is
the absence of diabetes [9] (p1). In a disease defined as luxury care, such as cosmetic sur-
model Zhang et al. [10] calculated the burden of gery [13]. Care for conditions covered by the
disease. They reported that globally, 12 % of first compartment is given regardless of an indi-
health expenditures and USD 1330 (Diabetes viduals financial situation and is regulated by the
induced expenditures (ID) 1478) per person were Algemene Wet Bijzondere Ziektekosten
spent on diabetes in 2010. The expenditure varies (AWBZ), or Exceptional Medical Expenses
by region, age group, gender, and countrys income Act. Both before and after the reforms, contribu-
level. Looking at Europe, the Netherlands was in tions to this fund were taken from income-related
the top 7 countries with expenditures of salary deductions, supplemented by a general
3,793,953,000 USD (the underlying assumption in government revenue grant [13]. The AWBZ was
these analyses was that a patient with diabetes is and continues to be applicable to all Dutch citi-
twice as expensive as a comparable person without zens. Before 2006, the provision and funding of
diabetes: this is conservative). Furthermore, IDF insurance for second and third compartment care
estimates expenditure of about 4113 USD per per- were determined by an individuals total income.
son with diabetes (Fig. 11.1), and this was expected In 2005, the wage ceiling was set at a gross
to rise to 6943.11 USD by 2014 [7]. This would annual income of 33,000 ($40,600) for employ-
mean that the Dutch performance is within the ten ees and 21,050 ($25,900) for the self-employed.
most expensive countries with health expenditure Those people earning less were determined eli-
rising to 4,311,488 USD by 2035 [10]. gible for the public system. Those who did not
qualify for the public system, could purchase pri-
vate insurance to cover potential short-term med-
Basic Principles of the Dutch ical needs in the second compartment and in
System some cases also for luxury care in the third
compartment [12]. However, the way in which
Prior to the health-care reforms of 2006, Dutch the provision and financing of the first and third
health care was characterised by extensive gov- compartment were organised did not change. The
ernment regulation and a dual insurance system main changes occurred to care covered by the
of public and private insurance, which had been second compartment.
perceived to be inefficient. By 2005, roughly Preceding the 2006 reforms, the second com-
two-thirds of the Dutch population had entered partment combined Social Health Insurance
11 Approaches to Integrated Diabetes Care in the Netherlands 187
Fig. 11.1 Diabetes induced expenditures [ID] per person and year in Europe (Data are from [10], the classification
limits are based on [11])
(SHI) the so-called Fund for the sick (ZFW) ernment, it is now the private health insurance
with a Private Health Insurance (PHI) scheme. market which is responsible for providing the
SHI was compulsory for people below a certain basic package of health insurance to all Dutch
income, funded through payroll contributions citizens. Extra government finance schemes
and managed by the government. The amount ensure that universality of care is maintained, no
paid by each individual was unaffected by their matter what your income, as well as providing a
medical situation. Resources were paid into a safety net for illegal immigrants [13].
Central Sickness Fund which provided a mech- The original idea was to limit costs by stimu-
anism for redistributing funds to compensate lating competition between the rival insurance
insurers for those considered high risk. Along companies. But with ongoing mergers of compa-
with the ABZW, the Fund allowed universal med- nies, there seems to be limited competition.
ical coverage. PHI was funded by employers or However, critics point out that large health insur-
individuals with higher incomes and insurers ance companies are said to squeeze health pro-
were allowed to take the risk of an individual into viders in order to lower their expenditures [14].
account, meaning that premiums varied widely. Adjustments to the Health Insurance Act cur-
The 2006 Dutch Healthcare Act (ZvW) over- rently being promoted will prevent patients from
turned the division between SHI and PHI in the choosing their own medical specialist. Presently
second compartment, thus creating a universally insurers have to reimburse a certain amount to
compulsory Social health Insurance scheme. patients who go to a specialist or facility that is
Instead of being managed primarily by the gov- not partnered with the insurance company,
offsetting at least part of the cost of treatment for the government should stimulate competition
the patient [15]. Furthermore, the monthly pre- rather than regulate the supply of health care:
mium for Dutch health insurance will rise by making the Dutch system the most extreme appli-
around 9.5 in 2015. This means people will cation of market mechanisms to stimulate effi-
spend roughly 114 extra per year, taking the ciency in a European health-care system.
annual cost of basic health insurance to 1215 in In order to achieve this, the Dutch came up
2015 [16]. Moreover, the amount of money for with a system of managed competition that
mandatory excess deductibles (eigen risico) included a statutory general insurance provision
increased from 360 to 385 per year in 2016 [12]. The basic concept demands that every
[17]. The eigen risico zorgverzekering or own Dutch citizen has to buy health-care and pharma-
risk insurance is the amount which an individual ceutical insurance from one of several private
has to pay out of pocket before health insurance providers [19]. The extent of coverage under
coverage sets in [16]. Some insurance companies these policies is government-mandated and iden-
offer larger excess deductibles (up to 900) com- tical, including a deductible, depending on the
bined with a lower annual standard price. specific insurance policy. This means that the
You are better off when you do not need insured patient has to pay additional expenses,
your deductible but when you do, you are worse ranging from the governmental fixed mandatory
off financially, This depends on your own calcu- amount of 385 in 2016 per year [17] to any rea-
lation of personal risk. sonably calculated amount balancing the contri-
The Primary Care Provider (PCP) plays the bution fee against the expected expenses [20].
leading role in providing care, acting as gate- Insurers must also charge the same premium to
keeper and the first point of contact (except in all, including those with pre-existing conditions.
emergencies). Every Dutch person has to register The only exception is that group discounts (e.g.,
with a primary care provider (PCP) [4]. Patients for an employer) are permitted [18] where per-
must obtain a PCP referral prior to a specialist sons are collectively insured. This could mean
visit, except for acute conditions such as trauma that the insured person of such a group can profit
or acute myocardial infarction. Nevertheless, this from a broader package at lower premium and at
also depends on the insurance package; with lower own risk. A special payroll tax also funds
more expensive policies, no referral is needed. the governments health regulator, which pro-
Nurse practitioners are employed to perform vides insurers with payments to help pay expenses
check-ups on the chronically ill. PCPs can deal related to high cost policy holders. Basically,
with routine health issues, perform standard gyn- health care embraces three overlapping markets
aecological and paediatric examinations, and as Fig. 11.2 shows: the acquisition of insurance
refer onto other services [13]. Most specialists contracts between individuals and insurers; the
work within a hospital setting. provision of health-care services between indi-
viduals and providers and between insurers and
providers for the pricing of those services [18].
Managed Care in the Netherlands: Over the years many approaches were intro-
Integrated Chronic Care duced to improve the quality and continuity of
and Bundled Payments care for chronic diseases. However, fragmented
funding made it difficult to establish long-term
The reform of the Dutch health-care system has programmes [21]. Therefore, the Dutch minister
been characterised as managed competition, or as of health approved, in 2007, the introduction of
an experiment in how far you can get with a bundled-care (known is the Netherlands as
system in which there is almost no direct govern- a chain-of-care) approach for integrated chronic
ment involvement [18]. The Government exe- care, with special attention to diabetes. This
cutes its responsibility indirectly, only. The bundled-payment approach was firstly introduced
leading principle of this reform (theory) was that on an experimental basis, accepted in 2010 and
Fig. 11.2 Medical specialist are rarely part of the Care health-care market. CG Care Group, GP General
Group and only provide treatment advices or suggestions Practitioner, ProVn Health-Care Provider, LAB Laboratory
without actually treating the patients themselves. Dutch (From Struijs [23])
subsequently implemented nationwide for diabe- by the care group and providers, are freely nego-
tes, chronic obstructive pulmonary disease tiated [21]. As Struijs et al. [21] point out, the
(COPD), and cardiovascular risk management aims of these care groups are similar to those of
[21]. Insurers negotiate and pay a single remu- Accountable Care Organizations (ACOs), as
neration [21] (lump sum) to a principal contractor currently designed in the United States or
(the care group) to cover a full range of care Clinical Commissioning Groups in the UK
services for specific chronic diseases, like diabe- [24]. However, there are some essential differ-
tes, COPD, or vascular diseases for a fixed period. ences: first, care groups (as with clinical commis-
Care groups (CG) are new legal entities which are sioning groups) are dominated by GPs, whereas
formed by health-care providers at local levels on ACOs may comprise a wide range of providers,
a regional scale [22]. Very often they are general at least primary care physicians, specialists, and
practitioners (GPs). As a principal contractor they one or more hospitals; second, patients are to be
negotiate with the insurers on price and products. assigned to ACOs on the basis of their patterns of
Finally, the care group takes on both clinical and service use, whereas patients here are assigned to
financial responsibility for all assigned patients in a care group on the basis of their disease (e.g.,
the particular diabetes care programme. The care onset of diabetes). Moreover, the care groups
group either delivers services itself or subcon- bear the full financial risk for the expenditures of
tracts to other care providers [21]. care [21], whereas ACOs will not take over the
With the bundled-payment approach, the mar- risk of higher expenditures than expected [25].
ket is divided into two segments: one in which
health insurance companies contract care from
the principal contractors (i.e., care groups) and Integrated Diabetes Care
one in which care groups conclude service con-
tracts from individual providers [23]. These pro- With a bundled payment approach or episode-
viders could be general practitioners, specialists, based payment multiple providers are reim-
dietitians, or laboratories. Both, the price for the bursed a single sum of money for all services
bundle of services by insurers and care groups, related to an episode of care (e.g., hospitalisa-
and the fees for the subcontracted care providers tion, including a period of post-acute care). This
is in contrast to a reimbursement for each indi- office space may also be included; nevertheless,
vidual service (fee-for-service), and it is these are difficult to budget under the existing
expected that this will reduce the volume of ser- bundled health-care model [28].
vices provided and consequently lead to a reduc-
tion in spending. Since in a fee-for-service
system the reimbursement is directly related to Organisation and Coverage of Care
the volume of services provided, there is little
incentive to reduce unnecessary care. The bun- Care groups are a core element of the bundled pay-
dled payment approach promotes a more effi- ment approach. Struijs et al. [28] outline the role of
cient use of services [26]. the principal contractor of the bundled payment
For example, the Washington State Hospital scheme in such a way that the groups are legally or
Association [27] identified three areas where contractually responsible for the coordination,
bundled payments should show progress: (1) consistency and quality of the diabetes care. In
Quality improvement and cost reduction by compliance with this role they can either contract
reducing administrative/overhead costs, sharing or coordinate health-care providers for the actual
risk, eliminating cost-shifting, outcomes man- provision of the specified health-care services or
agement and continuous quality improvement, they even provide certain or all of the care compo-
reducing inappropriate and unnecessary resource nents themselves. To ensure the required quality
use, efficient use of capital and technology; (2) and efficiency of care they have the option to
consumer responsiveness, i.e., seamless contin- selectively contract health-care providers.
uum of care and focus on the health of enrollees; The coverage of care offered by groups based
(3) community benefit by improving community on a bundled payment scheme is based on stan-
health status, and addressing the prevention of dards of care (CS). These standards are defined
social issues which affect community health. by the Dutch Diabetes Federation (NDF), build
Most integrated networks include a team-based on evidence-based guidelines and are updated
approach, as well as an emphasis on patient regularly [32]. However, a care group may have a
participation. specialist for internal medicine under contract for
Furthermore, with the set-up of a bundled pay- consultations. If such a specialist is consulted, an
ment model, it is reasonable to expect that multi- outpatient hospital treatment bundle for diabetes
disciplinary cooperation between health-care mellitus without secondary complications may
providers will be facilitated insofar as existing not be claimed. As soon as the treatment respon-
financial barriers between care sectors and disci- sibility for a patient is transferred from the PCP
plines will be eliminated [28]. Under this condi- to a specialist, a patient is no longer under the
tion so-called standard diabetes care can be care of the care group; this means that the bun-
offered, i.e., purchased, delivered and billed as a dled payment for this patient is terminated. The
single product [29]. From the point of view of the specialist then bills the health insurer directly for
Dutch Diabetes Federation (NDF) this scheme that patient. During that time, when the specialist
mainly serves people who have recently been activates the hospital payment scheme, the care
diagnosed with diabetes, people whose condition group cannot claim a bundled fee for that
is well controlled and those who have no serious patient [28].
complications [30, 31]. Bundled payment con-
tracts also cover consultations with secondary
care specialists. However, this consultation Care Based on Bundled Payment
opportunity does not include referral to and treat- Contracts
ment by those specialists. Accruing expenditures
(overhead costs) which are caused by the coordi- The extent to which care is provided to a diabetic
nation and interaction of the integrated care pro- patient is defined in the NDF care standard and is
cesses such as management, coordination and approved by all national providers and patient
organisations [21, 33]. However, it sets in only tracts. Laboratory testing was also included by
from the moment a diagnosis of diabetes mellitus nine care groups. Nevertheless, group nine,
is made [30]. Any activity which is needed to which was the exception, had a separate contract
diagnose diabetes falls outside a bundled pay- with a medical laboratory. Support in smoking
ment system. Struijs et al. [28] therefore stated reduction or cessation was not included in the
that in their study of tangible effects of bundled payment in five groups. Exercise counselling was
payment formal diagnosis was not included in included in all contracts, but supervised exercise
any of the contracts they reviewed. Initial risk counselling was mentioned in the bundled pay-
assessments, even if part of the diagnostic phase, ment contracts of group nine. Because the patient
were included in all the contracts. Table 11.1 had to pay 5 per year, it is unclear whether this
gives an overview of the performance of the group claim is part of the bundled payment. Medication
contracts with respect to the NDF standards. and psychosocial care were not included in any
In the contracts they reviewed, they found that of the bundled payment contracts. These services
periodic check-ups as well as specialist consulta- were not mentioned in the NDF standards either.
tion were included in all bundled payment con-
Table 11.1 Content of the bundled payment contracts by diabetes care group
Diabetes care group
Required by
NDF Health-
Care standard 1 2 3 4 6 7 8 9 10
Diagnostic phase
Formal diagnosis No
Initial risk assessment Yes + + + + + + + + +
Treatment and standard check-ups
12-monthly check-ups Yes + + + + + + + + +
3-monthly check-ups Yes + + + + + + + + +
Eye examinations Yes + + + + + + + + +
Foot examinations Yes + + + + + + + + +
Supplementary foot exams Unclear + + + +
Foot treatment No +
Laboratory testing Yes + + + + + + +b +
Smoking cessation support Yes + + + +
Exercise counselling Yes + + + + + + + + +
Supervised exercise No +c
Dietary counselling Yes + + + + + + + + +/d
Medication No
Psychosocial care No
Medical aids No a a
Additional GP consultations Unclear +/ +/ +/ +/ +/ +/ +/
(diabetes-related)
Additional GP consultations No +/
(non-related)
Specialist consultations Yes + + + + + + + + +
From Struijs [28]
a
Medical aids limited to blood glucose strips and billed at a maximum additional fee of 4.50 per patient per year
b
Supplementary fee paid for laboratory testing (27 per patient per year) via a module additional to the bundled fee
c
Exercise programme mentioned in contract at additional fee of 5 per patient per year
d
Dietary counselling contracted for new patients only (module 1) and for those in insulin adjustment phases (module 3)
but available to other patients on specific GP referral
A supplementary foot examination was covered Management of Type 2 Diabetes

in four contracts.
It was also not clear whether any extra GP The great majority of patients have type 2 diabe-
consultations were covered by the bundled pay- tes. All professionals agree that many patients
ment contracts. Even when distinguishing with type 2 diabetes can be treated well enough
between diabetes-related and non-diabetes- under community care (Dutch estimation: 80 %).
related visits there was no consistent picture. As The Netherlands is a small country and generally
Struijs et al. [28] show, the interviews indicated distances are no issue (except for traffic jams and,
that some insurance companies interpreted the for some, public transport fares). There is a
coverage more broadly than the care groups. national consensus (LTA: national transmural
agreement [34] between 1st and 2nd/3rd line)
that describes which patients would logically be
Provision of Care treated in the community care and which ones in
specialised care. In general terms, many patients
Type of diabetes, and associated treatment are therefore treated under community care, the
requirements, decide the place where care is GPs can consult the hospital-based specialist and
given and by whom. Care usually comprises all others are referred temporarily for a specific
aspects of diabetes care. In some cases, patients problem (some of them stay under hospital spe-
are for example referred temporarily or perma- cialist care however) or are referred permanently.
nently for podiatric care to the hospital while the Referral back to the GP is guided by the nature
usual diabetes care remains provided in commu- and severity of the diabetic condition, nondia-
nity care. Table 11.2 summarises the location of betic morbidity and the wish of the patient. In
care. modern terms shared decision-making. Very
generally outlined indications for (permanent)
Table 11.2 Location of care referrals are:
Type 2 diabetes mellitus:
Community care (1st line in our terms) Intensive insulin therapy or those having trou-
Specialised care ble achieving adequate control
General hospitals (2nd line in our terms) Insulin treatment and (recurrent)
University care (3rd line in our terms) hypoglycaemia
Type 1 diabetes mellitus Severe hypoglycaemia in any patient
Specialised care Difficult hyperlipidemia
General hospitals Difficult hypertension
University care Severe obesity
Diabetes and pregnancy (including GDM) Renal impairment (eGFR <45 in patients <60
Specialised care years, <30 in those >60 years) and/or macro-
General hospitals albuminuria/proteinuria
University care Difficult neuropathy
Secondary diabetes Complicated diabetic foot
Community care (selected individuals with stable Pregnancy-related issues
diabetic disease and primary morbidity (like renal
transplant, steroid related disease)
The LTA is then translated into a RTA (regional
Specialised care (genetic cases, syndromes,
drug-related, HIV-(drug)- related) transmural agreement) with local adaptations.
General hospitals Good implementation requires good communica-
University care tion and human relations. The personal factor(s)
Specific centres for CFRD and lung transplant, is/are essential to make this scheme work. The
other solid organ transplant, bone marrow/stem cell essential issue for the government is to provide
transplants) the best care near the patient, but basically they
appear to be seeking the cheapest care by health- tion and coordination, better collaboration among
care professionals with the minimally-required the providers and better adherence to care proto-
level of expertise (in theory). cols. On the other hand they also recognised a
There are some important issues diabetes pro- dominance of the Care Groups by GPs.
fessionals are confronted with: Furthermore, prices varied to a large extent
among the care groups, and, as they state, this
There is little room for innovation (costs money) could not be fully explained by differences in the
Restricted access to new medications services offered. Moreover, outdated information
Safeguarding adequate referral to secondary and information technologies led to an increased
care. administrative burden. Nevertheless, the intro-
duction of bundled payments might turn out to be
a useful step in the direction of risk-adjusted inte-
Evaluation of the Bundled Payments grated capitation payments for multidisciplinary
Approach provider groups offering primary and specialty
care to a defined group of patients [35].
As mentioned above, changes or improvements As far as efficiency of care is concerned, after
should occur most likely in three areas: (1) Quality 3 years of evaluation, several changes in care pro-
improvement and cost reduction by reducing cesses have been observed, including task substi-
administrative/overhead costs, sharing risk, elimi- tution from GPs to practice nurses and increased
nating cost-shifting, outcomes management and coordination of care [31, 36], thus improving
continuous quality improvement, reducing inap- process costs. However, Elissen et al. [31] con-
propriate and unnecessary resource use, efficient cluded that the evidence relating to changes in
use of capital and technology; (2) consumer process and outcome indicators, remains open to
responsiveness, i.e., seamless continuum of care doubt, and only modest improvements were
and focus on the health of enrollees; (3) commu- shown in most indicators. Struijs et al. [36] pres-
nity (e.g., whole village/town) benefit by improv- ent a more differentiated picture. Process
ing community health status, and addressing the indicators like measurement of HbA1c, body
prevention of social issues which affect commu- mass index, blood pressure, cholesterol, kidney
nity health. As the RAND Corporation states, the tests and foot examination have shown improve-
evaluation should cover at least: changes in con- ments. On the other hand, this improvement was
sumer financial risk, waste reduction (as a conse- accompanied by a decrease in annual eye testing.
quence of reduced (unnecessary) services), Some intermediate outcome measures like blood
reliability through focus on key processes and pressure and cholesterol level have improved
improved coordination, patient experience, health, slightly as well. Body mass index remained
coverage, capacity, operational feasibility [26]. unchanged, and the average HbA1c has increased.
Dutch integrated diabetes care, induced by the Struijs et al. found that patients in a bundled pay-
bundled payment approach, affects both horizon- ment diabetes care programme, used less special-
tal and vertical integration of providers. So far, ist care than patients receiving usual care [37].
goals like multidisciplinary care and seamless However, there has been no improvement in out-
care (especially seen from the patients view- come parameters like efficiency outcomes to date.
point) seems to have been reached. In a recent During the first year, the expenditure per
study de Baker et al. came to the conclusion that patient was actually higher than for patients
the bundled payment led to important changes receiving usual care. In their comparison of inte-
in the financing and delivery of chronic care in grated care outcomes in three countries, Busse
the Dutch health-care system. In a relatively short et al. identified an increased annual cost of $388
period of time, care groups were created through- per patient in the Dutch model. This was associ-
out the country, providing integrated, multidisci- ated with mixed clinical outcomes but better
plinary care for patients with diabetes . [35] experiences for patients and providers (Tables
(p430). They also identified improved organisa- 11.3 and 11.4) [38]. More than 90 % of the
194
Table 11.3 Quality of diabetes care in terms of process indicators in second and third years after bundled payment implementation, percentages by care group and for total sample
1 2 3 4 6 7 8 9 10 Total
Care group year 2 year 3 year 2 year 3 year 2 year 3 year 2 year 3 year 2 year 3 year 2 year 3 year 2 year 3 year 2 year 3 year 2 year 3 year 2 year 3
Patients with:
At least 4 58.3 70.3* 45.2 57.4* 52.9 68.7* 59.7 70.7* 38.1 66.1 54.5 66.3*
check-ups
Eye 38.4 32.6 31.3 60.1 69.1* 45.5 41.4* 27.1 52 42.8* 63.6 42.8 54.2* 77.3 73.5* 48.1 44.6*
examination
in past 12
months
Eye 44.8 88.0 54.5 72.7 71.6 92.4 63.0
examination
in past 24
months
Foot 63.7 68 79.2* 67.2 84.4* 92.9 91.7* 58.9 65.1 70.8* 73 76.3 77.8 88 84.8* 78.4 82.4*
examination
HbA1c test 74.2 97.0* 89.7 93.2* 93.6 94.3 94.9 94.8 89.2 82.6 81.3* 77.6 93.9 95.2 99.3 99.0 90.8 91.4*
Blood 98.5 97.3 91.1 91.8 96.7 96.9 99.2 99.2 89.9 86.8 85.4* 94.5 88.8 82.2* 98.7 98.2 93.9 93.4*
pressure
measurement
BMI
measurement
99.4 97.3* 92.1 92.1 96.6 96.9 99.1 99.1 90.6 79.9 83.1* 94.3 86.9 81.4* 91.6 91.5 91.7 92.3*
Creatinine 71.8 80.8* 91.2 94.2* 60.8 75.0* 86.5 86.8 83.6 75.9 73.0* 60 90.8 91.2 92.5 92.1 83.8 84.6*
clearance
calculation or
test
H.W. de Valk and H. Wenzel
11
Urine test 69.1 46.8* 63.8 74.7* 55.2 69.1* 68.6 65.6 68.9* 45.8 86.8 88.4 74.2 72.6 66.6 71.8*
(spot sample)
for albumin or
albumin-
creatinine
ratio
LDL test 71.5 78.7* 83.7 87.6* 66.2 83.0* 88 88.5 81.5 75.2 71.8* 63.1 90 93.0* 96.3 95.7 83.3 84.4*
HDL test 72.1 79.0* 86.5 90.6* 66.8 84.2* 83.8 77.7 73.9* 66.1 92.5 94.4 96.5 94.3 82.7 83.9*
Total 72.7 80.5* 87.3 91.3* 66.8 84.1* 89.7 90.2 83.9 77.7 74.0* 65.9 92.7 94.4 96.6 96.8 85.4 86.6*
cholesterol
Triglyceride 72.4 80.5* 87.2 91.2* 66.7 84.1* 88.5 88.3 83.7 77.5 73.7* 65.9 92.5 94.4 96.5 93.7* 84.9 85.5*
test
Smoking 100 100 96.1 78.8 100 93.7 100 98 68.5 88.5
status know
Dietician 21.6 21 16.1* 5.9 3.4* 2.8 0.1 5.2* 7.7 7.4*
consultation
Composite indicators
Tested/know: 58.3 57.8 72.9* 60.7 74.9* 83.8 84.3 51.1 52.5 60.5* 46.4 64.6 60.4 77.4 74.2 68.1 73.7*
HbA1c, blood
pressure,
Approaches to Integrated Diabetes Care in the Netherlands
BMI, total
cholesterol,
creatinine
clearance,
foot exams
Lipid 71.2 78.7* 83.7 87.5* 66 82.8* 81.5 74.8 71.6* 62.7 89.9 93.0* 96 90.3* 80.3 81.3*
profiling (total
cholesterol,
triglycerides,
HDL, LDL)
From Struijs et al. [36]
*
Significant (P < .05)
195
196
Table 11.4 Quality of care in terms of outcome indicators, at T1 (first value in second year after bundled payment implementation) and T3 (final value in third year after imple-
mentation), by care group and for total sample
Care group 1 2 3 4 6 7 8 9 10 Total
Assessment T1/T2a T3 T1 T3 T1 T3 T1 T3 T2a T3 T2a T3 T2a T3 T1 T3 T1 T3 T1 T3
Outcome indicator
HbA1c
Mean (mmol/mol) 48.3 49.4* 56.2 54.7* 48.3 50.2* 48.5 50.1* 51.2 55.0* 49.8 49.9 49 50.4* 53.0 51.6* 50.1 49.3* 50.5 51.0*
% of patients 53 76.6 74.9 42.3 50.2* 76.4 66.9* 73.0 66.7* 65.2 50.7* 70.1 69.1 72.8 71.9 56.2 61.7 69.9 66.8* 66 63.2*
mmol/mol
% of patients >69 0.8 1.7 14.3 11.7* 1.2 1.8 2.5 2.8 5.5 10.1* 2.3 2.4 0.9 4.4 8.7 6.8 4.8 3.1* 5.3 4.7*
mmol/mol
Systolic blood pressure
Mean (mmHg) 140 139 140 137* 139 139 141 140* 137 136 141 140* 138 139 144 140* 140 137* 141 139*
% of patients <140 47.4 52.0 47.6 58.2* 52.6 52.5 42.9 46.0* 54.1 56.2 44.9 46.6* 54.3 50.4 39.3 52.2* 48.3 61.7* 45.2 51.3*
mmHg
BMI
Mean (kg/m2) 29.71 29.73 29.32 29.26 29.48 29.35* 30.21 30.15* 29.29 29.17 29.61 29.53* 28.98 28.97 30.02 29.69* 30.09 30.04 29.93 29.86*
% of patients <25 15.2 15.9 18.9 20.6* 16.4 18.0 14.3 15.7* 19.3 20.0 17.5 18.0 19.7 18.6 12.7 14.2 13.0 14.3 15.3 16.8*
Kg/m2
% of patients >30 42.4 43.9 37.6 37.3 40.9 39.4 45.9 45.2 37.7 37.8 40.4 40.0 34.1 34.8 43.8 41.5 45.0 43.7 43.5 42.7*
Kg/m2
Kidney function
% of patients with
Clearance >60 ml/min
(MDRD)
% of patients with
microalbuminuria
% of patients with
proteinuria
Lipid profile
Mean LDL (mmol/l) 2.6 2.5* 2.6 2.4* 2.3 2.3 2.6 2.4* 2.7 2.5* 2.6 2.5* 2.7 2.6 2.5 2.5 2.6 2.1* 2.6 2.4*
% of patients with 49.0 57.8* 47.8 60.0* 59.9 62.2 46.4 55.8* 43.6 48.9* 50.3 53.4* 42.0 42.0 52.4 51 49.2 71.5* 48.2 58.6*
LDL <2.5 mmol/l
H.W. de Valk and H. Wenzel
11
Mean HDL (mmol/I) 1.3 1.3 1.2 1.3* 1.2 1.3* 1.2 1.2 1.4 1.4 1.3 1.3 1.3 1.3 1.1 1.2* 1.2 1.3*
Mean triglycerides 1.6 1.5* 1.7 1.6* 1.8 1.8 1.9 1.8* 2.0 1.9 1.7 1.7* 1.7 1.7 1.7 1.6 1.6 1.5* 1.8 1.7*
(mmol/l)
Mean total cholesterol 4.6 4.5* 4.6 4.4* 4.4 4.4 4.7 4.5* 4.8 4.6* 4.7 4.6* 4.8 4.8 4.5 4.5 4.3 4.0* 4.6 4.4*
(mmol/l)
% of patients with 48.0 55.6 47.6 57.2* 57.7 57.5 42.9 53.0* 41.0 44.7* 45.6 49.6* 41.4 35.7 50.8 51.4 57.4 73.1* 46.8 56.2*
total
Cholesterol <4.5
mmol/l
Smoking
% of smokers among 10.5 14.7 14.7 18.6 14.5 15.6 23.8 13.5 18.2 16.4
patients
With known smoking
behaviour
% of quitters
Complications
% of patients with
foot problems
% of patients with eye
problems
Approaches to Integrated Diabetes Care in the Netherlands
% of patients with
any complications
From Struijs et al. [36]
*
Significant (P < .05).
a
T1 values not known for care groups 1, 6, 7 and 8 (except blood pressure and BMI for group 1), hence T2 values reported (final value in second year after implementation) and
differences with T3 assessed; not included in totals
197
patients interviewed judged the cooperation and c o m m o n w e a l t h f u n d . o rg / p u b l i c a t i o n s / p r e s s -

releases/2014/jun/us-health-system-ranks-last.
coordination to be either good or excellent. The
3. van den Berg M, Heijink R, Zwakhals L, Verkleij H,
providers perceived improved quality and more Westert G. Health care performance in the
patient-centredness [38]. The adherence to care Netherlands: easy access, varying quality, rising costs.
standards (CS) improved from 79 % to 89.2 % Eurohealth. 2010;16(4):3.
4. Schlkopf M, Pressel H. Das Gesundheitswesen im
during the period 20102013 [34]. This positive
internationalen Vergleich:
trend was transferred into high levels of patient Gesundheitssystemvergleich und europische
satisfaction and their involvement in treatment [34]. Gesundheitspolitik. 2., aktualisierte und erweiterte
Auflage, rev. Ausg ed. Berlin: MWV Medizinisch
Wissenschaftliche Verlagsgesellschaft; 2014
5. Organisation for Economic Co-Operation and
Conclusions Development (OECD). Health at a glance 2013
OECD indicators 2013 [05/15/2015]. Available from:
The limited evidence for a positive impact of h t t p : / / w w w. o e c d - i l i b r a r y. o r g / d o c s e r v e r /
download/8113161e.pdf?expires=1445258459&id=i
bundled payments under the Dutch health system
d&accname=guest&checksum=B2A1378438D4D53
has led to discussions over whether the current DB27F2A5CE6B6174A.
methodological approaches are sophisticated 6. van Roojen M, Goedvolk R, Houwert T. A Vision for
enough to differentiate between differences due the Dutch health care in 2040 towards a sustainable,
high-quality health care a report from the World
to bundled payments compared with other influ-
Economic Forum prepared in cooperation with
ences of health-care reform that could be super- McKinsey & Company Amsterdam2013 [05/15/2016].
imposed [39]. Moreover, interactions of other Available from: http://www.mckinsey.com/global-
secular factors (e.g., regional, national or local locations/europe-and-middleeast/netherlands/en/
latest-thinking.
trends and characteristics) make it difficult to
7. International Diabetes Federation. IDF diabetes atlas.
assign outcomes to the influence of bundled pay- 6th ed. Brussels: International Diabetes Federation;
ments alone. As Struijs et al. [39] point out, new 2013.
methods are needed to distinguish between the 8. Shaw JE, Sicree RA, Zimmet PZ. Global estimates of
the prevalence of diabetes for 2010 and 2030. Diabetes
effects of the core elements of the payment
Res Clin Pract. 2010;87(1):414.
reform, the core elements of the provider-led 9. American Diabetes Association. Economic costs of
entities and the core elements of the health-care diabetes in the U.S. in 2012. Diabetes Care.
delivery transformations. Such analyses could 2013;36(4):103346.
10. Zhang P, Zhang X, Brown J, Vistisen D, Sicree R,
inform the global debate over how to implement
Shaw J, et al. Global healthcare expenditure on diabe-
integrated care. For example, Busetto et al. [40] tes for 2010 and 2030. Diabetes Res Clin Pract.
argue in favour of mixed methods studies. Overall, 2010;87(3):293301.
while the Dutch approach to integrated care, using 11. Slocum TA. Thematic cartography and geographic
visualization. 3rd ed. Upper Saddle River: Pearson/
a bundled payment system with a mixed payer
Prentice-Hall; 2009.
approach, has created a limited improvement in 12. Lynch R, Altenburg-van den Broek E. The drawbacks
integration, there is no evidence that the approach of Dutch-style health care rules: lessons for
has reduced morbidity and premature mortality: Americans. Backgrounder. Washington, DC: The
Heritage Foundation; 2010.
and it has come at an increased cost.
13. Daley C, Gubb J. Healthcare systems: The
Netherlands: CIVITAS; 2013 [05/15/2016]. Available
from: http://www.civitas.org.uk/research/health/
health-systems/.
References 14. Gowling A. Dutch health insurers squeeze health pro-
viders: IAM EXPAT; 2013 [05/15/2016]. Available
1. Bjrnberg A. Euro health consumer index 2014. from: http://www.iamexpat.nl/read-and-discuss/
Health Consumer Powerhouse Ltd; 2015. expat-page/news/dutch-health-insurers-
2. Mahon M, Fox B. US health system ranks last among squeeze-providers.
eleven countries on measures of access, equity, qual- 15. Gowling A. Dutch doctors strike to protest health
ity, efficiency, and healthy lives the commonwealth insurance amendment 2014 [05/15/2016]. Available
fund 2014 [05/15/2016]. Available from: http://www. from: http://www.iamexpat.nl/read-and-discuss/
expat-page/news/dutch-doctors-strike-health- Public Health and the Environment (RIVM) 2010

insurance-change. RIVM-rapportnummer: 260224002.
16. Clark B. Dutch health insurance premiums set to rise 29. Taakgroep Programma Diabeteszorg. Diabeteszorg
in 2015 2014 [05/15/2016]. Available from: http:// Beter rapport van de Taakgroep Programma
www.iamexpat.nl/read-and-discuss/expat-page/news/ Diabeteszorg. The Hague: Ministerie van
dutch-health-insurance-premiums-rise-2015#sthash. Volksgezondheid, Welzijn en Sport; 2005.
cRkDlxco.dpuf. 30. Nederlandse Diabetes Federatie. NDF
17. Health Insurance Information Centre. Compare health Voedingsrichtlijn Diabetes 2007.
insurances 2015 [05/15/2016]. Available from: http:// 31. Elissen A, Duimel-Peeters IGP, Spreeuwenberg C,
www.zorgverzekering.org/eng/compare/. Vrijhoef HJM, Nolte E. The Netherlands. In: Nolte E,
18. MedTech Europe. The Netherlands: market-based Knai C, editors. Assessing chronic disease manage-
reform n.d. [05/15/2016]. Available from: http://www. ment in European health systems country reports.
reforminghealthcare.eu/economist-report/some- 39th ed. Copenhagen: World Health Organization;
roads- ahead-innovative-approaches-in-five-west- 2015.
european-countries/the-netherlands-market-based- 32. Raaijmakers LGM. Diabetes care and the introduction
reform. of a new standard in the Netherlands a closer look
19. Government of the Netherlands. Health insurance from the perspectives of health care professionals and
standard health insurance n.d. [05/15/2016]. patients. Dissertation. Maastricht: Maastricht
Available from: https://www.government.nl/topics/ University; 2014.
health-insurance/contents/standard-health-insurance. 33. Nederlandse Diabetes Federatie. NDF Zorgstandaard
20. Koenraadt B. Healthcare insurance in the Netherlands: diabetes type 2 volwassenen. 2015 [05/15/2016].
an overview 2014 [05/15/2016]. Available from: Available from: http://www.zorgstandaarddiabetes.nl/
http://www.iamexpat.nl/read-and-discuss/expat-page/ type-2/.
articles/healthcare-insurance-in-the-netherlands- 34. Raaijmakers LG, Kremers SP, Schaper NC, de Weerdt
overview. I, Martens MK, Hesselink AE, et al. The implementa-
21. Struijs JN, Baan CA. Integrating care through bun- tion of national action program diabetes in the
dled payments lessons from The Netherlands. N Netherlands: lessons learned. BMC Health Serv Res.
Engl J Med. 2011;364(11):9901. 2015;15:217.
22. de Wildt JE, Leusink GL, Beek GJ. Nulmeting zorg- 35. de Bakker DH, Struijs JN, Baan CB, Raams J, de
groepen: Een beschrijvend onderzoek van de karak- Wildt JE, Vrijhoef HJ, et al. Early results from adop-
teristieken. Utrecht: LVG; 2008. tion of bundled payment for diabetes care in the
23. Struijs J. Integrating care by bundled payments: les- Netherlands show improvement in care coordination.
sons from the Netherlands: Rijks Instituut voor de Health Aff (Millwood). 2012;31(2):42633.
Volksgezondheid en Milieu; 2012 [05/15/2016]. 36. Struijs JN, de Jong-van Til J, Lemmens L, Drewes H,
Available from: http://www.kingsfund.org.uk/sites/ de Bruin S, Baan C. Three years of bundled payment
files/kf/jn-struijs-integrated-care-netherlands- - for diabetes care in the Netherlands. Bilthoven:
kingsfund-sept12.pdf. National Institute for Public Health and the
24. Ham C, Zollinger-Read P. What are the lessons from Environment (RIVM); 2012.
the USA for clinical commissioning groups in the 37. Struijs JN, Mohnen SM, Molema C, de Jong-van Til
English National Health Service? Lancet. JT, Baan CA. Effects of bundled payment on curative
2012;379(9811):18991. health care costs in the Netherlands an analysis for
25. Luft HS. Becoming accountable-opportunities and diabetes care and vascular risk management based on
obstacles for ACOs. N Engl J Med. nationwide claim data, 20072010. Bilthoven:
2010;363(15):138991. National Institute for Public Health and the
26. Bertko J, Effros R. Analysis of bundled payment. Environment (RIVM); 2012.
RAND Corporation; 2010 [05/15/2016]. Available 38. Busse R, Stahl J. Integrated care experiences and out-
from: http://www.rand.org/pubs/technical_reports/ comes in Germany, the Netherlands, and England.
TR562z20/analysis-of-bundled-payment.html. Health Aff (Project Hope). 2014;33(9):154958.
27. Washington State Hospital Association. Integrated 39. Struijs JN. Payment reform and integrated care:
delivery systems. n.d. Available from: http://www. the need for evaluation. Int J Integr Care.
wsha.org/files/62/Gov_Board_Manual_integrated. 2013;13:e056.
pdf. 40. Busetto L, Luijkx KG, Vrijhoef HJ. Implementation
28. Struijs JN. Experimenting with a bundled payment of integrated care for type 2 diabetes: a protocol for
system for diabetes care in the Netherlands the first mixed methods research. Int J Integr Care.
tangible effects. Bilthoven: National Institute for 2014;14:e033.
in Sweden 12
Helmut Wenzel, Stefan Jansson,
and Mona Landin-Olsson
Basic Principles of the Swedish and municipalities [2]. Figure 12.1 shows the
Health-Care System make-up of the organisation, and the financial
flows.
The Swedish health-care system is a form of the In principle, the central government sets the
Beveridge model, with a strong orientation health-care policies, whereas the local govern-
towards subnational levels, i.e., municipalities ments organise the delivery of services to
and regions. Thus, the responsibility is shared ensure that their own residents receive the med-
between the central government, county councils ical care they need [3]. Health care is mainly
(landsting in Swedish) and municipalities tax funded; this applies at Government level as
(kommuner in Swedish). The Health and well as to the levels of the county councils and
Medical Service Act regulates the responsibilities the municipalities. The latter also levy propor-
of the different actors, giving the local govern- tional income taxes on the population to cover
ments more freedom in this area [1]. The role of the services that they provide [4]. Therefore,
the central government is to establish principles the county council tax, supplemented by a gov-
and guidelines, and to set the political agenda for ernment grant, is the main means of financing
health and medical care. This is primarily the health-care system. In addition, small user
achieved through laws and ordinances or by fees are paid at the point of use. Long-term care
reaching agreements with the Swedish for the elderly is financed and organised by the
Association of Local Authorities and Regions municipalities. Each county council and region
(SALAR), which represents the county councils is governed by a political assembly, whose rep-
resentatives are elected for 4 years in general
elections [5]; Anell [4] gives a detailed over-
view on the financial flow.
Swedish health care is also characterised by its
H. Wenzel (*)
Health Economist, Konstanz, Germany universal access and both visits and prescriptions
e-mail: [email protected] of drugs are heavily subsidised meaning that vir-
S. Jansson tually all people can get help with their medical
Department of Family Medicine, Brickebacken problems. The Ministry of Health and Social
Primary Health Care Center, rebro, Sweden Affairs is primarily responsible for drafting health
e-mail: [email protected] policy legislation. It also works with the county
M. Landin-Olsson councils and municipalities to determine how to
Department of Endocrinology, Lund University best finance and deliver health care to the citizens
Hospital, Lund, Sweden
e-mail: [email protected] of Sweden [6]. It supervises 25 agencies, of which

202 H. Wenzel et al.
Fig. 12.1 Organisation of the Swedish Health-Care System and Financial Flows (Used with permission from Anell
et al. [4])
11 are directly involved with health care at the Jnkping, Kronoberg, Skne, Vstra Gtaland,
national level. These are: The National Board of rebro and stergtland as well as the Gotland
Health and Welfare, The Health and Social Care municipality are called regional councils and
Inspectorate, The Public Health Agency, The have assumed responsibility for regional devel-
Swedish Agency for Health and Care Services opment from the state. There is no hierarchical
Analysis, The Swedish Agency for Health relation between municipalities, county councils
Technology Assessment and Assessment of and regions. Around 90 % of the work of Swedish
Social Services, The Swedish eHealth Agency, county councils concerns health care, but they
The Medical Products Agency, The Dental and also deal with other areas such as culture and
Pharmaceutical Benefits Agency, The Medical infrastructure.
Responsibility Board, Apoteket AB, and The The municipalities are responsible for care for
Social Insurance Agency [7]. The central govern- the elderly in the home or in special accommoda-
ment provides funding, and distributes money and tion, i.e., nursing homes. Their duties also include
resources to each of the agencies, mentioned care for people with physical disabilities or psy-
above and establishes the general role of each the chological disorders and providing support and
agencies are then free to act independently, decid- services for people released from hospital care as
ing on their own how to address particular issues. well as for school health care. Chronic diseases
Sweden is divided into 290 municipalities and that require monitoring and treatment, and often
21 county councils (Fig. 12.1). Nine of the county life-long medication, place significant demands
councils: Gvleborg, Halland, Jmtland, on the system [8].
12 Integrated Diabetes Care in Sweden 203
In this way, the health system is highly decen- of the so-called economic crisis (especially in
tralised and organised; however, this does lead to 2008) [15].
variation in care provision. Primary and second-
ary care are funded and delivered at county level.
Municipalities are responsible for nursing and Performance of the System:
residential homes as well as home care and other The OECD Health Care Quality
social services [9]. This, along with other issues, Reviews
leads to fragmented care. For example, hgren
identifies three major causes: decentralisation, The OECD Health Care Quality Indicators proj-
which leads to rather independent processes, a ect, initiated in 2002, aims to measure and com-
high degree of specialisation, driven by medical pare the quality of health service provision in the
development, and professional organisation, different countries, based on a set of quality indi-
which means that nurses and doctors focus on cators at the health systems level, which allows
their part of the care process for which they are an assessment of the impact of particular factors
responsible. Aiming for common health-care on the quality of health services [16]. According
goals has a low priority [10]. to the Health Care Quality Review of 2013, the
quality of health care in Sweden is generally
good. Rates of avoidable hospitalisation for
Expenditure and Evaluation chronic conditions such as asthma (22.2 per
of the System 100,000 population) are among the lowest in the
OECD (average 45.8) and long-term complica-
In 2013 Sweden spent the equivalent of 4,904 tions of diabetes including amputation rates
USD per capita on health whereas the among diabetic patients are all lower than the
Organisation for Economic Co-operation and OECD average [17]. In the 2011 survey, 90 % of
Development (OECD) average was at 3,453 USD, primary care patients in Sweden said the particu-
with 84 % of the expenditure coming from public lar staff treated them respectfully. Swedens qual-
sources. Both numbers are above the OECD aver- ity registers, for tracking the quality of care that
age [11]. Compared with the OECD average of 73 patients receive and the corresponding outcomes
%, Swedish Government spending has one of the for several conditions, are among the most devel-
highest proportions of public expenditure among oped across the OECD [17].
OECD countries. Similarly high proportions can Yet, the coordination of care for patients with
be found in the Netherlands (88 %), Norway (85 complex needs is less good. Only one in six
%) and Denmark (84 %) [11]. Health spending as patients had contact with a physician or specialist
a proportion of Gross Domestic Product (GDP), nurse after discharge from hospital for stroke,
as an indicator for affordability, grew from 7.4 % again with substantial variation across counties.
in 2000 to 11 % in 2013; the corresponding OECD Fewer than half of patients with type 1 diabetes,
average in 2013 was 8.9 % [11, 12]. Health expen- for example, have their blood pressure adequately
diture in the Netherlands, as a percentage of GDP, controlled, with a considerable variation (from
increased from 7.4 % in 1980 to 11.9 % in 2011 26 % to 68 %) across counties [17].
[13, 14]; in Sweden it was 10.6 % in 2011. In As with primary care, the governance struc-
Germany the comparable numbers rose from 9.8 ture around long-term care is divided.
% in 2011 to 11.1 % in 2014; in Switzerland the Municipalities are responsible for institutional
share was 11.1 % in 2013, and it was 16.4 % in care and nursing care in private homes, and
2013 in the USA [12]. county councils are responsible for the delivery
Interpreting this time series (i.e., share of and financing of medical care. There are few
GDP) needs caution: increasing rates could be built-in incentives for coordination across these
due to increased spending, as well as due to a governance levels, or across the health and social
reduction in overall GDP, e.g., as a consequence care components of long-term care services.
Whereas central government has set out a holistic Coordination of care between hospitals, pri-
vision for care of the elderly, this has not success- mary care providers and local authorities is
fully transformed to local implementation. Joint becoming one of the biggest challenges to the
care planning, joint purchasing or bundled pay- continued quality of Swedens health and social
ments, which should help to integrate local health care system. This means, the central government
and long-term care services, are frequently will have to set out responsibilities very clearly,
absent. Consistently high levels of unnecessary by developing standards, building the evidence
hospitalisation for the elderly are a clear indica- base and sharing knowledge. For example, cen-
tor of this fragmentation. For example, at 260 tral authorities should be given a more defined
admissions per 100,000 people aged over 80, role in assuring the quality of services by setting
avoidable hospital admissions for uncontrolled out national quality standards. Moreover, clear
diabetes in Swedens elderly population are the standards are needed to underpin the new inter-
sixth highest in the OECD, and about 1.5 times mediate care facilities being developed by munic-
higher than in Denmark. On the other hand, the ipalities. The information infrastructure must
average length of stay in hospital after a myocar- improve by developing new indicators of quality
dial infarction in Sweden is less than 5 days, of care provided by GPs and elderly care ser-
which is among the lowest in the OECD, and a vices. Finding ways to link across different data
sign of efficiency [17]. sources is also necessary, to allow a complete
Municipalities are often not adequately picture of an individuals care to be built up [18].
equipped to manage patients after hospital dis- Waiting times have long been a cause of dis-
charge. Home care services needs up to 5 days satisfaction [19]. In an OECD ranking of 2011,
to reorganise their work, and allocate resources, Sweden was rated second worst [20]. Therefore,
to ensure a safe and convenient situation in the Sweden introduced a health-care guarantee in
home after hospital stay. This situation leads to 2005. This means that all patients should be able
crowding in the hospitals, with many patients to be in contact with a primary health care the
unable to leave the hospital, staying for social same day they seek help, and have a doctors
rather than medical reasons. Only around 20 % appointment within 7 days. After an initial exam-
of primary care doctors in Sweden report that ination, no patient should have to wait more than
they receive the information necessary to man- 90 days to see a specialist, and no more than 90
age a patient within 48 h after hospital dis- days for an operation or treatment, once it has
charge, compared to almost 70 % in Germany. been determined what care is needed. If the wait-
Swedens health and elderly care systems ing time is exceeded, patients are offered care
deserve their reputation as being among the best elsewhere; the expenditures, including those for
in the world. Yet, an ageing population with travelling, is then paid by their county council [8,
growing chronic conditions and requiring more 21] . Statistics from 2015 indicate that about
complex health services are testing Swedens eight out of ten patients see a specialist within 90
ability to continue delivering high-quality care, days and receive treatment or are operated on
according to a new OECD report. Sweden has a within a further 90 days. In 2013, 78 % felt they
larger share of elderly people than most OECD received the care they needed. In 2006, the figure
countries: 5.2 % are over 80, compared to the was 74 % [8]. Most patients who appeal under
average of 4.2 %. Spending on elderly care is the health-care guarantee and prioritised in the
3.6 % of GDP, compared to an OECD average queue had acute conditions rather than medical
of 1.7 %. The country also has the highest num- problems as a consequence of an underlying
ber of care workers per capita, and they deliver chronic disease. Patients waiting for a hip
care where it is generally most wanted - at replacement or a cataract surgery are cured after
home. Seven out of ten dependent elderly peo- surgery and no life-long follow-up is needed.
ple receive care in their homes [17]. When such patients are prioritised, the long-term
care for patients with chronic diseases is
crowded out, lowering their priority and risking to include all health care that is provided for a
worse outcomes. The health-care guarantee can specific patient group within a county council,
therefore lead to longer intervals between check- i.e., all health care produced within the catch-
ups, with difficulties in accessing health care if ments area of the county council in question.
their pre-existing condition has deteriorated. Consequently, patient flow within a hospital or
The OECD summarises Swedens generous within a primary care centre is not a chain of care
health care system performs well on most quality [10]. Chains of care provide evidence-based
indicators but like all other OECD countries, it health care and have to take into account clinical
faces a number of challenges.... (17 p. 13). guidelines, e.g., agreements on the distribution of
Challenges are, for example, better coordination medical work between different providers of
between primary, secondary care and community health care within a county council catchment
health services. There is now a greater interest in area [10]. In other words, chains of care involve
enabling patients to make choices and in encour- coordinating multidisciplinary care based on
aging quality-based competition between provid- clinical guidelines. This approach is similar to
ers. Taking the consumer and patients point of the concept of Managed Clinical Networks,
view, Powerhouse in their Euro Health Consumer which aims to bring together different health pro-
Index 2014 stated that, in principle, the Swedish fessionals and organisations and help them work
healthcare services are excellent, but their ranked in a more coordinated manner without restric-
position is dragged down by the seemingly never- tions [23]. Applying Fulops typology these char-
ending story of access/waiting time problems, in acteristics reflect more of a virtual clinical
spite of national efforts such as Vrdgaranti integration through contractual agreement than a
(National Guaranteed Access to Healthcare); in real organisational integration [24]. This kind of
2014, Sweden dropped to 12th place with 761 virtual integration between commissioners and
points (22 p. 5). providers can often involve several responsible
authorities and medical providers [10].
Commissioners set up the contractual agreements
Chains of Care: The Swedish with providers which specify volume, cost, qual-
Approach to Integration ity and method of delivery [24]. Chains of care
may have a manager responsible for organising
Between 2000 and 2011 several reforms were activities resources and finance. The two largest
implemented, aiming at introducing the waiting- Swedish county councils have each developed
time guarantee, improving the transparency of more than 50 chains of care [9].
quality comparisons between counties, address- Although numerous chains of care have been
ing the national coordination of highly special- established, appraisals have concluded that there
ised care, fostering choice and privatisation in is little evidence that significant change have
primary care. In 2008 the government also intro- been achieved [24]. Ham et al. [25] also conclude
duced a bonus payment to county councils that that studies show limited impact of chains of
met national waiting-time targets [4]. care. Moreover, they highlight the challenges,
However, the ever-increasing fragmentation in which result from overcoming professional and
health care called for approaches to integrate the structural barriers to realise integrated care. Yet,
care of the various providers. Quality of care others see some benefits such as a reduction in
issues were the most important reason for devel- the number of hospital beds and other improve-
oping chains of care. These chains of care are ments in efficiency [26]. On the other hand, a
coordinated activities within health care, linked number of problems were created, for example, a
together to achieve a qualitative final result for lack of physicians in nursing homes. Additionally,
the patient. A Chain of Care often involves sev- some resistance from GPs towards integration,
eral responsible authorities and medical provid- the policy shift towards free choice for patients,
ers (10 p. 2). This means that a chain of care has and the competing demands of managing
competition alongside collaboration have also and is estimated to be close to 89 for women and
caused challenges [26, 27]. Other obstacles 87 years for men in 2060 [30]. The percentage of
include resistance among some health-care man- people above the age of 65 is now 19.6 % [31]
agers who are afraid of changes. Some authors and by 2060 it is estimated to be 25 % [32]. This
also point to perceived threats to clinical auton- development of a growing proportion of elderly
omy; moreover, it is also expected that the dif- will increase the pressure on the health-care sys-
ferentiation of clinical functions, which comes tem to change and be prepared to meet people in
from sub-specialism, could further contribute to older age with different claims about their own
fragmentation [28, 29]. For Curry et al. [24] this health. Diabetes mellitus is no derogation from
indicates limits of contractual integration, and this development and the prevalence of diabetes
emphasises the point that commissioners face is increasing, mostly due to people with diabetes
major challenges in using their influence to cre- living longer with the disease than previously
ate better integrated care. [33]. Currently, the total diabetes prevalence in
For hgren the most frequently stated reason Sweden is estimated at 4.7 %, which is equivalent
for this lack of success is the intense compart- to almost 450,000 individuals [34].
mentalisation of responsibilities between differ- Diabetes is essentially treated effectively by
ent professions and departments, while the multidisciplinary teams consisting of doctors and
responsibilities and power remain in the vertical diabetes nurses in collaboration with other pro-
organisational structure. Limited or even absent fessionals, i.e., dieticians, podiatrists, physiother-
participation among some local authorities has apists, psychologists (Fig. 12.2a, b). Regular
also been perceived to be responsible for chains follow-up of patients with type 1 diabetes are car-
of care, which were developed, missing some ried out in hospitals by specialist endocrinolo-
activities throughout a patients treatment [10]. gists while type 2 diabetes patients are followed
To ensure or improve quality of care, county up in primary health-care centres (PHCC) by
councils will continue to develop chains of care GPs. Much of the care given is closely assessed
in spite of their doubts whether they have been through the Swedish National Diabetes Register
successful in their development [10]. (NDR) with 100 % of specialist clinics and 95 %
hgren et al. note that policy makers pro- of PHCC reporting to the register. The reporting
moted two important changes in health care with is not mandatory, but all clinics and PHCC are
a strong impact on the provision of unfragmented encouraged to do so. The Swedish Association
care: free patient choice in primary health care for Diabetology initiated the NDR in 1996. The
and mergers of hospitals and clinical depart- purpose of NDR is to endorse evidence-based
ments. These decisions will also have an adverse development of diabetes care by offering up-to-
effect on the development of integrated care. date information about changes in the treatment
Moreover, they recommend putting more efforts of glycaemia and other risk factors, as well as
into evaluation of integrated care solutions in diabetic complications. Data on quality indica-
order to provide a basis for a more rational tors are continuously collected from different
decision-making and to replace political beliefs clinics and primary health-care centres. For every
about the benefits of such policies with evidence clinic/PHCC it is then possible to make compari-
[27]. son with other units in the region or on national
level. Anyone can easily access all available data
instantaneously by internet. This means that pro-
Growing Burden of Diabetes viders as well as patients can seek information
about a certain diabetes care unit to compare the
Sweden has a population of almost ten million health care given there with other units. Moreover,
people with an increasing life expectancy for the overall aim of NDR is to reduce morbidity
women and men over the past decades, reaching and mortality and to conduct a cost-effective
84 years for women and 80 years for men in 2013 diabetes care while every participating unit is
Fig. 12.2 (a, b) The care of the diabetic

patient is more complex if more
complications exist
interested in knowing how well they are doing in highest score ever observed in a HCP Index) as
their care of people with diabetes. the country with the best diabetes care delivery in
Having a very high quality registry provides Europe [22]. The secret of the high performance
good opportunities to engage in a high-quality is probably the art of knowing what you are
diabetes care. This is the probable main reason doing. It was the only country out of 30 coun-
why Sweden turned out so well in a recent diabe- tries assessed that could provide data on all 28
tes investigation among EU member states. The indicators and areas, including prevention, case
Euro Diabetes Index 2014 shows Sweden (936 finding, range and reach of services, access to
points out of a maximum possible of 1,000; the treatment and care, procedures and outcomes.
Organisation of Diabetes Care some specified diseases. Some of these decision

supports have been developed through coopera-
Within each region / county council the care of tion between the profession and an authority, for
patients with diabetes is divided. Patients with instance sick leave. Guidelines for treating diabe-
type 1 diabetes get their care at specialist clinics tes with different conditions and associated diag-
in hospitals and the majority of patients with type noses are solely developed by the profession.
2 diabetes in primary care. Patients with type 2 Different specialists provide education for
diabetes who have severe complications are GPs and nurses and in some cases even for com-
referred to the Diabetes Clinics at the hospital. munity nurses. As an example, in the diabetes
Approximately 10 % of all patients with type 2 field, there is also a collaboration between diabe-
continue their care at the hospital clinics. They tologists and some GPs who have deeper skills in
are almost always on insulin in high doses often diabetology in providing education for primary
in combination with oral agents but despite mas- health-care providers, e.g., GPs and nurses spe-
sive medication many of these patients have dif- cialised in diabetes. This education can be done
ficulties to achieve metabolic balance. Patients physically as webinars or linked by video.
with advanced complications such as foot ulcers,
macroangiopathic manifestations and treatment
with dialysis are also treated at the hospitals. Diabetes and Pregnancy
Since these patients have multiple medical prob-
lems, the care has to be carried out in close col- Women with type 1 are informed early about
laboration with primary care and home health pregnancy risks and advised to plan their preg-
care. There is also a shared use of resources, nancies in advance. When pregnancy is detected
where hospital doctors can refer patients with the woman should be referred immediately to a
type 1 diabetes to primary health-care providers, specialised maternal diabetes team to ensure
e.g., if patients have asthma/chronic obstructive optimised metabolic control combined with
pulmonary disease, psychological disease or intense obstetric management. Women with type
musculoskeletal disorders. Primary care physi- 2 diabetes have contact with a midwife at the
cians assess referrals and decide which profes- PHCC in the beginning of the pregnancy. During
sion will best take care of the patient, for example, the rest of the pregnancy the women are followed
a doctor, a counsellor or a physiotherapist. The up through local and regional treatment pro-
specialist in the hospital can also write a referral grammes, which differ in different parts of
to another specialist at another hospital in the Sweden. For instance, in some regions, the diabe-
region or to other hospitals outside the region for tes care of women with type 1 diabetes and
patients with type 1 diabetes. There may be women with type 2 diabetes treated with oral
patients who are in need of, for example, surgery, antihyperglycaemic agents and/or insulin treat-
or other more specific treatments carried out only ment is provided by the hospital maternal care
in some hospitals in the country. unit through a multidisciplinary team of diabe-
In the majority of regions and county councils tologists, diabetes nurses and midwives (Fig.
all records are electronic. Some county councils 12.3). All other contacts during the pregnancy are
have the same electronic medical records (EMR) provided by the midwives at the PHCC. The
in primary and secondary care and therefore can treatment follows a standardised programme
electronically refer between each other. In other which varies in different regions of the country. A
regions EMR can be read by all, but referrals national care programme is under development.
between primary and secondary care have to be For women with type 2 diabetes, the follow-up
postal. For both doctors and nurses there are elec- continues at the PHCC according to a special
tronic decision supports at the point of care for care programme. The midwives are responsible
Fig. 12.3 Integrated and

multi-professional care of
women with type 1 diabetes
and pregnancy
for the antenatal care, but diabetes nurses and Care at Municipalities
GPs are also involved. Women are referred to a
hospital for evaluation by a diabetologist. Women When patients with type 1 or type 2 diabetes need
remain at the PHCC level as long as the glucose social services including short-term and long-
control is good according to the care programme. term special housing (e.g., for the elderly and
When insulin treatment needs to be started to disabled), this becomes the municipalities
optimise glucose control, the woman is then responsibility for providing care. The municipali-
referred to the maternity care unit in the hospital. ties are also responsible for various types of
For women diagnosed with gestational diabetes domestic assistance for people living in their own
mellitus (GDM) pregnancy care continues at the housing. Many persons also receive home-based
PHCC as long as glucose control is good. When curative health care provided by nurses and aux-
glucose values deteriorate, care is transferred to iliary staff employed by the municipalities.
the maternity unit at the hospital. All deliveries
occur at a hospital. After pregnancy, the care con-
tinues as usual at the hospital unit for women Primary Health-Care Level
with type 1 diabetes and at the PHCC for women
with type 2 diabetes. Women who have had GDM For patients with type 2 diabetes, doctors in pri-
during pregnancy are followed up according to a mary care write referrals to various units in the
special care programme including visits to the hospital for patients in need of different types of
PHCC annually to receive advice on healthy life- medical assistance, for instance orthotic shoes,
style, e.g., physical activity, diet, and to undergo eye examination, or if they need to be examined
an oral glucose tolerance test (OGTT) for the by, e.g., a neurologist or surgeon. Patients with
exclusion of incident diabetes. more complex diabetes may need a second opin-
ion, and then the patient is referred to the ity. The same authority concluded in another
diabetologist for assessment. Many of these report in 2013 that there is a need for a more in-
patients use the hospital clinic as advisory, to depth analysis of the various forms of health care
obtain recommendations for their ongoing care. and of possible inequalities based on socioeco-
The patient will come back for follow-up and nomic factors rather than diagnosis of illness [36].
receive revised recommendations with certain Studies regarding the effect of the reforms on the
time intervals but the main care is still given at care of patients with diabetes are few but a recent
the PHCC. Each PHCC also has a medical study in 7,121 adult diabetes patients living in
responsibility for patients living in nursing south of Sweden showed that continuity of care
homes. Usually this means that primary care doc- given by GPs in PHC decreased substantially after
tors make rounds once a week to the care accom- the reform. The decrease was independent of sex,
modation. The round may involve both direct and type of diabetes, age, and income [37].
indirect contact with the patient. The charge
nurse at nursing homes is always present.
Type 1 Diabetes
Primary Health-Care Changes Children defined as patients younger than 18

years are treated at 1 of the 43 Swedish
During the past years the structure of primary Departments of Paediatrics. At the onset of the
health care has been changed. Before 2010, disease, the children are hospitalised for about a
patients were, by law, assigned to the PHCC in the week. During the stay, the whole family is offered
geographic area in which the patient lived. The contact with both the diabetes team and the child
patients could apply to be listed at another PHCC psychiatry clinic. Parents are given up to 2 weeks
outside the living area, for instance close to their of sick leave after leaving the hospital in order to
job, but there was no guarantee that they would be adapt life to a child with diabetes. The hospital
listed there. In 2010 the Swedish government initi- seeks contact with the childs school to inform
ated Health Choices, which allowed patient free- and educate the school staff about the patient and
dom in choosing the PHCC wherever they like. his or her special needs. During the first year the
Most Swedish studies of such freedom of choice diabetic child and the parents have several visits
systems are built around having studied the to the clinic and the diabetes team consisting of
change before and after the introduction of health- doctor, nurse and dietician. For continuous fol-
care choice. This makes it difficult to state whether low-up, the patient and his/her family have at
a change in waiting times, for example, is an effect least four visits per year when they meet both a
of health-care choice in itself, or an effect of some doctor and a diabetes nurse. The patients with
other health-care reform introduced by The diabetes at the Paediatric Departments are seen
Swedish Agency for Health and Care Services by diabetes staff and registered in two registers.
Analysis [35]. However, some conclusions from The first register, Better Diabetes Diagnosis
the reports are worth mentioning. Those who (BDD), aims to provide the correct classification
made an active choice of provider were more satis- of the diabetes type by human leucocyte anti-
fied, although patients chose their health-care pro- gens (HLA) typing and testing for beta cell spe-
vider on the basis of proximity and reputation, not cific autoantibodies [38]. About 90 % of the
on medical results. The costs have remained under children are classified as having Type 1 diabe-
control, the effects on patient waiting times were tes based on positive autoantibodies and a few
small and the patients now live at a shorter dis- percent receive a diagnosis of Maturity Onset
tance from their medical centre. Finally, the Diabetes of the Young (MODY) [39]. Type 2
Vrdanalys stated that there are deficiencies in diabetes among children is very rare in Sweden.
knowledge about the effects of freedom of choice The departments also take part in a second
in terms of efficiency, medical quality and equal- register, a quality register called Swed-diab kids.
This register keeps track of the clinical perfor- About 22 % of patients aged between 18 and
mance for all Paediatric Clinics by registration of 21 years have insulin pumps but the usage varies
HbA1c and secondary complications. During largely in different parts of the country [42]. The
recent years a significant improvement measured remaining patients use either disposable or car-
as a decrease in median and mean HbA1c has tridge insulin pens. All patients with Type 1 dia-
been observed across the whole Sweden. The betes are treated with insulin analogues for both
mean HbA1c during 2015 was 57 mmol/mol basal and bolus doses. Retinal photos occur every
which is the lowest value ever in the register [40]. other year and the pictures reviewed by an oph-
In 2014, a majority, i.e., 52.2 % of the patients thalmologist. Foot inspections occur at least
between 0 and 18 years had insulin delivered by annually and help with shoes or insoles are given
continuous subcutaneous insulin infusion (CSII if needed. If complications occur, contact is
or insulin pump), [40] and an increasing propor- established with the appropriate team in other
tion of the patients use continuous glucose moni- departments such as cardiologist, vascular sur-
toring (CGM). At the age of 18 years, the patient geon, orthopaedist or ophthalmologist.
is transferred to the adult diabetes clinic at the Pregnancies in women with Type 1 diabetes are
hospital since patients with Type 1 diabetes are treated at specialised maternity care units with a
not referred to primary care. team of obstetricians, midwives and a diabetolo-
When adults develop type 1 diabetes they are gist with experience with these women.
initially hospitalised for 35 days but referred to Almost all of patients with Type 1 diabetes
day care units as soon as possible. Insulin treat- receive care primarily from outpatient diabetes
ment is started immediately, and patients receive clinics in hospitals. Most diabetes clinics are
training in how to inject insulin and laboratory organised within a hospitals medicine or endo-
tests are carried out. The main purpose of the crine unit and the diabetes care is managed by
inpatient stay is to teach the person to manage specialists in internal medicine or endocrinology/
insulin and learn to self-care. This education, as diabetology and by specialised nurses. In a
mentioned, is provided by diabetes nurses and is broader perspective the diabetes care is often pro-
continued into outpatient care. Two weeks of sick vided by multi-professional teams to allow the
leave is standard after discharge from hospital. opportunity to meet different patients specific
After the initial inpatient care episode, the outpa- needs. A diabetes team beyond physicians and
tient care starts, including fundal photography diabetes nurses may include dieticians, podia-
and meeting with the dietitian. trists, counsellors, physiotherapists, specialists in
In some cases, when the patient needs inten- vascular surgery and infectious diseases.
sive care, the care process can be followed by up Diabetes nurses play an important role in dia-
to 10 days in hospital. In the first year, approxi- betes care. These nurses are employed in both
mately six follow-ups occur, to primarily educate diabetes clinics and in the primary health-care
the patient about the disease and to manage their services. Diabetes care includes different compo-
own care. These follow-ups include, e.g., labora- nents, but one of the more important parts is
tory tests, selection of insulin and the injection patient-education, which is managed by diabetes
technique. At least two visits to the doctor and nurses. The goal for such education is to teach
two visits to the diabetes nurse per year are rec- patients to manage their own care on a daily basis
ommended, but in practice, often only one visit and to recognise risk factors in order to maintain
per year to the physicians and diabetes nurses a good quality of life. Many hospital-based dia-
respectively takes place. The incidence of Type 1 betes clinics and primary health-care clinics offer
diabetes in the age group 1834 years. has almost group-based patient-education. Consequently,
doubled during the last decades, which is a simi- specialised diabetes nurses responsible for
lar trend to what has been observed among chil- patient-education must not only have specific
dren [41]. diabetes knowledge, but must also have compe-
tence in adult learning approaches.
Type 2 Diabetes decentralised and organised; this leads to varia-

tion in care provision. To overcome this, the gov-
Similar to many other developed countries, ernment has created national guidelines for
patients with type 2 diabetes are usually treated common diagnoses including diabetes. The
under primary health care. Only if severe compli- health-care system is trying to combine decen-
cations arise is the care shifted to specialised tralisation, a high degree of specialisation, and
clinics in hospitals. This is the case for about 10 professional organisation in a system where com-
% of the patients and these patients have the most mon health-care goals can be maintained. To
expensive care due to costly treatment of for avoid fragmentation, chains of care have been
example foot ulcers and renal insufficiency. identified to bridge different care givers.
Uncomplicated patients continue at the PHCC Diabetes is managed by multidisciplinary
with regular medical check-ups, usually through teams consisting of doctors and diabetes nurses
one visit per year to the general practitioners in collaboration with other professionals at pri-
(GPs) and diabetes nurse respectively. The check- mary or secondary care level. This means that
ups also include patient education provided by virtually all patients with type 1 diabetes have
GPs and nurses. Patient education is given both their care given at hospitals while patients with
individually and as group-based meetings four to type 2 diabetes are managed in primary care.
six times in groups of six to eight patients. The Diabetes care in Sweden is mainly decentral-
latter are normally led by a diabetes nurse. The ised to PHCCP but with close connection to both
providers follow national standard guidelines home care and specialised diabetes units at the
from the National Board of Health and Welfare, hospitals. To maintain comparable quality across
updated February 2015. These guidelines contain the country, Sweden has emphasised participa-
a total of 140 recommendations, over 50 of which tion in the NDR with an annual quality registra-
are of particular significance to the finances and tion. NDR enables comparison with other units
organisation of the health service and to ensuring and deviations from national goals are easily
that people with diabetes receive a consistently detected. EMR are used by the majority of care
high standard of care [43]. The NDR plays a key providers and are linked to the NDR as well as to
role in the assessment of diabetes care as it pro- other national registers. EMR are also used for
vides health-care professionals individualised referrals within or between county councils. For
information for each patient to help establish tar- the most common diagnosis or complications
gets for, e.g., HbA1c. However, despite the find- Sweden has created quality indicators and all
ings of Euro Diabetes Index 2014, new research units shall try to achieve these goals. The most
shows inequalities in terms of resource allocation costly care takes place at the specialised units at
and implementation of organisational features the hospital where patients with severe complica-
within Swedish diabetes care in primary health tions are treated. These patients also have needs
care [43, 44]. from primary care and home services.
Coordination activities are needed to optimise
resource use. For this purpose, chains of care
Summary have been developed and established in order to
facilitate collaboration between different levels
The Swedish health-care system has been built of care.
with a strong orientation towards subnational lev-
els, i.e., municipalities and regions. The responsi-
bility is shared between the central government,
county councils and municipalities. Health care References
is mainly tax funded both at the Government
1. Rttsntet. Hlso- och sjukvrdslag (1982:763) n.d.
level and at the level of the county councils and [05/15/2016]. Available from: https://www.notisum.
municipalities. The health system is highly se/Pub/Doc.aspx?url=/rnp/sls/fakta/a9820763.htm.
2. Sveriges kommuner och Landsting. Swedish 18. OECD. Sweden has excellent health care but must
Association of Local Authorities and Regions n.d. improve care co-ordination Press release 2013
[05/15/2016]. Available from: http://skl.se/tjanster/ [05/15/2016]. Available from: http://www.oecd.org/
englishpages.411.html. newsroom/sweden-has-excellent-health-care-but-
3. Svensk frfattningssamling. Lag om ndring i hlso- must-improve-care-co-ordination.htm.
och sjukvrdslagen (1982:763). 2009. 19. Bidgood E. Health care systems: Sweden & local-
4. Anell A, Glenngard AH, Merkur S. Sweden health ism an example for the UK? : Institute for the Study
system review. Health Syst Transit. of Civil Society (Civitas); 2013 [10/5/2015]. Available
2012;14(5):1159. from: http://civitas.org.uk/2013/10/22/new-report-
5. Hanning M, Schmidt I. Peer Review Health System the-merits-of-swedish-style-localism-for-improving-
Performance Assessment (Brussels, 1920 May uk-healthcare/.
2014) Sweden: SB Consulting; 2014 [05/15/2016]. 20. OECD. Waiting times. In: OECD, editor. Health at a
Available from: http://ec.europa.eu/social/keyDocu- Glance 2011: OECD Indicators: OECD Publishing;
ments.jsp?advSearchKey=Peer+Review+Health+Sys 2011.
tem+Performance+Assessment+&mode=advancedSu 21. Vogel H. Health Care Guarantee 2015 [05/15/2016].
bmit&langId=en&policyArea=&type=0&country=12 Available from: http://www.1177.se/Other-languages/
&year=2014. Engelska/Regler-och-rattigheter/Vardgaranti/.
6. N.N. The Swedish Health Care System The 22. Cebolla Garrof B, Bjrnberg A, Yung Phang A. Euro
Organization and Delivery of Care n.d. [05/15/2016]. Diabetes Index 2014 Health Consumer Powerhouse
Available from: http://fall09hpm101sweden.provi- Ltd; 2014.
dence.wikispaces.net/The+Organization+and+Delive 23. NHS Lanarkshire. Managed Clinical Networks 2009
ry+of+Care. [05/15/2016]. Available from: http://www.nhslanark-
7. Government Offices of Sweden. Ministry of Health shire.org.uk/Services/MCN/Pages/default.aspx.
and Social Affairs 2014 [05/15/2016]. Available from: 24. Curry N, Ham C. Clinical and service integration
http://www.government.se/government-of-sweden/ the route to improved outcomes. London: The Kings
ministry-of-health-and-social-affairs/. Fund; 2010.
8. Swedish Institute. Health care in Sweden 2015 25. Ham C, Curry N. Integrated care what is it? Does it
[05/15/2016]. Available from: https://sweden.se/soci- work? What does it mean for the NHS? London: The
ety/health-care-in-sweden/. Kings Fund; 2011.
9. Robertson H. Integration of health and social care. A 26. Burgess L. Integration of Health and Social Care:
review of literature and models. Implications for International Comparisons 2012 [05/15/2016].
Scotland. Prepared for the RCN in Scotland. 2011. Available from: http://www.scottish.parliament.uk/
10. hgren B. Chain of care development in Sweden: ResearchBriefingsAndFactsheets/Factsheets/
results of a national study. Int J Integr Care. SB_12-48.pdf.
2003;3:e01. 27. hgren B, Axelsson R. A decade of integration and
11. OECD. How does health spending in Sweden com- collaboration: the development of integrated health
pare? 2015 [05/15/2016]. Available from: http://www. care in Sweden 20002010. Int J Integr Care. 2011;11
oecd.org/els/health-systems/Country-Note-Sweden- Spec Ed:e007.
OECD-Health-Statistics-2015.pdf. 28. hgren B, Axelsson R. Determinants of integrated
12. OECD. Health Expenditure n.d. [11/27/2015]. health care development: chains of care in Sweden.
Available from: http://www.oecd.org/els/health- Int J Health Plann Manag. 2007;22(2):14557.
systems/health-expenditure.htm. 29. Trgrdh B, Lindberg K. Curing a meagre health care
13. Schlkopf M, Pressel H. Das Gesundheitswesen system by lean methods translating chains of care
im internationalen Vergleich: in the Swedish health care sector. Int J Health Plann
Gesundheitssystemvergleich und europische Manag. 2004;19(4):38398.
Gesundheitspolitik. 2. aktualisierte und erweiterte 30. Statistiska centralbyrn. Life expectancy by sex
Auflage ed. Berlin: MWV Medizinisch 19602014 and forecast 20152060 2015
Wissenschaftliche Verlagsgesellschaft; 2014. 300 S p. [05/15/2016]. Available from: http://www.scb.se/en_/
14. OECD. Health at a Glance 2013 OECD indicators. Finding-statistics/Statistics-by-subject-area/
Paris: OECD Publishing; 2013. Population/Population-projections/Population-
15. OECD. Health at a glance 2015: OECD indicators. projections/Aktuell-Pong/14505/Current-forecast/
Paris: OECD Publishing; 2015. The-future-population-of-Sweden- 20152060/
16. Kelley E, Hurst J. Health care quality indicators proj- 273436/.
ect conceptual framework paper OECD Health 31. Statistiska centralbyrn. Swedens Population by sex
Working Paper No 23 DELSA/HEA/WD/ and age on 31/12/2014 2015 [Available from: http://
HWP(2006)3 OECD; 2006. www.scb.se/en_/Finding-statistics/Statistics-by-
17. OECD. OECD reviews of health care quality: subject-area/Population/Population-composition/
Sweden executive summary, assessment, and rec- Population-statistics/Aktuell-Pong/25795/
ommendations. 2013. Yearly-statistics--The-whole-country/262459/.
32. Statistiska centralbyrn. Population by age 1960 och islet autoantibodies in childhood and adolescent type
2014 and forecasts 2060 2015 [05/15/2016]. 1 diabetes. Pediatr Diabetes. 2013;14(2):97105.
Available from: http://www.scb.se/en_/Finding- 39. Carlsson A, Kockum I, Lindblad B, Engleson L,
statistics/Statistics-by-subject-area/Population/ Nilsson A, Forsander G, et al. Low risk HLA-DQ and
Population- projections/Population-projections/ increased body mass index in newly diagnosed type 1
Aktuell- Pong/14505/Current-forecast/The-future- diabetes children in the better diabetes diagnosis study
population-of-Sweden-20152060/273430/. in Sweden. Int J Obes (Lond). 2012;36(5):71824.
33. International Diabetes Federation. Sweden 2015 40. SWEDIABKIDS Nationellt register fr barn-och
[11/24/2015]. Available from: http://www.idf.org/ ungdomsdiabetes. rsrapport 2014 2015
membership/eur/sweden. [01/17/2016]. Available from: https://swediabkids.
34. Jansson SP, Fall K, Brus O, Magnuson A, Wandell P, ndr.nu/Documents/NDR-Child/AnnualReport-2014.
Ostgren CJ, et al. Prevalence and incidence of diabe- pdf.
tes mellitus: a nationwide population-based 41. Rawshani A, Landin-Olsson M, Svensson AM,
pharmaco-epidemiological study in Sweden. Diabet Nystrom L, Arnqvist HJ, Bolinder J, et al. The inci-
Med. 2015;32(10):131928. dence of diabetes among 034 year olds in Sweden:
35. Vrdanalys. The effects of freedom of choice in new data and better methods. Diabetologia.
healthcare 2012 [05/15/2016]. Available from: 2014;57(7):137581.
http://www.vardanalys.se/Support/In-English/ 42. Nationalla Diabetesregistret (NDR) Registercentrum
Reports1/The-effects-of-freedom-of-choice-in- Vstra Gtaland. rsrapport 2014 rs resultat 2015
health-care/. [05/15/2016]. Available from: https://www.ndr.nu/#/
36. Vrdanalys. Who has benefited from choice in health- arsrapport.
care? 2013 [05/15/2016]. Available from: http://www. 43. Socialstyrelsen. National Guidelines for Diabetes
vardanalys.se/Support/In-English/Reports1/ Care support for governance and management 2015
Who-is-benefiting-from-choice-in-healthcare/. [05/15/2016]. Available from: http://www.social-
37. Tillman C, Beckman A. Smre lkarkontinuitet efter styrelsen.se/publikationer2015/2015-4-12.
vrdvalet fr patienter med diabetes. Lakartidningen. 44. Swedish Association of Local Authorities and Regions
2015; 112 pii: DFDZ. (SALAR). Success factors in Swedish Diabetes Care
38. Andersson C, Vaziri-Sani F, Delli A, Lindblad B, 2015 [05/15/2016]. Available from: http://webbutik.
Carlsson A, Forsander G, et al. Triple specificity of skl.se/bilder/artiklar/pdf/5388.pdf?issuusl=ignore.
ZnT8 autoantibodies in relation to HLA and other
for Adults with Diabetes: A Patient 13
Organisation Perspective
Heather Bird and Bridget Turner
What We Mean by Integrated Care: The Need to Improve Outcomes

A Patient Perspective for People with Diabetes in the UK
Integrated care is an approach that seeks to The need to improve outcomes for people with
improve the quality of care for people with diabe- diabetes in the UK is clear.
tes by ensuring that services are well coordinated There are currently 3.3 million people who
around their needs. Integrating care around the have a diagnosis of diabetes in the UK [2]. Each
needs of the person with diabetes must be about year, 20,000 people die prematurely from diabe-
improving outcomes and at a wider level tes [3]. It is responsible for more than 100 ampu-
reducing the incidence of complications and the tations a week, is the leading cause of preventable
proportion of people who do not feel equipped to sight loss in people of working age and a major
self-manage their condition. This can only be contributor to kidney failure, heart attack and
achieved by designing a system that focuses on stroke [46].
the patients perspective of care, enables health- Too many people with diabetes are still not
care professionals and people with diabetes to getting all the checks that they need. Too few are
work in partnership and supports the individual being offered or are accessing diabetes education
to take control and self-manage their condition. to support self-management. Large numbers of
Integrated care is about maximising the value people with diabetes have not received or had
both healthcare professionals and people with access to emotional or psychological support
diabetes get from the time they spend in routine when needed [7].
consultation [1]. It is about making sure that peo- We know that people living with diabetes
ple are supported and able to self-manage their face daily challenges managing their condition.
condition and about people being able to access Such as, diet and exercise, treatment-taking,
the right care as soon as they need to. Integrated psychological stress, education, illness and dis-
diabetes care is underpinned by vertical struc- ability. We also know that people are not as
tural integration between primary, community engaged in their care as they might be. A study
and specialist care services. by the Royal College of General Practitioners
(RCGP) found a lack of engagement in consul-
tations, including a failure to attend follow-up
appointments. It has also been found that up to
H. Bird (*) B. Turner 50 % of the medication prescribed for manag-
Diabetes UK, London, UK ing long term conditions is not taken, or is not
e-mail: [email protected]; Bridget.
[email protected] taken as prescribed [8].

216 H. Bird and B. Turner
The English Health System: finances and, in addition, regulation of services is

Understanding the Context at an organisation level rather than across a path-
for Integrated Care way of care [11].
The National Health Service (NHS) in the UK is

publicly funded and free at the point of use. The Need for Change
Health is a devolved responsibility within the
UK. So, the structure of the service in England, The boundaries between primary and specialist
Wales, Scotland and Northern Ireland is differ- services are reinforced by payment systems, the
ent. The following is an explanation of the physical location of healthcare providers and the
English system. separation of responsibility for the commission-
Overall political responsibility rests with the ing and provision of services [12]. This is partic-
Secretary of State for Health, supported by the ularly problematic for delivering integrated
UK Department of Health. There have been a diabetes care because diabetes is such a complex
number of reconfigurations, following the cre- condition and people need access to a range of
ation of the NHS in 1948. A purchaser provider healthcare professionals across specialist, com-
split (i.e. service commissioning and provision munity and primary care.
were separated into different organisational The need to join up health services across the
structures) was introduced in 1991 to develop an National Health Service (NHS) to centre on the
internal market within the NHS [9]. needs of patients rather than the needs of the sys-
Commissioning is the planning and purchas- tem is well recognised in national policy. The
ing of NHS services. Since 1991 it has been car- NHS Mandate, issued annually by the Department
ried out by organisations acting as purchasers of of Health, challenges the NHS to improve the
services from healthcare providers. From 2013, provision of care to ensure it is coordinated
NHS England has been commissioning primary around the needs, convenience and choices of
care services, through its Area Teams. At a local patients, rather than the interests of organisations
level, 211 Clinical Commissioning Groups that provide care [13]. This was more explicitly
(CCGs) have responsibility for commissioning called for in The NHS Five Year Forward View.
services across a range of clinical areas (hospital Published in October 2014, this is the vision for
services, emergency and urgent care, ambulance the NHS in England and explains the need for
services and community services) [10]. The services to change to meet the needs of patients
CCGs report to NHS England, again through its and be sustainable over the next 5 years. In par-
Area Teams. ticular, this document encourages local areas to
In the UK, diabetes care is delivered by a dissolve traditional boundaries between primary,
range of providers. There can be local variation community and specialist services to improve the
in the way services are delivered but in general, way care is delivered and enable clinicians to
people with diabetes will be registered with a focus on delivering personalised care [14].
general practitioner working in primary care.
Primary care includes general practitioners and
practice nurses. Access to specialist care is Supporting the Patient:
through referrals by general practitioners who The Importance of Supported
serve as gate keepers. Specialists work in mul- Self-Management
tidisciplinary teams, with links to allied health
services such as podiatry who may be based The input and skills of healthcare professionals
in a hospital or community setting. There are dis- across primary, community and specialist care
tinct funding streams for primary care, specialist are essential to provide high quality care for peo-
care and sometimes community care services. ple with diabetes, which meets their individual
Each organisation is responsible for their own needs and enables them to be engaged in their
13 Integrated Diabetes Care for Adults with Diabetes: A Patient Organisation Perspective 217
own care. Integrated care is about making sure an annual surveillance review, less than half of
that the time people spend with healthcare pro- people with diabetes felt that this time was being
fessionals is joined up and it is clear to the indi- maximised to facilitate discussion between them
vidual how each aspect of care contributes to and clinicians that would support them to self-
supporting them to live their life with manage. The study therefore looked at how to use
diabetes well. this time more effectively to make sure discus-
Part of this is about making sure that people are sions focused on the needs of the person with dia-
seen by the appropriate healthcare professional to betes and supported them in their day to day
manage their condition at the right time. For management of their diabetes [1].
example, in general, people with Type 1 diabetes
should have their care coordinated by a multi-dis-
ciplinary specialist team which includes a con- Changing Attitudes to Focus
sultant diabetologist [15]. Everybody with diabetes on the Person with Diabetes
should have straightforward, rapid access to spe-
cialist care services based on needs. A key focus of the Year of Care was demonstrat-
It is also about making sure that the time a per- ing the need for a fundamental change in the way
son with diabetes spends in consultation with a healthcare professionals and people with diabetes
healthcare professional supports their ability to interact with each other and the system to man-
self-manage their condition. It is estimated that age the individuals diabetes. The pilot showed
99 % of diabetes care is self-management and, as that healthcare professionals must abandon the
outlined above, people living with diabetes face traditional approach to care delivery, where they
daily challenges [16]. Although, on average, peo- are the primary decision maker, in favour of a
ple with diabetes spend approximately only 3 h a partnership model, working with the person with
year with healthcare professionals (versus 8757 h diabetes [18]. At the same time, people with dia-
of self-management) this time can be used effec- betes need to take an active role in determining
tively to support self-management [16]. This has their care and support needs. This is not straight-
most clearly been shown through the Year of forward, due to the diversity of need and the
Care pilot, which ran in three pilot sites, conclud- degree of engagement people with diabetes have
ing in 2011 [1]. The Pilot ran for 3 years and said they want in their care, but it is essential in
focused on the delivery of routine care to make every case to improve individual outcomes [17].
sure that the annual review appointment was The partnership approach should focus on identi-
transformed into a meaningful and constructive fying the individuals priorities and goals. It is
discussion between the person with diabetes and about making healthcare relevant to the life and
their healthcare professional. The pilot showed circumstances of the person with diabetes and
support for self-management to be an effective identifying additional aspects of support, such as
intervention in improving engagement among peer support, that can help to deliver the out-
people with diabetes. In Tower Hamlets, as a comes they want to achieve.
pilot site in the Year of Care, care planning was The 3 year pilot was split into two phases: fea-
implemented in 31 out of 32 practices, over 90 % sibility and implementation. The Year of Care
of patients had an annual care and support plan- offered a practical approach that engaged staff at
ning review and patients reported that involve- a practice level, and involved patients through
ment in care rose from 52 % to 82 % over the participation events in pilot sites. Local engage-
course of the pilot [17]. In addition, in Tower ment and ownership is key to implementing
Hamlets, the proportion of people achieving an change but a challenge for the Year of Care was
HbA1c of less than 7.5 % steadily improved from to ensure healthcare professionals embraced the
36 % to 55 % over the same time period. The pilot philosophy of partnership working and did not
recognised that although the UK is comparatively just implement the practical steps involved in the
successful in ensuring people with diabetes have new way of working, e.g. sending out test results
prior to consultations, as an add-on to the exist-

ing approach. Ongoing training that reacted to
this challenge was a central part of the Pilot [1].
Practices were incentivised to participate in
the project and each site committed resources but
no further extra funding was available. The Year
of Care found the care planning approach to be
cost neutral once it was up and running. This cost
did not include the large amount of additional
work that practices had to do to change the way
they worked, and for introducing changes to the
IT systems to properly support care planning [1].
At its conclusion, people with diabetes
reported an improved experience of care and real
changes in their self-management. Healthcare
Fig. 13.1 The house of care framework (Used with per-
professionals also reported improved knowledge mission from Year of Care: Report of findings from the
and skills, and greater job satisfaction [18]. pilot programme. June 2011. Year of Care Partnership.
Diabetes UK/NHS Diabetes; 2011)
The Importance of the House

of Care Commissioners should work together to make
sure that all aspects of the diabetes pathway are
The house of care is a metaphor introduced through being delivered by the appropriate provider, with
the Year of Care to illustrate the fact that all parts of well-defined referral pathways between services.
the system, people with diabetes and healthcare Within an integrated diabetes pathway local ser-
professionals must be aligned in support of the vices should have clear protocols for who does
partnership approach to care [19]. See Fig. 13.1. what and what services are provided where.
In the Year of Care pilot, the house of care ini- Within the full pathway of diabetes care, as out-
tially explained the engagement needed for this lined below, care pathways and referral pathways
but it has subsequently been applied much more for particular aspects of care, such as diabetes
widely as a framework, explaining how a whole foot care, should be defined to enable delivery of
system might adapt to deliver collaborative care the right care, in the right place and at the right
planning to support integrated care [20]. time. Figure 13.2 illustrates the care pathway for
Delivering the philosophy of the house of care diabetes.
requires localities to define and agree their model Although the various services that make up
of care and approach to delivery, with complete the pathway are constant, how the pathway is
commitment from all involved. delivered in each locality can vary and should be
Commissioning all the aspects of the diabetes defined in a local model of care. All areas in the
care pathway is the foundation in the house of UK should have a well-defined model of diabetes
care metaphor and provides the basis on which an care. Broadly, a model of care describes the way
integrated system can be implemented. health services are configured [22]. It can be
applied to health services delivered in a provider
or organisation, within a team or across a whole
The Diabetes Care Pathway local system of care. For diabetes, it will need to
set out the care to be commissioned and delivered
The provision of integrated care for people with by provider organisations defining who does
diabetes must be underpinned by access to all what, where and how. In any configuration, effec-
aspects of the diabetes care pathway [21]. tive delivery must be underpinned by partnership
Fig. 13.2 The diabetes care pathway (Used with permission from Year of Care: Report of findings from the pilot pro-
gramme. June 2011. Year of Care Partnership. Diabetes UK/NHS Diabetes; 2011)
working between generalists and specialists to the methods and processes are in place to facili-
support and deliver care whether it is provided in tate integrated care [25].
a general practice, in a community setting or in
the hospital.
The National Clinical Director for obesity and Supporting Integrated Care
diabetes in England has produced a Diabetes Through the Integration of Systems
Sample Service Specification Fig. 13.3 [23]. This and Processes
outlines the provision of high quality care for all
those with diabetes and, as appropriate, differen- The diabetes community in the UK have identified
tiates the care needs of adults with Type 1 diabe- five integration enablers that must be in place to
tes (T1DM) and those with Type 2 diabetes support the delivery of integrated diabetes care
(T2DM). It describes all the services needed to [26]. Within the house of care framework four of
provide a complete care pathway for people with the enablers are organisational processes and local
diabetes, including those with the long-term infrastructure which need to be in place to support
complications of diabetes. It also meets the the development of integrated care focused on the
National Institute for Health and Care individual with diabetes, through care planning (at
Excellences (NICE) Quality Standard for diabe- the centre of the house). This left and right walls of
tes (QS6), which defines clinical best practice in the house need to be in place to deliver care plan-
the delivery of diabetes care [24]. ning whereby engaged and informed individuals
As well as making sure that all the compo- and carers and healthcare professionals are com-
nents of the pathway are in place commissioners mitted to partnership working [19].
must make sure the appropriate support for The five distinct but mutually reinforcing inte-
delivery is commissioned. This includes ongo- gration enablers are:
ing training and education to assure staff com-
petency in diabetes as well as training for 1. Integrated IT
healthcare professionals to enable them to work 2. Aligned finances and responsibility
in partnership with people with diabetes through 3. Care planning
care planning. It is also about making sure that 4. Clinical engagement and leadership
5. Robust clinical governance [26]
Fig. 13.3 The diabetes sample service specification (Used with permission from Valabhji J). [23]
Integration Enablers to Facilitate This was driven by the enthusiasm for change of
Integrated Care GPs and consultant diabetologists locally. The
redesign saw care delivery supported by a single
Wherever possible, localities should make sure IT system and a single budget. Evaluation of
all the enablers are in place. Although they can be this model found that in the first 6 months, there
introduced independently and drive some was a reduction of mean length of inpatient stay
improvement, it is evident that they are mutually from 11 to 6.5 days in patients with a primary
reinforcing so the ability of an area to provide diagnosis of diabetes from participating GP
integrated care is further enhanced when the practices compared with the same time period
enablers are developed at the same time to sup- 12 months previously.
port the delivery of integrated care. Formal feedback was obtained from patients
Where these enablers are in place to support annually using a questionnaire which explored
the delivery of integrated diabetes care, early their experiences in the service using a 5 point
evidence suggests patient experience is score with 5 being excellent and 1 poor. The
improved and the cost efficiency of the service questionnaire was administered to 50 users in
increased. Derby introduced a new model of 2012. The process was repeated in 2013, but this
delivering diabetes care in 2009 commis- time the survey was administered by users
sioned by the Primary Care Trust (the body instead of administrators. The results were
responsible for commissioning at the time). similar. Overall:
85 % rating the service as excellent or very appointment was in primary or specialist care, to
good optimise care and make the referral process more
70 % felt that the service was excellent or very efficient [28].
good compared with their previous care An information system, where GPs and spe-
80 % felt the waiting time between referral cialists can see the same record, can be used to
and first appointment was excellent or very automatically identify and target at risk patients
good. [29]. This is the approach being used in
90 % felt that the service was able to help Wolverhampton. Wolverhampton has developed
them with the problem they were referred with a system focused on delivering integrated diabe-
95 % stated they would recommend the ser- tes care for many years, which supports the
vice to another person with diabetes. development of primary care led diabetes ser-
vices. Specialist care is delivered in partnership
with primary care to meet the clinical needs of
Understanding the Five Enablers the patient. The model of care is based on self-
care through education, patient centeredness and
Integrated IT Systems empowerment. As part of their model of care, the
Clinical Commissioning Group (CCG) (explained
Why Integrated IT Is Important earlier in this chapter) as the relevant commis-
In 2013, National Voices (a coalition of health sioning body has introduced a central portal
and social care charities) was commissioned by (Graphnets Care Centric Portal). This is used to
NHS England to produce a narrative for person- extract data from 49 GP practices. These data are
centred coordinated care [27]. This document is fed into the trusts Diabeta3 system and a locally
focused on understanding what integrated care developed algorithm stratifies patients according
looks like from a patient perspective, and uses a to risk [30]. Patients are rated against the NICE
series of I statements to do this. In this, patients recommended nine diabetes care processes
stress the importance of being able to tell their (Table 13.1) and based on their risk status for
story once, without needing to repeat themselves micro and macro vascular complications of dia-
across the system [27]. A culture change to betes they are flagged as red, amber or green. The
remove boundaries between organisations and results are then used to decide where care should
encourage meetings and dialogue between clini- be provided to that patient along the pathway and
cians goes some way to facilitate this. Optimal what should be done to improve care for that par-
information sharing, however, should be under- ticular person [30]. Using data in this way allows
pinned by an information system that provides the clinicians to operate outside of the constraints
clinicians across primary, community and spe- of the system, enabling them to identify patients
cialist care with a patients clinical record regard- most at need and enable treatment at an early
less of setting. point and in the right setting, removing delays
Integrated IT: In Practice

The ideal is for all providers in a pathway to use Table 13.1 Nine care processes
the same system. For example, in the Derby 1. Blood glucose level measurement
model all GP practices and the hospital in Derby 2. Blood pressure measurement
use SystmOne (TPP). There can be initial frustra- 3. Cholesterol level measurement
tions with this, as clinicians must get consent 4. Retinal screening
from patients to share their data and the neces- 5. Foot and leg check
sary data sharing and governance structures need 6. Kidney function testing (urine)
to be introduced. However, once the system is 7. Kidney function testing (blood)
fully established clinicians can see a patients 8. Weight check
records, regardless of whether their previous 9. Smoking status check
in referral and looking to reduce emergency by consultant specialists [31]. This saw the con-
admissions. sultants focus on super-specialist areas of diabe-
tes care in the hospital and refer all other care,
which is was felt did not need to be managed
Aligned Finances and Responsibility exclusively by specialists, to community and pri-
mary care. This model is based on an increased
Defining Who Does What role for primary care in the management of dia-
Clearly defining responsibility for all aspects of betes, and required the engagement of GP prac-
service delivery is essential for delivering the full tices within the local area. For this approach to
diabetes care pathway. The pathway should be work it is essential that all healthcare profession-
clearly defined in the model of care, and the als have the skills, clinical support and infrastruc-
infrastructure supporting the delivery of care ture necessary to be able to provide high quality
should ensure that all clinicians are able to deliver diabetes care. Under the new alignment, the dia-
against national standards. As illustrated by the betologists in Portsmouth were given two func-
Diabetes Sample Service Specification, the spe- tions: to continue in their role as deliverers of
cific care needs of people with Type 1 diabetes specialist care; and a new function as healthcare
should be drawn out in the configuration of ser- professional educators [32]. All the GP practices
vices [23]. As stated above, in general, support involved in the super six initiative have virtual
for people with Type 1 should be coordinated by access to consultant support (telephone and
a multi-disciplinary diabetes specialist team. The email) and each practice is visited by a diabetes
team can be based either in hospital or in a com- specialist nurse and consultant biannually to
munity setting. This is because managing Type 1 deliver training and support as needed. This is
diabetes is complex and requires significant accompanied by a programme of accredited
expertise, and there can be serious consequences training [33]. There are indications that clinical
if things go wrong [15]. At the same time people outcomes have improved since 2011 when the
with complex Type 2 diabetes will also need new service model was introduced. For example,
timely access to specialist care. The ongoing care the hospitals diabetic hypoglycaemia admis-
of people with Type 2 diabetes is generally pro- sions fell from 224 to 198 between 2011/2012
vided by GPs with clear systems in place for call and 2013/2014 [34].
and recall, regular review and referral when In North West London, a pilot was started in
necessary. 2011 aimed at improving the delivery of diabetes
A common feature that has preceded change care and care for older people in North West
in areas that have looked at improving the deliv- London. This pilot focused on better coordinat-
ery of diabetes care across the whole pathway is ing good practice to enable clinicians to work
an unacceptable variation in the quality of care efficiently across provider boundaries. Investment
(particularly across primary and community was made in IT and the model of care introduced
care) and a lack of consistency in the delivery of a stratification process which segments people
services. with Type 2 diabetes according to need and refers
In Portsmouth, for example, the diabetes clini- them to the appropriate part of the system. The
cal lead at Portsmouth Hospital, a GP with spe- multidisciplinary group structure provides GPs
cial interest in diabetes and the commissioning with direct access to specialist knowledge links
managers developed a proposal for change. The which had previously not been made to discuss
initiatives focused on clarifying the role of the complex cases and develop their skills [35]. Over
consultant diabetologist in the delivery of diabe- three quarters of professionals surveyed, as part
tes care. The model of care defines who does of the evaluation of the pilot, reported that this
what within the system and is widely known as new structure enhanced inter-professional work-
the super six. The super six are the areas of ing and levels of professional knowledge [36].
diabetes care that it was agreed must be managed No outcomes data have been released.
The Role of Payment Systems attain endorsement as a GP practice with an

in Supporting the Delivery of Care interest, and high level of competency, in the
The implementation and use of payment systems delivery of diabetes care [28].
must support the delivery of services as explained
in the local model of care. In a traditional model
of diabetes care the rigid divide between primary Care Planning
and specialist care is exacerbated by the provi-
sion of funding. For example the tariff system Through the Year of Care pilot, care planning has
used in England, to pay for activity in specialist been shown to be an effective way of engaging
care, can create incentives for one part of the sys- people with diabetes and healthcare professionals
tem to hold on to patients who might be better in a way of working which effectively supports
treated elsewhere. This system was originally the individuals ability to self-manage [1].
introduced to incentivise providers to increase Care planning is a continuous process in
elective activity and reduce waiting times. Whilst which clinicians and patients work together to
it has been effective for improving access to agree goals, identify support needs, develop and
planned care, it is not so well suited to achieving implement action plans and monitor progress
the continuity of care needed to facilitate inte- [18]. People with diabetes should have active
grated care [37]. involvement in the care planning process of
Going forward, a key commitment in The deciding, agreeing and owning how their diabe-
NHS Five Year Forward View is a central focus tes will be managed [1].
on introducing national flexibilities in the cur- For care planning to be truly collaborative, the
rent regulatory, funding and pricing regimes to person with diabetes must be engaged in the pro-
assist local areas to transition to better models of cess and allowed time to think about his/her own
care [13]. Within the current constraints of pay- priorities. In Wolverhampton the care planning
ment systems, however, there is an acceptance process is initiated through a questionnaire sent
that the providers involved need to focus on the to patients prior to their annual review appoint-
needs of the whole health economy rather than ment, which includes a list of questions for them
their own organisation. to consider and identify their priorities. This is
discussed at their consultation and an action plan
Using Payment Systems to Support based on this is designed in collaboration with
Integrated Care: In Practice the clinician to inform their ongoing care [39].
Derby introduced a new NHS organisation,
which held a pooled budget to deliver diabetes
care based on historical prices across the care Clinical Engagement and Leadership
pathway [28].
North West London had a set budget to sup- Engaging the Right People
port the delivery of their pilot project. This in Designing a Model of Care
budget was used to compensate providers for To maximise the chances of the model of diabe-
the time spent working on additional features tes care meeting the needs of people with diabe-
of the pilot. For example, GPs were paid for tes and healthcare professionals all relevant
the time spent attending multi-disciplinary stakeholders (Table 13.2) should be engaged
group meetings and putting together care collaboratively in discussion at an early point.
plans [38]. The North West London integrated care pilot
Portsmouth, Derby and Wolverhampton have clearly illustrated the central importance of this
introduced a locally enhanced service pay- [40]. The initial meeting to develop the ICP was
ment which is paid to GP practices to incentiv- attended by the Chief Executive of Imperial
ise attendance at diabetes training courses and Hospitals NHS Trust, diabetologists, diabetes
interaction with the diabetes specialist team to leads, commissioners, representatives of Central
London Community Healthcare Trust, GP leads, help determine priorities such as education and
psychiatrists, Diabetes UK and Age UK. Chief training needs. People with diabetes were there-
executives and senior managers gave the pilot fore included on all the groups and an additional
their backing from an early point, giving clini- patient and users committee was established to
cians the financial and managerial support neces- discuss the progress of the pilot as a whole [41].
sary to enable them to better focus the model of
diabetes care on the needs of people with diabe-
tes. The trust chief executives were particularly Clinical Governance
supportive of the diabetologists spending time
away from the hospital to work more closely with What Is Clinical Governance and Why It
colleagues across the pathway [41]. Is Important
Clinical governance is a system through which
Getting Clinicians Support NHS organisations are accountable for continu-
One of the most significant challenges for North ously improving the quality of their services and
West London was getting clinicians from all pro- safeguarding high standards of care by creating
viders involved and supportive of the pilot. an environment in which excellence in clinical
Initially, clinicians were concerned about the care will flourish [42].
challenges the pilot posed to their position and Implementing a clear and effective clinical
current way of working. These concerns were governance structure helps to align the ambitions
gradually overcome as an external chair was of clinicians with those of commissioners and
appointed to lead the pilot and bring people most importantly with people with diabetes. An
together on equal terms. The number of clini- integrated system which removes barriers
cians attending the meetings to develop the pilot between care providers and overcomes perverse
steadily grew as people got to know one another financial incentives allows services and the peo-
and unite behind the aim of improving the service ple involved to align and take shared responsibil-
for people with diabetes (41). ity for a single goal [42].
Engaging People with Diabetes Implementing Clinical Governance

In the initial developmental discussions about the In Derby and North West London establishing a
pilot, people with diabetes were represented by structure of clinical governance for their respec-
Diabetes UK. As the pilot developed it was tive initiatives was core in their drive to improve.
agreed that people with diabetes must have a They took different approaches to this, reflecting
direct role throughout the structure of the pilot to the scale and complexity of the project.
In the Derby model a clinical board structure
Table 13.2 Initiating change: who to involve has been established to oversee the model of
care. The board has responsibility for ensuring
Local diabetes network, to include:
the quality of the service delivered to patients. As
People with diabetes and groups representing them
Healthcare professionals from the full range of
a new NHS organisation was set up as a company
relevant specialties limited by shares in Derby, the two shareholders
From primary, specialist and community care (the group of GP practices and the Derby Hospital
CCG clinical lead NHS Trust) were directors of the board. The
Area Team representative directors had responsibility for holding the board
Additionally: to account for delivery of the commissioned ser-
Senior managers (inc. hospital chief execs) vice specification and the financial state of the
Trust finance managers company. In addition to the directors, the clinical
Trust IT leads board is made up of consultant diabetologists,
Medicines management GPs, practice nurses, diabetes specialist nurses,
Existing network (e.g. SCN) representatives dietitians, service managers, consultant ophthal-
mologist and people with diabetes. An additional Health and Social Care Hospital Episode Statistics
2007/82011/12. London: HSCIC; 2013.
patient board meet every 2 months to discuss
5. Liew G, Michaelides M, Bunce C. A comparison of
how to improve the service [28]. the causes of blindness certifications in England and
Wales in working age adults (1664 years). 1999
2000 with 20092010. BMJ Open. 2014;4:e004015.
6. Morrish NJ, Wang SL, Stevens LK, Fuller JH, Keen
Conclusion H. Mortality and causes of death in the WHO multina-
tional study of vascular disease in diabetes.
Integrated care must be about improving the way Diabetologia. 2001;44 Suppl 2:S1421.
diabetes care is delivered to better meet the needs 7. Health and Social Care Information Centre. National
Diabetes Audit 201213. Report 1: care processes and
of people with diabetes.
treatment targets. London: HSCIC; 2014.
Healthcare systems are complex and improve- 8. Mathers N, Roberts S, Hodkinson I, Karet B. Care
ment requires leadership to orchestrate all the planning: improving the lives of people with long
items, people and processes to work together term conditions. London: Clinical Innovation and
Research Centre; 2011.
toward a common purpose [43]. Achieving inte-
9. UK Parliament. Health Committee. Commissioning
grated care needs the full commitment of com- 19482010 [internet]. 2010. Available from: http://
missioners and providers to work together, in www.publications.parliament.uk/pa/cm200910/cmse-
partnership with people with diabetes to realign lect/cmhealth/268/26805.htm.
10. Holder H, Thorlby R. The new structure of the NHS in
the system.
England [internet]. 2013. Available from: http://www.
In many areas of the UK progress is being nuffieldtrust.org.uk/talks/slideshows/
made, through different approaches to delivery, new-structure-nhs-england.
design and support. It is challenging but change 11. NHS England. Understanding the new NHS. London:
BMJ; 2014.
needs to be driven locally to make sure people
12. Ham C, Dixon J, Chantler C. Clinically integrated
with diabetes have access to a whole system systems: the future of NHS reform in England. BMJ.
pathway of care, which provides excellent ongo- 2011;342:d905.
ing management and rapid access to specialist 13. Department of Health. The NHS mandate. London:
Department of Health; 2014.
services when required. This must be under-
14. NHS England. The NHS five year forward view.
pinned by the structural integration enablers of London: NHS England; 2014.
shared information systems; aligned finances and 15. Diabetes UK, Association of British Clinical
responsibility, clinical engagement and leader- Diabetologists. Care coordination for adults with
Type diabetes. London: Diabetes UK; 2015. Available
ship and robust clinical governance. Clinicians
from: https://www.diabetes.org.uk/About_us/What-
and people with diabetes need to be engaged as we-say/Specialist-care-for-children-and-adults-and-
equal partners in the management of diabetes complications/Adults-with-type-1-diabetes/.
through the ongoing care planning process. 16. Roberts S. Working together for better diabetes care.
UK: Department of Health (UK); 2007.
17. Eaton S, Roberts S, Turner B. Delivering person cen-
tred care in long term conditions. BMJ. 2015;350:h181.
References 18. Coulter A, Roberts S, Dixon A. Delivering better ser-
vices for people with long term conditions. Building
1. Year of care partnership. Year of care: report of find- the house of care. London: The Kings Fund; 2013.
ings from the pilot programme. Diabetes UK/NHS 19. Coalition for collaborative care [internet]. Available
Diabetes; 2011. Available from: http://www.yearof- from: http://coalitionforcollaborativecare.org.uk/
care.co.uk/key-documents. aboutus/house-of-care/.
2. Health and Social Care Information Centre. Quality 20. NHS Improving Quality [internet]. Building the house
and Outcomes Framework (QOF) 201314. London: of care the house of care toolkit. Available from:
HSCIC; 2014. http://www.nhsiq.nhs.uk/improvement-programmes/
3. Health and Social Care Information Centre. National long-term-conditions-and-integrated-care/long-term-
Diabetes Audit 201213. Report 2: complications and conditions-improvement-programme/house-of-care-
mortality. London: HSCIC; 2015. toolkit.aspx.
4. Health and Social Care Information Centre. National 21. Department of Health. National service framework
Diabetes Audit 2011/12: the Information Centre for for diabetes. London: Department of Health; 2001.
22. Queensland Government. Queensland health. 34. Robertson R, Sonola L, Honeyman M, Brooke B,
Changing models of care framework. 2000. Kothari S. Portsmouth and South East Hampshire dia-
23. Valabhji J. Diabetes sample service specification betes service. London: Kings Fund; 2014.
[internet]. NHS England. 2014. Available from: www. 35. Greaves F, Pappas Y, Bardsley M, Harris M, Curry N,
diabetes.org.uk/integrated-diabetes-care. Holder H, Blunt I, Soljak M, Gunn L, Majeed A, Car
24. National Institute for Care and Health Excellence. J. Evaluation of complex integrated care programmes:
Diabetes in adults quality standard. 2011. Available the approach in North West London. Int J Integr Care.
at: http://publications.nice.org.uk/diabetes-in-adults- 2013;13:e006.
quality-standard-qs6/quality-statement-6-insulin- 36. Bardsley M, Smith J, Car J. Evaluation of the first
therapy. year of the Inner North West London Integrated Care
25. Shaw S, Rosen R, Rumbold B. What is integrated Pilot. 2013. Nuffield Trust. Int J Integr Care.
care? London: Nuffield Trust; 2011. 2013;13:111.
26. NHS Diabetes. Best practice for commissioning dia- 37. Ham C, Smith J, Eastmure E. Commissioning inte-
betes services: an integrated care framework. London: grated care in a liberated NHS. London: Nuffield
NHS Diabetes; 2013. Trust; 2011.
27. Department of Health. Integrated care and support: our 38. NHS NW London. Integrated care in North West
shared commitment. London: Department of Health; 2013. London. 2012 Feb. Available from: http://www.nuff-
28. Diabetes UK. Improving the delivery of diabetes care ieldtrust.org.uk/sites/files/nuffield/daniel_elkeles_
through integration. London: Diabetes UK; 2014. integrated_care_in_nw_london_290212.pdf.
29. Ham C, Imison C, Goodwin N, Dixon A, South 39. Wolverhampton Diabetes Centre [internet]. My dia-
P. Where next for the NHS reforms? Making the case betes plan. Available from: http://www.wdconline.
for integrated care. London: The Kings Fund; 2011. org.uk/downloads/education/MyDiabetesMyPlan.
30. Todd R. Wolverhampton shares info on diabetics. pdf.
Digitalhealth.net. 24 February 2014 Sect. News. 40. Steeden A. The integrated care pilot in North West
Available from: http://www.digitalhealth.net/news/ London. Lond J Prim Care. 2012;5:811.
EHI/9246/wolverhampton-shares-info-on-diabetics. 41. Vize R. Integrated care: a story of hard won success.
31. Kar P. The super six for the acute trust; all else for BMJ. 2012;344:e3529.
primary care? Pract Diabetes. 2011;28(7):3089. 42. Scally G. Clinical governance and the drive for qual-
32. Kar P. The Super Six model of diabetes care: two ity improvement in the new NHS in England. BMJ.
years on. Diabetes Prim Care. 2013;15:2115. 1998;317:61.
33. Cummings M. Professional education programme 43. Langley G, Moen R, Nolan K, Nolan T, Norman C,
[internet]. Available from: http://www.porthosp.nhs. Provost L. The improvement guide: a practical
uk/departments/Diabetes-and-Endocrinology/ approach to enhancing organizational performance.
professional-education-programme.htm. New York: Wiley; 2009.
Training for Diabetes Integrated
Care: A Diabetes Specialist 14
Physician Perspective
from the English NHS
Anne Dornhorst
The provision of healthcare is challenging for HEE. HEE delegates the training and education
everyone, the politicians, the providers, the work- programmes, both clinical and non-clinical, to 13
force as well as the patients on the receiving end. Local Education and Training Boards (LETBs)
The provision of education for the healthcare within 13 separate areas in England [1].
workforce is equally challenging for the politi- In the United Kingdom the Secretary of State
cians, the universities, the training and regulatory for Health, a cabinet minister in the UK elected
bodies and the trainees on the receiving end. government has financial control and oversight
Clearly the medical workforce of tomorrow of NHS delivery and performance; however,
should be trained to meet the challenges of deliv- since 1998, this has been largely restricted to
ering a sustainable high quality healthcare sys- England. Today the majority of non-English
tem of the future. Currently in the UK there is a related NHS policy is devolved from the UK par-
miss-match between what the healthcare policies liament to its member parliaments, or assemblies,
require and what the workforce is actually being in Scotland, Wales and Northern Ireland, leaving
trained for. The specialist training for the long- the Department of Health (DH) responsible for
term conditions, for which diabetes is one, is a health and adult social care policy mostly in
good example of this miss-match. The last 5 England. The principals of the NHS that pledges
years in the UK and England, in particular, has a comprehensive health service, available to all
seen seismic changes in the healthcare landscape with access based on clinical need, not an indi-
including how the education of its workforce is viduals ability to pay, remains a fundamental
funded. Health Education England (HEE) has, tenet across all four UK health systems [2].
since 2015, been an autonomous national body In late 2014 NHS England published their
responsible for the education and training of the 5-year forward plan for the NHS [3]. Central to
NHS healthcare workforce, and is overall respon- this plan was commissioning new models of inte-
sible for commissioning under and postgraduate grated care that would promote different provid-
medical education. Supporting education and ers including GPs, hospital consultants and social
training for integrated care is a priority for care to work more closely together to allow more
non-elective healthcare to move out of secondary
care back into the community and reduce unnec-
essary hospital admissions. A drive shared across
A. Dornhorst (*) all political parties. However, current medical
Department of Diabetes and Endocrinology,
training is not aligned to this. For example, the
Imperial College Healthcare NHS Trust,
Hammersmith Hospital, London, UK current Certificate in Endocrinology and
e-mail: [email protected] Diabetes, awarded on a multiple choice

228 A. Dornhorst
examination of best of 5 of 200 questions and a medical education and training, each represent-
prerequisite for completion of higher specialist ing multiple members (Table 14.2). All represent
training in diabetes, includes only 10 questions the interests of their respective membership and
relevant to integrated care [4]. While the training some with conflicting interests around supporting
curriculum does cover delivery of diabetes care integrated care.
(Table 14.1), this knowledge can all be acquired The General Medical Council (GMC) is an
solely through reading not actual exposure to independent regulatory organisation responsible
integrated care, and currently represents a very for setting standards for the delivery of under-
small fraction of the total curriculum [5]. graduate and postgraduate training including the
Overall undergraduate and speciality post- final Certificates of Completion of Training
graduate medical training today has little expo- (CCT), the universal requirement for all GPs and
sure to the different community healthcare teams consultants, in any specialty, be it diabetes, respi-
working across a local population. This lack of ratory or other specialty, to obtain to work within
community exposure and hospital centric focus the UK [7].
has resulted, not surprisingly, in a negative per- To date the GMC has not been involved in the
ception among trainees and consultants on com- content of either the undergraduate or postgradu-
munity work, with hospital based consultant ate training programmes. The content of the
posts being considered more prestigious to com- undergraduate curriculum is the responsibility of
munity based posts. Going forward, undergradu- the 34 different UK undergraduate medical
ate and postgraduate medical training will need schools, all members of the Medical Schools
to change as will the perception around commu- Council. Although all medical schools do offer
nity work if integrated care as envisaged by the some community-based teaching in their curri-
5-year forward plan, is to flourish. The concept cula, the time spent and emphasis on cross-
that postgraduate medical training and workforce organisation and cross-discipline training,
planning need to be intimately interwoven with community placements and population health
health service policy and delivery is not new and varies among the medical schools [8]. As a gen-
was emphasised by Sir Professor John Took in eralisation, undergraduate teaching is given pre-
2008 in his report on Modernising Medical dominantly by university lecturers and hospital
Careers [6]. consultants, with little experience in chronic dis-
In the UK from the selection of medical stu- ease management in the community. A signifi-
dents through to higher speciality training and cant percentage of undergraduate clinical
the revalidation process of doctors to practise, placements are timetabled in the hospital envi-
there is no one unifying professional body ronment. There also remains a real financial dis-
responsible. There are eight main bodies involved
in the regulation, commissioning and delivery of
Table 14.2 Organisations responsible for the regulation,
commissioning and delivery of medical education and
post graduate training in the UK
Table 14.1 Aspects of endocrinology and diabetes cur-
riculum of Joint Royal Colleges of Physicians diabetes 1. The General Medical Council (GMC)
and endocrinology Training Board (JRCPTB) relating to 2. The Medical Schools Council (MSC)
the management of delivery of diabetes care 3. Education England (HEE)
1. The factors which influence commissioning 4. NHS Education Scotland (NES) l the Northern
diabetes care within the NHS Ireland Medical and Dental Training
2. Which aspects of diabetes care can be appropriately 5. Northern Ireland Medical and Dental Training
delivered in different clinical settings Agency (NIMDTA)
3. The role of information technology in integrating 6. The Conference of Postgraduate Medical Deans of
care across different providers the UK (COPMeD)
4. The role of diabetes networks and advisory groups 7. Wales Deanery
in the organisation of care 8. The Academy of Medical Royal Colleges (AoMRC)
14 Training for Diabetes Integrated Care: A Diabetes Specialist Physician Perspective from the English NHS 229
incentive for universities and their university 2014 [11]. This report recommended a greater
hospitals to actively encourage clinical place- amount of training during these first 2 years to be
ments in the community as this would result in undertaken in community-based settings, antici-
lost income. University hospitals are Local pating the need for the next generation of founda-
Education Providers (LEPs) and receive tariffs tion doctors to be better equipped to provide
for education and training for each undergraduate integrated care. A major recommendation from
placement from the Local Education and Training this report was that at least 80 % of foundation
Boards (LETBs), which in turn receive their doctors should undertake a community place-
money through HEE. These tariffs would be lost ment or an integrated placement starting in
if more clinical placements took place in the August 2015. A view echoed in the 2014 Shape
community. Currently a tariff for an annual clini- of Training report [12].
cal placement is in excess of 33,000 [9]. The responsibility of postgraduate speciality
For the profile of undergraduate teaching on training after the foundation years is dissolved to
chronic disease management in the community to the different medical royal colleges, faculties and
increase requires not only more out of hospital specialty associations to deliver the curricula and
based clinical placements, but for the assessment to assess trainees competencies. For example,
of population health to be part of the final quali- diabetes speciality training is the responsibility
fying exams. The Medical Schools Council of the Joint Royal Colleges of Physicians
Assessment Alliance (MSCAA) a partnership of Endocrine and Diabetes Training Board
its members has already agreed to include a pro- (JRCPTB) [13].
portion of finals examination questions from a The GMC involvement and influence on cur-
shared question bank. Potentially the MSCAA riculum content as well as training standards may
could play a role in supporting integrated care by increase following two major reports in the last 3
mandating its inclusion into all curricula and years. Firstly, the publication on the Shape of
final assessment exams [10]. Training Review by Professor Greenaways in
There is a significant focus from HEE and the 2013 [12], an independent review commissioned
individual LETBs to support medical workforce by the four UK governments sponsored in part by
planning and educational commissioning that the GMC, reported what changes were required
fosters doctors to meet the changing needs of and in medical postgraduate training to meet the
ageing population with complex health needs and future healthcare needs across the UK. The report
high expectations. There is a real appreciation by came up with 19 recommendations. Although
the LETBs that there will need to be fundamental these continue to be hotly debated, changes to
changes in postgraduate training to equip medi- medical training are likely to follow. This will
cal physicians for integrated care and recognition equip tomorrows medical specialist to be better
that current training is not doing this. suited to work in integrated care settings. This
The first 2 years of postgraduate medical edu- will include closer training with GPs and other
cation and training after qualifying, are under- healthcare professionals to deliver out of hospital
taken in a foundation programme that provide a speciality care at a population level in the com-
generic training to bridge the transition from munity. Another one of the 19 recommendations
medical school into specialist and general prac- of this report was for more subspecialty training
tice training. Both the GMC and the LETBs to be undertaken following qualification as a doc-
assess these educational programmes for the tor. The GMC is in favour of credentialing; how-
standards of training they provide. Placement ever, who exactly would pay for and accredit this
opportunities are in broad specialty areas with extension to training remains unclear. Potentially
opportunities to work in both primary and sec- training in integrated care for the long-term con-
ondary care settings. Integrated care was widely ditions both for general practitioners and hospital
referred to in the Health Education England specialists could become a recognised post CCT
Broadening the Foundation Programme report of credential [12].
230 A. Dornhorst
The second factor that might lead to the GMC ences. Placements that span the acute sector,
having a greater influence on the core medical community and private/voluntary sector organ-
curriculum is it support for a national licensing isations will require training programme direc-
examination to be taken by all graduates and doctors and educational supervisors, at a local level,
tors wishing to work in the UK, with 2021 being who meet the GMC standards for training. This
the provisional date for its implementation. Such may initially be difficult to establish in organisa-
an exam could support integrated care by includ- tions that are unfamiliar with training. Certain
ing greater focus on the nature of multidisci- specialities, such as respiratory medicine, have,
plinary team-work, the impact of differing UK through their speciality society the British
health systems and the interface between acute Thoracic Society (BTS), developed a curriculum
and primary healthcare and social care [14]. for integrated respiratory physician training [16,
Currently, following foundation training, 17]. Other speciality societies, including those
those trainees wishing to pursue a career in one involved with diabetes training, have yet to
of the 27 medical specialties enter a 2 year core develop a curriculum for integrated care. The
medical programme in which they rotate through BTS Working Group on Integrated Respiratory
generic medical disciplines before a competitive Care recognises it is essential in the future for all
selection process during year 3 of their post grad- specialist trainees to have some experience of
uate training (ST3) into one of the specialist primary care at least once during their training,
training programmes. These specialty training probably twice a year and late in the course of
programmes are usually an additional 45 years their training. The BTS acknowledges, in the
of training. Those wishing to pursue a career as a future, the roles of consultant involvement in
general practitioner, enter 3 years of GP Specialty respiratory care is likely to increase to include
Training (GPST) that normally includes 18 supervision of community sleep services, reviews
months in an approved training practice with a of those dying from airway disease and provision
further 18 months in approved hospital posts. of medical input into and care for those with idio-
The Royal College of General Practitioners pathic pulmonary fibrosis.
(RCGP) has been a long-term champion of inte- While exposure to community work during
grated care [15]. The Shape of Training report training is available in other medical specialties,
concentrates on medical as opposed to general including diabetes, this exposure is extremely
practice training. It proposes expanding the num- patchy and other specialty curricula could learn
ber of trainees working purely in general medi- from the proactive endorsement the respiratory
cine to 3 rather than 2 years before entering specialty society, BTS, has given to specialty
specialist training, and to continue with a com- training in integrated care. Diabetes UK, the
mitment to general medicine throughout their UKs leading diabetes charity, has widely sup-
specialty training. Current funding for these ported integrated care as the way forward in their
training years is paid half by the LETBs with the 2014 published report on the subject; however,
other half from the hospital trusts for clinical ser- this report failed to address the needs for profes-
vice. While there is a general acceptance that sional medical training to deliver such care, and
there needs to be a balance between training and the on going training consultants in diabetes
clinical service along with greater integrated would need to support out of hospital services
work, there is reluctance among the different spe- [18]. While this report had the endorsement of
ciality Royal College training boards and special- the Association of British Clinical Diabetologists
ity medical societies to shorten or dilute speciality (ABCD), the national organisation of Consultant
training. Physicians in Britain who specialise in Diabetes
For LETBs to commission a specialist training (the increasingly influential body) has remained
post that supports integrated care, training place- lukewarm concerning integrated care concentrat-
ments will need be outside of the acute setting ing more on the role of consultant diabetologists
and be able to provide suitable training experi- as specialty hospital based [19]. The ABCD has
14 Training for Diabetes Integrated Care: A Diabetes Specialist Physician Perspective from the English NHS 231
representation on the RCP training and Specialist Table 14.3 Supporting professional activities
Certificate Exam Board Specialist training 1. Continuing professional development
Committee boards and is in a very strong position 2. Local clinical governance activities
to influence change in diabetes postgraduate 3. Multidisciplinary training
training if it so wished. While the ABCD do 4. Formal teaching
acknowledge a role for consultants to provide 5. Management
specialist leadership for the local health economy 6. Appraisal
in designing a high quality and cost-effective 7. Job planning
integrated model of diabetes care, the training 8. Audit
required to do this has not yet been addressed. 9. Research
This contrasts with the emphasis on diabetes edu-
cational training for all healthcare professionals
in primary care and the need for clinical up skill- job plan for supporting the professional activities
ing the workforce that has come from general (SPAs) necessary to underpin direct clinical care
practice and the Diabetes UK Primary Care work. The SPAs for a community consultant post
Network [20] and Primary Care Diabetes Society will differ from those in a traditional hospital
[21]. For true integrated care in diabetes and the based consultant job plan but are equally as
other long term condition specialties to work, the important (Table 14.3).
education and training needs for both general At the very core of integrated care is working
practitioners and hospital specialists need to be across primary and secondary care. Advertised
more closely aligned. consultant appointments that are purely in the
The Royal Colleges, both the RCP and the community are not the way forward as this just
Royal College of General Practitioners, have a replaces one form of working in silos for another.
pivotal role in supporting integrated care, for all The RCP Advisory Appointments Committees
specialities. Not only through the RCP speciality should not sanction such posts and should
training boards but also through on going educa- strongly encourage the host institution holding
tional programmes and support for those consul- the contract to be an NHS trust, preferably a hos-
tants involved in community care. The RCP pital NHS trust. A major reason for this is that if
through its Advisory Appointments Committees the contract for the community specialist service
(AAC) is involved in scrutinising consultant job is not re-commissioned or is commissioned by a
plans and the appointment process of most con- non-NHS private provider, the consultant would
sultant posts in the England. They are also work- still hold an NHS contract with the hospital. This
ing with medical staffing departments to make is important as this has implications for pensions
them compliant with British Medical Association and other employment benefits. Again the RCP is
guidelines. The RCP continues to have a repre- in a strong position to actively encourage this
sentative on most medical consultant interviews. when approving consultant posts.
The RCP should take a lead from the respiratory The royal colleges are also well placed to raise
society, the BTS, which has published a generic the clinical profile and stature around integrated
job plan for community consultant. care by hosting faculty with expertise in inte-
For new diabetes and other speciality consul- grated care within their colleges and supporting
tant posts that have a community component, and research and educational conferences on inte-
an increasing number do, the RCP is well placed grated care. These programmes could include
to ensure that there is a balance between commu- programmes in leadership skills required for con-
nity and hospital work, that a newly appointed sultants in these roles. At the end of the day it
consultant is supported by the hospital specialist will be action not purely words that will dictate
team to guarantee that they have the skills neces- the success and implementation of high quality
sary, there is access to continual professional integrated care clinical services. Action around
development and appropriate time set aside in the education is key to this success.
232 A. Dornhorst
References 11. https://www.hee.nhs.uk/sites/default/files/documents/

Broadening%20the%20foundation%20report.pdf
12. http://www.shapeoftraining.co.uk/static/documents/
1. https://hee.nhs.uk
c o n t e n t / S h a p e _ o f _ t r a i n i n g _ F I NA L _ R e p o r t .
2. https://www.gov.uk/government/uploads/system/
pdf_53977887.pdf
uploads/attachment_data/file/480482/NHS_
13. http://www.jrcptb.org.uk/specialties/endocrinology-
Constitution_WEB.pdf
and-diabetes-mellitus
3. h t t p s : / / w w w. e n g l a n d . n h s . u k / w p - c o n t e n t /
14. http://www.gmc-uk.org/information_for_you/
uploads/2014/10/5yfv-web.pdf
25688.asp
4. https://www.mrcpuk.org/sites/default/files/documents/
15. http://www.rcgp.org.uk/policy/rcgp-policy-areas/~/
sce-endocrinology-and-diabetes-blueprint_0.pdf
media/Files/Policy/A-Z-policy/General_Practice_and-
5. http://www.jrcptb.org.uk/sites/default/files/2010%20
the_Integration_of_Care%20_An_RCGP_Report.ashx
Endo%20%26%20Diabetes%20%28amendment%20
16. https://www.brit-thoracic.org.uk/document-library/
2012%29.pdf
delivery-of-respiratory-care/integrated-care/
6. http://www.medschools.ac.uk/AboutUs/Projects/
role-of-the-respiratory-specialist-in-the-integrated-
Documents/Final%20MMC%20Inquiry%20Jan2008.
care-team-june-2014/
pdf
17. www.jrcptb.org.uk/sites/default/files/2010%20
7. UK. http://www.gmc-uk.org/about/role.asp
Respiratory%20Medicine.pdf
8. Lee SW, et al. The current provision of community-
18. h t t p s : / / w w w. d i a b e t e s . o r g . u k / D o c u m e n t s /
based teaching in UK medical schools: an online
About%20Us/What%20we%20say/Integrated%20
survey and systematic review. BMJ Open.
diabetes%20care%20(PDF,%20648KB).pdf
2014;4(12):e005696.
19. http://www.diabetologists-abcd.org.uk/home.htm
9. https://www.gov.uk/government/uploads/system/
20. h t t p s : / / w w w. d i a b e t e s . o rg . u k / P r o f e s s i o n a l s /
uploads/attachment_data/file/390006/2015- 16_
Professional-groups/Primary-Care-Network/
Tariff_guidance_-_DRAFT_-_web_version.pdf
21. http://www.pcdsociety.org
10. http://www.medschools.ac.uk/Pages/default.aspx
Diabetes Integrated Care: Are
We There Yet? 15
David Simmons, Helmut Wenzel,
and Janice C. Zgibor
Introduction Health Service: surely experiences of patients

well beyond the United Kingdom. The scale of
The preceding chapters describe a plethora of the epidemic described for Africa in Chap. 6 is
definitions of, and perspectives on, integrated particularly concerning: clearly we need afford-
care and diabetes integrated care, including a able, population based approaches for preventing
multitude of components that are thought to be diabetes, and to ensure that those who do develop
important in the running of the system. There diabetes have a better quality of life, are less
are common elements, but also aspects, which likely to develop complications and do not die
differ significantly. Some have been associated prematurely. Diabetes integrated care sounds as
with tangible clinical benefits, others less so. though it should be a major part of the solution,
Each chapter describes the growth in numbers but is it?
with diabetes and its complications, the persist- Our chapters show that integrated care for
ing under-performance in the clinical systems to people with diabetes has now been tested in a
optimise clinical assessment and metabolic con- range of settings and with a variety of approaches.
trol, maximise self-care (an accepted key feature So, can we now say yes, we are able to roll out
of quality diabetes care) and treat detectable tis- diabetes integrated care systematically? Are we,
sue damage early. The patient perspective is well as suggested in the recent Royal College of
described in Chap. 13, detailing the breadth and Physicians report [1], still dependent on local
depth of non-engagement and lack of articulation personal relationships between primary and sec-
between different parts of the English National ondary care rather than a framework which will
work in most areas? Or do we need to re-invent
the diabetes care wheel repeatedly, depending on
D. Simmons (*)
School of Medicine, Western Sydney University, whether the lead specialists, lead GPs and other
Sydney, New South Wales, Australia health professionals get on. We would suggest
e-mail: [email protected] not, but implementing diabetes integrated care is
H. Wenzel certainly neither simple nor easy. Finally, should
Health Economist, Konstanz, Germany we be looking at more creative and innovative
e-mail: [email protected] ways to improve diabetes care by more system-
J.C. Zgibor atically integrating the services and expertise of
Department of Epidemiology and Biostatistics, others including pharmacists and peer support
College of Public Health, University of South
Florida, Tampa, FL, USA facilitators, and for those under more medically-
e-mail: [email protected] centred systems, nurses and diabetes educators?

There are other initiatives globally not Table 15.1 The diabetes care project (DCP): Australias
national cluster randomised control trial of primary care
described in these chapters, which we also draw
coordination of diabetes care
upon to help define where we are, and perhaps
what we need to do to implement diabetes inte- The components of the DCP [2]:
1. An integrated information (IT) platform for general
grated are more widely. These include the
practitioners, allied health professionals and patients.
Australian Diabetes Care Project (DCP) (Table But not specialists
15.1) [2] and the work of Kaiser Permanente and 2. Continuous quality improvement (CQI) processes
other US providers, particularly since the intro- informed by data-driven feedback within primary care
duction of the Affordable Care Act [3]. There are 3. Flexible funding for primary care, allocated based
also many smaller front line approaches on patient risk stratification.
designed to integrate care around individuals with 4. Quality improvement support payments linked with
a range of patient population outcomes
additional needs, such as recurrent diabetic keto-
5. Funding for care facilitation, provided by dedicated
acidosis (DKA) with a combined diabetes spe- Care Facilitators
cialist service case manager-mental health Study cost: A$33.4 million over the period 20112014
approach [4] and among rural indigenous people Intervention groups
[2]. The former included a diabetes specialist 1. Group 1 IT platform and CQI processes (i.e., no
dietitian with approval to manage continuous sub- funding changes)
cutaneous insulin infusion (pump) therapy and 2. Group 2 all five components
carrying out some aspects of social work, while 3. Control group
the latter including a dedicated Aboriginal Health Study population
Worker. These provide examples of how diabetes 184 general practices and 7781 people with diabetes
integrated care can benefit from broader skills and enrolled over 6 months
broader membership. One thing that is clear is 18 months intervention
that there are some fundamentally different per- Included type 1 and type 2 diabetes
spectives on diabetes integrated care and what it Outcomes
involves. We therefore commence this chapter 1. Group 2: HbA1c dropped 0.2 %, systolic blood
pressure (1 mmHg), LDL cholesterol (0.06 mmol/l),
with a fundamental dichotomy is diabetes inte- waist circumference (0.4 cm) care-plan take-up,
grated care an entity to be developed in itself (i.e., completion of recommended annual cycles of care,
a front line diabetes approach) or should it be sim- and allied health visits (educator, podiatrist, dietitian)
ply a product of a system carefully crafted by the improved
2. Group 1: increased care plan take-up
health system masters (i.e., waiting for the wider
3. No relationship between health care need and the
system to become integrated)?
amount of chronic disease funding received
We then describe the components of diabetes 4. Having a care plan or completing an annual cycle of
integrated care systems reported to date, and com- care did not influence metabolic control
pare how and whether different diabetes integrated 5. Significant increase in cost per person, no
care approaches dealt with them. The validity of significant decrease in costs of hospitalisation
the methods for evaluation is crucial of course, so Comment
we make a few comments on how the different No significant clinical improvements
projects have been assessed. We finalise by trying More expensive
to build up to suggested foundations for function- Hypoglycaemia not considered in the risk
stratification
ing and sustainable diabetes integrated care.
Specialists excluded from the payment system-only
GP and allied health Medicare item numbers
included
Top Down or Front Line
Approaches?
Act in the USA [3] and the Quality Outcomes
The improvements in care (at least for some Framework [5] in the UK reflect the reality that
patients, for a period of time) that are possible diabetes care (including self-care) does not sit
with policy changes, such as the Affordable Care alone, but is part of total health and social care
15 Diabetes Integrated Care: Are We There Yet? 235
systems, and indeed, overall society. This gener- This year vs the future: The majority of
ates an immediate dichotomisation of perspec- adverse diabetes outcomes (e.g., stroke, myo-
tives that goes to the heart of the diabetes cardial infarction, blindness, amputation, end
integrated care debate: stage renal failure) are many years in the
future, so a system based on this years budget
Is the way to improve diabetes care through a will often not prioritise the future (with harms
top down approach, i.e., change the broader predicted by, e.g., HbA1c, blood pressure).
health system and diabetes care will inevitably Cost attribution: Even for these adverse dia-
be optimised as a result OR betes outcomes, other clinical factors con-
Is the way to improve diabetes care through a tribute to the end result. This is even more the
front line approach that defines each com- case for, e.g., infections, falls and other condi-
ponent of care, related barriers to component tions which are more common or more expen-
optimisation and finds holistic ways to address sive with diabetes. As a result of this
them complexity, attribution to diabetes may not be
so obvious to those seeking ways to minimise
The former reflects a more managerial policy, expenditure. It has previously been shown in
public health perspective, often encapsulated by England that the population attributable frac-
the think tanks (e.g., Kings Fund [6]) and tion of the excess hospitalisation costs from
accounting firms (McKinsey [7, 8]), while the diabetes is approximately 40 % [10], compris-
latter reflects a more grounded, clinical perspec- ing excess admission and readmission rates,
tive. What we can see from the chapters in this length of stay, and cost of stay.
book is that both perspectives are needed: an Cost impact of drugs: Payment for drugs can
integrated approach! We know clearly that there range from personal, insurance and govern-
are significantly different perspectives between ment subsidised budgets. Evidence for reduc-
patient, primary care and secondary care [9]. The ing future health costs is complex and time
paradox that we are dealing with is that in spite consuming to collect and the pharmaceutical
of health professionals wanting the best for their companies themselves decide the final price of
patients on a patient by patient basis, the way the drugs. In England, drugs may be shown to
that individuals and institutions are organised improve quality of life at an acceptable cost
and paid, directly influences the clinical deci- when benchmarked against other drugs
sions that are made. In South Africa (Chap. 6), it (through NICE), but might still not be taken
was reported that many. programs concen- up because of the cost impact on local bud-
trate on cost-savings rather than patient service gets: budgets held by the same primary care
utilization and improved clinical outcomes. practitioners who are managing the people
Furthermore, the ease with which the best with diabetes.
decision can be made, directly determines what Separation of ambulatory and inpatient bud-
action is taken: the more time/effort required, the gets: Payment for ambulatory diabetes care,
less likely that it will take place. A simple skel- which is essentially the preventative part of
etal representation of the paradox is as follows diabetes care, usually sits in a different budget
(Fig. 15.1). to the inpatient budget where the big expenses
Naturally, optimising personal care and the are. Increasing real funding for ambulatory
provider/purchaser-commissioner budget may be diabetes care for possible reductions in hospi-
aligned, but this is where diabetes poses substan- talisation in the future is hard to sell as the
tial problems from a health system point of view: perspective that if you have a bed you will fill
Incentives that optimise patient

personal care and this years budget
Short term and Best decisions

not the best this year, not
decisions the future
Barriers to the best Facilitators of the best

decision decision
Long term but Best decisions

not the best for the future
decisions
Incentives that optimise patient

personal care but not necessarily
this years budget
Fig. 15.1 How optimisation of care and annual budgeting can compete in diabetes
it is common among commissioners. Qualitative impacts: There is a mantra that

Furthermore, good evidence for reducing hos- fragmentation of care and reductions in conti-
pitalisation through diabetes integrated care is nuity of care are likely to harm the quality of
limited, although the South African and Hong care [14], but hard evidence is difficult to
Kong models do show that it is possible: of obtain.
course this will depend upon context and start-
ing point. So, the payment, funding or commissioning
Obstruction to referral to a more specialist system can clearly be a major determinant of the
service: There is ample evidence [11, 12] quality of care around the person with diabetes.
where clinicians own, and profit from, other The problems outlined above, suggest that any
services (e.g., laboratory, radiology), that health system that fails to take account of the
referral rates are increased, often inappropri- need to integrate the payment system from both
ately (although showing this on a case by case an immediate and long term perspective, must be
basis can be difficult). Under the English at greater risk of their diabetes integration
NHS, the converse exists, where GPs, either attempts failing and/or being unsustainable. The
holding health budgets, or receiving payments Derby example (Chap. 8) shows that even when
for maintaining health budgets [13], reduce the immediate payment system is merged at the
their referrals to more specialist care. While provider level, and benefits have accrued, sus-
this may be appropriate in many cases, it may tainability is not assured unless vertical integra-
result in delays and avoidance of referrals, tion has occurred. Similarly, the East Cambs and
even when specialist care is likely to be of Fenland approach (Chap. 7) tried to address each
benefit. care component, but the insufficient movement in
vertical integration was the likely cause of the economic health, and the latter within the health
lack of effect or sustainability. The bundled system.
care and chains of care in the Netherlands As per Fig. 15.1, both patients and HCPs
(Chap. 11) and Sweden (Chap. 12) have tried to have to work with facilitators and obstructions to
link providers in different ways, but neither optimal diabetes care and self-care. Obstructive
seems to have reduced cost. components sit within the barriers to diabetes
Essentially, the front line approach seems to care framework described in Chap. 7 (East
be required to ensure that the strategy is grounded Cambs and Fenland initiative). Many of the facil-
in clinical best practice, but the overarching itators of the best patient decisions also facilitate
framework of the health economy needs to sup- best clinical practice. However, the health care
port an integrated approach within, and poten- professional sits upon a range of systems facilita-
tially, beyond diabetes. As the changes to tors (or otherwise) within a context that can pro-
governance and commissioning are substantial, it mote a population based approach to health
may be more acceptable to change the whole sys- (goals of population based outcomes, finance,
tem on the other hand, the details are such that shared clinical governance) or otherwise (organ-
an integrated care pilot within diabetes should be isational finance, separate clinical governance).
attractive if those involved appreciate the com- Table 15.2 compares the context of each of the
plexity of what is required to create an integrated approaches described in this book, and adding the
system. Australian DCP, a trial of integrated care compo-
nents in a fragmented health system [2]. Two suc-
cessful front line (micro) approaches (managing
Components of Diabetes UK patients with repeat diabetic ketoacidosis [21]
Integrated Care and managing patients within an Australian
Aboriginal Health Service [22] are also included.
Although the patient is often appropriately placed Table 15.3 compares the components within
in the centre of policy diagrams to reflect patient each of the approaches. There are clearly a num-
centred care, in fact, from a clinical perspective, ber of common factors and several that differ
outcomes depend upon the patient-health care between successful and less successful models.
professional (HCP) interaction. This relationship
is strengthened through consideration of factors
outlined in the American Diabetes Association/ Data and Its Interpretation
European Association for the Study of Diabetes
position statement which include risks or side Success in these models is usually described in
effects associated with treatment, disease dura- terms of hospitalisation (including, e.g., DKA,
tion, life expectancy, important comorbidities, amputation, cardiovascular disease events, hypo-
vascular complications, patient attitude and glycaemia, eye disease, renal disease, all cause),
expected treatment efforts, and resources and metabolic outcomes (e.g., HbA1c), health costs
support system [15]. At the Veterans (Chap. 2), and access to complex care. Some have described
this is called the patient-doctor dyad. The impor- patient related outcomes, quality of life and other
tance of this relationship is emphasised in the staff satisfaction, but the methodology and biases
chronic care model [1620]. have often not been open to scrutiny. There are
Figure 15.2 summarises pictorially, the com- some methodological issues that suggest that
ponents of diabetes integrated care included in many of those with positive results may be illu-
the chapters in this book. The figure places the sory and reflect the pre-existing landscape and/or
patient-HCP dyad in the centre, but, of course, wider changes, particular to that locality. It is
both patient and HCP sit within their own con- important to note that these do not in any way
text: the former in relation to their family, mental denigrate the models described, simply that ben-
(including motivational and spiritual) and socio- efits may not be as large as described.
Fig. 15.2 Components of diabetes integrated care placed within the patient-health care professional dyad and the bar-
riers to diabetes care framework: the four slice sandwich
Secular Trends Relativity and Starting Points
The reported success of intermediate diabetes Kaiser Permanente is often described as a great
clinics run by English General Practitioners with diabetes success story in the USA due to its
a Special Interest led to extension of the model to higher than peer levels of, e.g., HbA1c testing
other areas. This was finally tested in a ran- [23]. However, in the 2015 HEDIS data, levels of
domised controlled trial as described in Chap. 9 testing, metabolic control achieved and compli-
(Warwickshire) and shown to be a more costly cation rates show quality metrics lower than the
model with no real benefit for patients or the sys- English NHS, in spite of the problems with the
tem. Similarly in East Cambs and Fenland, the 1 latter [23]. Furthermore, HbA1c rates above 9 %
year results suggested major reductions in hospi- remain at approximately 20 %, in Southern
talisation and costs in practices participating California [24] or 19 % in Northern California
fully in the integrated care initiative, compared [25], a level much higher than that in the UK [7],
with those who engaged later [9]. However, and nearly abolished in an Aboriginal Medical
once the trends in neighbouring areas and among Service [22]. Similarly, the Super Six model, an
those without diabetes were accounted for, it approach with commendable primary care based
became clear that the benefits originally reported interventions, has been lauded as a success, as a
were actually due to wider hospitalisation reduc- result of reductions in patients with, e.g., amputa-
tions, not just in those with diabetes. Studies of tions. However, these complications were in the
hospitalisation/hospital costs that do not compare bottom quartile of performance for these out-
with rates in the non-diabetic population need to comes in England [26] and hence improvement
be interpreted with caution. The Affordable Care would be expected with the additional diabetes
Act in the USA provides a natural experiment in resources invested into the area. Amputation rates
which the role of insurance coverage, provider remain higher than the national average, in spite
incentives and outcomes can be evaluated, of a less heterogenous population than in many
although it may be too early to tell. other parts of England [27]. Studies showing
Table 15.2 Comparison of different contexts within which the models of diabetes integrated care sit
US-Vet SF UPMC HK SA ECF Derby Cov-W Ger NL Swe DKA-P A-DCP Rumba
CF National 8508 8508 8508 3405 3405 3405 4495 5099 3925 3405 3800 3800
$ /capita 2011 [33]
Coordinated [33] 6/11 6/11 6/11 1/11 1/11 1/11 10/11 5/11 11/11 1/11 4/11 4/11
care CF-Rank
Outcomes
Hospitalisation ?b ?b a a 40 %a c a ?b a ?b b a a ?b
Glycaemia+ (e.g., 0.6 %a a a a 1.5 %a b a RCT IT ?b c b a 0.2 %b a
HbA1c) 0.8 %b
Costs Productivitya b ?b a a ?b a c a c b a c ?b
Access to complex a ?b b a a c a 1st? IT yesb ?b c easyb a b a
spec care
HCP context
Funder Federal Govt Local Insured Govt Insured Govt Govt Trial/Pilot Insured Insured Govt Govt Govt Govt
Govt Govt
Defined Veterans Low Members Capitated Members Capitated Capitated Capitated Members National Regional >1 DKA Capitated Indigenous
population income admit
Levels covered 2 1-I 1-I 2 :90 % 1-I-2 1-I-2 1-I-2 I 1-I-2 1-I-2 1-I-2 2 1-I 1-I-2
of all DM 70 %DM
Organisations Hospital GP GP+ Hospital Specialist CS JV Specialist JV GP Local Hospital GP AHS+
leading hospital Govt hospital
Number of 1 1 1 1 Multiple Multiple 1 1 Multiple Multiple 1 1 Multiple 2
employers
Financial Single Single Single Single Single Multiple Shared Single Shared Multiple Single Single GP AHS+
governance saving hospital
Service Single Single Single Single Within Multiple Single Single Multiple Multiple Single Single GP GP+
management provider specialist
Clinical Single Single Single Single Within Within Shared Shared Shared Within Within Within Within Within
governance provider provider provider provider provider provider provider
Information Singleb Singleb Singleb Singleb Some Within Shareda Singleb Shareda Within National/ Within Within Within
governance sharedb providerc providerc internalb providerc providerc providerc
$ Incentives No Yes Yes No No QoF QoF QoF Shared Profit No No Yes No
saving multiple
Key: Levels = 1 Primary, 2 Secondary, I intermediate, CF Source: Commonwealth Fund [8] QoF Quality Outcomes Framework, IT IT/Education based pilot project, JV Joint Venture
a
Good/improved between primary and secondary care
b
maybe-unclear
c
worse/not included
Table 15.3 Comparison of the content of different models of diabetes integrated care
DKA-
US-Vet SF UPMC HK SA ECF Derby Cov-W Ger NL Swe P A-DCP Rumba
Patient factors
Continuity of care Yesa Yesa yesb Yesa Yesa Yesa Yesa Nob Yesa Yesa Recently Yesa Yesa Yesa
reducedc
Team working Yesa Yesa yesc Yesa Yesa Partlyb Yesa Partlyb Yesa Yesa Yesa Yesa Yes not specialistb Yesa
Overcoming Personalised carea Team yesc Yesa Personalised Yesa Yesa Noc Someb Noc Some Yesa Noc Yesa
patient barriers carea carea socialb
Co-morbidity Yesa Yesa Noc No-GP No-GP onlyc 1-yesb Partlyb No-GP Partlyb Noc Noc Yesa Noc Yesa
Mx articulation onlyc 2-nob onlyc
Pt Co-payment Someb Yesc Yesc Yesc Noa Noa Noa Noa Noa Yesc Yesc Noa Partlyb Noa
Patient education Yesa Yesa Yesa Yesa Yesa Yesa Yesa Yesa Yesa Yesa Yesa Yesa Yesa Yesa
Case management Dyada Yesa yesc Noc Noc Noc Noc Noc Noc Noc Noc Yesa Care facilitatorsb Noa
Peer support Nob Yesa Nob Yesa Nob Nob Nob Nob Nob Nob Nob Nob Nob Nob
Patient e-records Noc Yesa yesc No-but Noc Noc Noc Noc Noc Noc Noc Noc Yesa Noc
reportsb
HCP factors
Electronic health Yesa Yesa yesc Yesa Yesa Noc Yesa Yesa Yesa Noc National Noc Yesa Noc
records datab
Electronic Yesa Yesa yesc Yesa Noc Noc Yesa Noc ?b Noc Noc Noc Noc Noc
communication
Referral p ways Yesa Someb Someb Yesa Yesa Yes partlyb Yesa Yesa Yesa Maybeb T1-2a Yesa Yesa Yesa
risk stratification T2 1a
Guidelines Yesa Yesa Yesa Yesa Yesa Yesa Yesa Yesa Yesa Yesa Yesa Yesa Yesa Yesa
HCP Decision Noc Noc yesb Yesb Regular auditb Joint Yesa No 1stb Noc Noc Yesa Yesa Yesa Yesa
support reviewsa Yes 2ndb
HCP education Noc Noc Yesa Yesa Yesa Yesa Yesa Yesa Noc Noc Noc Yesa Yesa Yesa
Diabetes
Prevention Noc Noc Noc Yesa Noc Noc Noc Noc Yesa Noc Noc Noc Noc Noc
Care facilitators coordinated but did not deliver care
Key: Levels = 1 Primary, 2 Secondary, I intermediate, CF Source: Commonwealth Fund [33], QoF Quality Outcomes Framework, IT IT/Education based pilot project, JV Joint Venture
a
Good/improved between primary and secondary care
b
maybe-unclear
c
worse/not included
improvement from a low baseline do not neces- behaviours. It is not credible to expect these skills
sarily provide a best practice model, but perhaps across primary care, and most patients with type
a change from a system that required improve- 1 diabetes continue to have substantial glycaemic
ment. Was it the greater resource, a change in variation even with a good HbA1c and are at
leadership or the new model or a mixture of all: high risk of premature death [29]. However,
without well conducted randomised controlled given differential access to specialty care, sup-
trials (which are probably impossible), we will port for primary care through technology or other
not know. mechanism may provide a solution.
HbA1c Changes What Facilitates Diabetes

Integrated Care
Several projects report improvements in HbA1c,
however, to show improvements in mean or pro- Figure 15.2 shows the large number of enablers
portion with better control in an area with worse and components for integrated care; taken from
than average glucose control does not necessarily the models described, these include the need for
denote improvements due to integrated care. integrated information technology systems,
More importantly, improvements in HbA1c, aligned finances and responsibility, care planning
without reports of hypoglycaemia rates and and robust clinical governance. However, the key
weight gain, may be associated with worse out- to integrated care remains the relationships
comes as suggested from the ACCORD trial [28]. between patients and the different health care
However, moving the population to better control provider. The elements of this are discussed next.
by following a patient-centred approach should
minimize the likelihood of adverse outcomes.
Activated and Informed Patients
Future Changes Most of the models attempt to activate and inform

patients about their diabetes. This may be through
Surely, what is intended with any new approach access to their own electronic records as in
is to create a sustainable approach to improved Warwickshire or under the House of Care
population based diabetes management, not just approach described in Chap. 13, where patients
within primary care and within secondary care, had access to their results. The role of social
but with all aspects working well together and media and other electronic approaches to infor-
properly articulated. It is here that the Super Six mation and participation are yet to be fully
model is of particular concern. Notwithstanding explored and remain an exciting area for devel-
its exemplary close collaboration between GP opment. Conversely, peer support, so effectively
and specialist services, a model that prevents used in both San Francisco and Hong Kong,
access to the specialist team by patients with type appears to be an important mechanism to improve
1 diabetes (besides those with poor glucose con- self-management and avoid hospitalisation in
trol, youth or continuous subcutaneous insulin some patients.
infusion therapy), those with rare forms of diabe-
tes and those with difficult to control type 2 dia-
betes is unlikely to produce long term benefit to Activated and Informed Health Care
patients and create sustainably low hospitalisa- Professionals
tion rates. Type 1 diabetes management in par-
ticular, requires experience in assessing, e.g., Rao (Chap. 2) raises an important point that
hypoglycaemia, hypoglycaemia unawareness, where a single clinician takes responsibility, the
alternative diagnoses, and wider life impacts and resulting personalised care might be a key factor
in achieving metabolic outcomes, rather than health sectors. A culture of collaboration and a
simply prospective measurement of metabolic population based outcomes approach is unlikely
status. The House of Care reported improved to be natural in a market or competitive system,
knowledge and skills, and greater job satisfaction and part of the leadership training that is required
among HCPs [30]. Prior research by Pringle et al for those involved in governance/management
showed that the patients of GPs with an interest and leadership should be built upon this philoso-
in diabetes had significantly better glycaemic phy, when patients with diabetes are involved [1].
control [31]. Similarly, the Royal College of Physicians [1]
have emphasised the need for a shared vision
across organisations and professions, built around
Team/Collaborative Care the users perspective and supported by an ongo-
ing process for co-design, continued stakeholder
Being within the same organisation does not engagement and improvement. The bottom line
mean that integrated working is inevitable, but is that diabetes integrated care in a population
looking at Table 15.2, would certainly appear to will not occur by chance currently, but will ini-
be an important facilitator. Co-location and con- tially need investment in local leadership training
current team working has been suggested to be a and engagement frameworks. This must recog-
key factor in the achievement of the Derby, nise the time constraints among many of those
Veterans and Pittsburgh UPMC outcomes. involved.
Mapping and planning the joint work, with clear
definition of outcomes, process and roles, was
also crucial to improving the efficiency of the Evaluation and Feedback
care provided. The Veterans chapter provides a
step by step guide for those wishing to improve The Warwickshire approach (Chap. 9), included
their own approach to care. One wonders to what Diabetes Manager, a tool not only providing real
extent the fact that all HCPs are regional employ- time feedback to improve decision-making, but
ees, i.e., have the same employer, contributes to allowing benchmarking with peers. Decision
Sweden being the country with the best diabetes support was associated with a 37 % improvement
care delivery in Europe (936 points out of a maxi- in HbA1c in meta-analyses [34], Such an
mum possible of 1000: the highest score ever approach is not new in diabetes, and was part of
observed in the Euro Diabetes Index 2014 [32], the Diabcare movement over 20 years ago [35]
in spite of relatively limited integration (accord- and is a regular part of, e.g., the UPMC quality
ing to the Commonwealth Fund [33]). improvement efforts.
Where HCPs are not within the same organ-
isation (and even if they are), it is clear that train-
ing for integrated working should be implemented What Obstructs Diabetes
[34]. Integrated Care
The importance of the perverse incentives that

Clinical Engagement and Leadership can be embedded in the reimbursement system is
well described in Chap. 2 (veterans), Chap. 3
Different health systems and different aspects of (Pittsburgh) and Chap. 7 (East Cambs and
heath have different leadership models. Diabetes Fenland). In the former, the system obstructed a
UK have emphasised that whoever is leading, contemporaneous and synchronised multidis-
time and resource need to go into engaging clini- ciplinary approach and promotes increased vol-
cians to create a collaborative and constructive umes of activity (but neither quality nor
culture between those with diabetes and health outcomes) activity. In the latter, the reimburse-
care professionals and between the different ment system systematically blocked joint
working and promoted box ticking rather than Models of Clinical and Corporate
effective care. Under both, there remained practi- Governance
tioners dedicated to best practice, but this was
in spite of the reimbursement system. The Royal If leadership, financial and governance structures
College of Physicians [1] also describe short- are the upstream factors that obstruct diabetes
term service contracts, funder-provider split, integrated care (even before we get to the hard-
competing organisational budgets, activity-based ware, software, patient and health care profes-
tariff, inadequate resources, continual organisa- sional factors), what are the governance models
tional change, inadequate training, poor manage- that are available? The Veterans, UPMC, and
ment support from acute trusts and the lack of an Kaiser are single organisations, so all compo-
evidence base for ensuring sustainable, effective nents of their systems come under a single gover-
services as barriers to integrated care. The nance and management framework. All
Center for Medicare and Medicaid Services responsibility lies with a single entity. However,
(CMS) has instituted meaningful use criteria to in most cases, multiple organisations (e.g., GPs,
improve quality, safety efficiency and to reduce hospitals, community services, private providers)
health disparities [36]. CMS provides payment may be involved. In Cambridgeshire, the Kings
incentives as certain meaningful use parameters Fund [37] was commissioned to describe the
are adopted. While CMS funded 12 sites to eval- advantages and disadvantages of the various gov-
uate the impact of meaningful use, these results ernance models available in a setting where a
are not yet available for diabetes measures. single entity was not possible (Table 15.4).
Table 15.4 Governance models

Model Advantages Disadvantages
Virtual Inexpensive Little potential for progression to focus on
network system-wide decision-making through informal
lines of communication
Uncomplicated Dependent upon individual relationships
Potential for cost savings Unclear accountability
Hub and Uncomplicated and intuitive model of delivery Limits communication flows
spoke Allows easy entry of new provider Centralisation of expertise may create professional
model organisations divide
Clear lines of accountability Lack of joint ownership might exacerbate
traditional divides
Managed Transparent decision-making processes Potential inconsistency with espoused government
clinical policy (i.e., reduction of quangos)
network Dedicated management function Cost and resource implications
Clear lines of accountability Possible additional bureaucracy
Equal Local stakeholders are receptive Possibility of bureaucratic approach to decision-
partner making (slowing down progress)
network Potential to allow more systematic approach Cost and resource implications
to patient safety and clinical effectiveness
Builds on good relationships
Legally Establishment of a single, clear framework for Cost implications
binding clinical governance
joint Could facilitate single vision and culture (not Could be bureaucratic
venture guaranteed) Possible conflict with emergent government policy
Wouldnt necessarily lead to integrated care
Could damage existing good relationships
Used with permission from Kings Fund. Consultancy Report: Partnership for Sustainable Health Care Development:
Options paper for approaches to governance and clinical governance for an integrated care approach to diabetes. Final
Report. London: October, 2010 [37]
Whereas the Derby model was a legally binding education for people living with diabetes, their
joint venture, most of the other models used in families and healthcare professionals in all the
England have been virtual networks. countries surveyed. Furthermore, due to depen-
dency on subsidized or free diabetes medicines
and devices, many people either have to pay for
Is There a Tension Between Access their own treatment or just do without. People
to Quality Diabetes Care and Paying with diabetes in Spain reported spending on aver-
for Diabetes Care? age 428 USD per year in 2013 [39] (300 euros) on
their diabetes medicines and devices. This is, in
The cost of medications and devices has been turn, about 1.9 % of the median disposable
known to be a significant barrier to self-care for monthly income in 2011 (1.265 euros per month
many years [38]. However, costs remain a major [40]). In 2013, people with diabetes in Poland
barrier to care where out of pocket expenses reportedly spent around 560 USD per year (about
remain significant. The International Diabetes 400 euros) for their treatment; this is 3.6 % of the
Federation Europe analysed access to quality dia- Household Net Adjusted Disposable Income [39].
betes care in 47 countries of the European Region In Bulgaria, Russia and Azerbaijan, people
[39] based upon availability, access and afford- reported having to spend well over 821 euros a
ability. Whereas availability reports the presence year up to 1200 euros [39].
of products in a country that meet the needs,
accessibility describes physical access to prod-
ucts, including access to prescribers and educa- What Would Ideal Diabetes
tion. Affordability depicts the way of paying for Integrated Care Look Like? One or
products and care, especially reimbursement of Many Variants?
the health care system and out-of pocket pay-
ments [39]. Naturally the situation varied from We would suggest that while a diabetes inte-
region to region. Nevertheless, the survey identi- grated care approach may vary in its implementa-
fied several problems that are directly or indi- tion depending on local relationships and
rectly linked to the financing of care. leadership, the components are very clear. We list
Budgetary constraints and austerity measures these in Table 15.5 with some of the ways they
(Portugal, Spain, Greece) are especially a prob- can be implemented.
lem in southern European countries. Half of the
countries that were included in the survey reported
scarcity and supply problems (availability). More Sustainability and Replication
than one third of the respondents of those coun-
tries also reported increasing difficulties in getting Although the Pittsburgh Veterans integrated team
their prescription. They encountered delays of up model improved clinical outcomes and productivity,
to several months to see their healthcare profes- it has reportedly not been replicated elsewhere.
sional, or because there are no healthcare profes-
sionals close to where they live (accessability).
The economic and financial crisis aggravates these Wider Benefits
issues. This is predominantly true for Mediterranean
countries, such as Greece and Portugal, where Besides the total health system learning and tools
local healthcare centres have had to reduce their that can arise from implementing diabetes
staff or have simply closed down due to austerity integrated care, the Hong Kong approach
measures [39]. Overall, the study also identified a (Chap. 5), demonstrates the phenomenal research
general lack of access to continuous diabetes benefits that can accrue. As a result of the diabetes
Table 15.5 Blueprint for an integrated diabetes care service

Empowered/enabled/activated and informed Education, access to electronic records, HCP telephone support,
patients nominated diabetes care promoter or care lead/care manager,
peer/health coach support/case management
All skills represented in the organisation, all All members of the team work at the top of their competency (not
HCPs competent at their level just license/certification as much of the world has often moved on
from this training)-Ongoing HCP education
Clarity over roles and limitations and easy Mapping and planning the joint work, with clear definition of
access to others for advice outcomes, process and roles, was also crucial to improving the
efficiency of the care provided. Chapter 2 provides a step by step
guide for those wishing to improve their own approach to care.
Similar re-engineering of work flow occurred in Hong Kong
Primary care support by specialists In the Super six model and in the East Cambs and Fenland
model, payments were made for specialists not only to see
patients but to support primary care including:
Virtual clinics (case-based discussions)
Database reviews to discuss individual patients with regards to
targets, hospitalisation, referrals
Reviews of audits completed by the GP practice on diabetes
care
Educational sessions on areas of diabetes management of the
practices choice
Patient reviews (in conjunction with GPs or practice nurses)
All information and communication shared Electronic communication, electronic records
readily-Information management optimised
Methods to overcome clinical inertia Benchmarking, audit, decision support, QA, use of different
health disciplines
Aligned finance-no disincentives Link hospital costs with ambulatory costs-single budget for pts
with diabetes. As for incentives-jury is still out
Single organisational governance/management Single organisation, joint venture, Hub and Spoke model,
structure even if made up of more than one probably not network models
organisation
Endocrinologist/diabetes specialist leadership The successful models integrating primary and secondary care
or joint venture model were either led by specialists or were joint venture models
Leadership clearly defined and supported ..and trained
Nuts and bolts in place Define risk categories, targets and wider guidelines, e.g.,
education, drugs, self-monitoring, support, diet, lifestyle and, e.g.,
palliative care, mental health, complications management
Define minimum visit/care expectations As per the Veterans in the USA and Paediatrics/Transition under
Best practice tariff in the UK
Define pathways T1DM = is specialist in nearly all models
T2DM = largely primary care but including a range of disciplines
Other specialist = pre-pregnancy, antenatal, postpartum, inpatient,
emergency, renal, foot, ?eye, cardiovascular, rare forms of
diabetes
Morbid obesity and type 2 diabetes: medical or surgical?
Importance of using the data for not only Electronic database-complementary epidemiological, social
improving care but advancing knowledge science and health economic (and health informatics) expertise
through research (Hong Kong)
integrated care system and databases that have Conclusion

been developed, the Hong Kong team have been
able to generate important new knowledge relat- The creation of a local health system that can
ing to the epidemiology of diabetes, the genetics integrate primary, secondary and community dia-
of diabetes and testing different clinical assess- betes care, sharing the work while getting the
ments and interventions. Their approach has also best from each, would seem to be an obvious and
generated biobanks and patient pipelines into relatively simple and sensible way forward. We
clinical trials. Their JADE programme has also have shown in this book that while this can work,
been able to support other health systems. A fur- it is neither simple nor straightforward.
ther benefit of the integrated care approach has Integration will require its own resource. Health
been to assist with workforce planning, another systems are too complex for integration to hap-
key strategic component of maintaining a quality pen just because it is good/best practice.
health system. We provide individual components that appear
to be required for successful integrated care.
Some, such as integrated IT systems, may need to
The Two Cinderellas of Diabetes come later; other components, such as achieving
Integrated Care a registered population and aligned financial
incentives may take some time and substantial
There remain two Cinderellas of diabetes facilitation, but this should not stop a stepwise
care that few of the diabetes focussed models approach across each local health economy.
of care appear to address: integration with Governments should move to systems that will
mental health services and with diabetes facilitate integration, and away from the market
prevention. systems that appear to increase cost and reduce
The DKA prevention case management model the quality of diabetes care.
[41] included a substantial and operational inte-
gration between the diabetes specialist service
and the mental health services. While many pri- References
mary and secondary care services have staff
trained in, e.g., motivational interviewing, and 1. Hussain S, Dornhorst A. Integrated care taking spe-
cialist medical care beyond the hospital walls.
can offer mental health first aid, few appear to be
London: Royal College of Physicians; 2015.
integrated seamlessly with mental health services. 2. Australian Department of Health. Evaluation report of
While this could be said for many co-morbidities, the diabetes care project. Canberra, Australia; 2015.
there are good examples of other wider integrated 3. Office of the Legislative Counsel. Compilation of
patient protection and affordable care act 2010.
team working (e.g., multidisciplinary foot teams,
Available from: http://housedocs.house.gov/energy-
obesity teams). commerce/ppacacon.pdf
Prevention of type 2 diabetes is the other 4. Randall L, Begovic J, Hudson M, Smiley D, Peng L,
important issue that appears to continue to remain Pitre N, et al. Recurrent diabetic ketoacidosis in inner-
city minority patients: behavioral, socioeconomic,
generally outside the remit of diabetes care. Hong
and psychosocial factors. Diabetes Care.
Kong does have its nurse-led OPAL programme, 2011;34(9):18916.
the evaluation of which is keenly awaited. Many 5. Health & Social Care Information Centre. Quality and
would see diabetes prevention as in the domain of Outcomes Framework (QOF) (2014) [05/15/2016].
Available from: http://www.hscic.gov.uk/
public health and primary care and primary care
media/15053/QOF-Publication-Consultation-
is, of course, a key component of diabetes inte- Results-2014/pdf/QOF_Publication_Consultation_
grated care. However, there are growing aspects Results_2014.pdf
where those with diabetes may be able to facili- 6. Alderwick H, Ham C, Buck D. Population health sys-
tems: going beyond integrated care 2015 [05/15/2016].
tate prevention [42] and where those needing pre-
Available from: http://www.kingsfund.org.uk/sites/
vention are not being followed up adequately in files/kf/field/field_publication_file/population-health-
primary care. systems-kingsfund-feb15.pdf
7. De Smet A, McGurk M, Vinson M. Unlocking the in chronic illness care: a collaborative approach. Jt
potential of frontline managers. New York: McKinsey Comm J Qual Improv. 2001;27(2):6380.
& Companies; 2009. [05/15/2016]. Available from: 21. Savage MW, Dhatariya KK, Kilvert A, Rayman G,
http://www.mckinsey.com/business-functions/ Rees JA, Courtney CH, et al. Joint British Diabetes
organization/our-insights/unlocking-the-potential-of- Societies guideline for the management of diabetic
frontline-managers. ketoacidosis. Diabet Med. 2011;28(5):50815.
8. McKinsey & Company. How companies manage the 22. Simmons D. Impact of an integrated approach to dia-
front line today: McKinsey survey results (2010) betes care at the Rumbalara Aboriginal Health
[05/15/2016]. Available from: http://www.mckinsey. Service. Intern Med J. 2003;33(12):5815.
com/business-functions/organization/our-insights/ 23. NCQA. Comprehensive diabetes care (n.d.)
how-companies-manage-the-front-line-today- [05/13/2016]. Available from: http://www.ncqa.org/
mckinsey-survey-results report-cards/health-plans/state-of-health-care-
9. Simmons D, Lillis S, Swan J, Haar J. Discordance in quality/2015-table-of-contents/diabetes-care
perceptions of barriers to diabetes care between 24. Permanente K. Measuring care quality: Southern
patients and primary care and secondary care. California Region. Rockville: Kaiser Permanente; 2015.
Diabetes Care. 2007;30(3):4905. 25. Permanente K. Measuring care quality: Northern
10. Simmons D, Wenzel H. Diabetes inpatients: a case of California Region. Rockville: Kaiser Permanente; 2015.
lose, lose, lose. Is it time to use a diabetes-attributable 26. NHS. The Information Centre. National Diabetes
hospitalization cost to assess the impact of diabetes? Audit: PCT summary. Year 2009/10 (n.d.)
Diabet Med. 2011;28(9):112311230. [05/15/2016]. Available from: http://www.hscic.gov.
11. Relman AS. Dealing with conflicts of interest. N Engl uk/catalogue/PUB06517/nati-diab-audi-pct-lhb-eng-
J Med. 1985;313(12):74951. scen-port-pct-rep.pdf
12. Schaffer W, Holloman F. Consultation and referral 27. Diabetes Specialist Centre. Pattern review and out-
between physicians in new medical prac-tice environ- comes from SMBG data in primary care. An evaluatu-
ments. Ann Intern Med. 1985;103:6005. ion of communication and technology support.
13. Simmons D, English P, Robins P, Craig A, Rush Diabetes Voice Summer 2014(26).
A. Should diabetes be commissioned through multi- 28. Action to Control Cardiovascular Risk in Diabetes
disciplinary networks, rather than practice based com- Study Group. Effects of intensive glucose lowering in
missioning? Primary Care Diabetes. Type 2 diabetes. N Engl J Med.
2011;5(1):3940. 2008;358(24):254559.
14. Shih A, Davis K, C SS, Gauthier A, Nuzum R, 29. Huxley RR, Peters SA, Mishra GD, Woodward
McCarthy D. Organising the U.S. Health Care M. Risk of all-cause mortality and vascular events in
Delivery System for High Performance. The women versus men with type 1 diabetes: a systematic
Commonwealth Fund (2008) [05/12/2016]. Available review and meta-analysis. Lancet Diabetes
from: http://www.commonwealthfund.org/usr_doc/ Endocrinol. 2015;3(3):198206.
Shih_organizingushltcaredeliverysys_1155.pdf 30. Coulter A, Roberts S, Dixon A. Delivering better ser-
15. Inzucchi SE, Bergenstal RM, Buse JB, Diamant M, vices for people with long term conditions. Building
Ferrannini E, Nauck M, et al. Management of hyper- the house of care. London: The Kings Fund; 2013.
glycemia in type 2 diabetes, 2015: a patient-centered 31. Pringle M, Stewart-Evans C, Coupland C, Williams I,
approach: update to a position statement of the Allison S, Sterland J. Influences on control in diabetes
American Diabetes Association and the European mellitus: patient, doctor, practice, or delivery of care?
Association for the Study of Diabetes. Diabetes Care. BMJ Br Med J. 1993;306(6878):6304.
2015;38(1):1409. 32. Cebolla Garrof B, Bjrnberg A, Yung Phang A. Euro
16. Wagner EH, Austin BT, Von Korff M. Improving out- diabetes index 2014. Health Consumer Powerhouse
comes in chronic illness. Manag Care Q. Ltd; 2014.
1996;4(2):1225. 33. Davis K, Stremikis K, Schoen C, Squires D. Mirror.
17. Wagner EH, Austin BT, Von Korff M. Organizing Mirror on the wall, update: how the U.S. Health Care
care for patients with chronic illness. Milbank Q. System compares internationally. The Commonwealth
1996;74(4):51144. Fund; 2014.
18. Wagner EH, Bennett SM, Austin BT, Greene SM, 34. Tricco AC, Ivers NM, Grimshaw JM, Moher D,
Schaefer JK, Vonkorff M. Finding common ground: Turner L, Galipeau J, et al. Effectiveness of quality
patient-centeredness and evidence-based chronic ill- improvement strategies on the management of diabe-
ness care. J Altern Complement Med. 2005;11 Suppl tes: a systematic review and meta-analysis. Lancet.
1:S715. 2012;379(9833):225261.
19. Wagner EH, Austin BT, Davis C, Hindmarsh M, 35. Piwernetz K. DIABCARE Quality Network in
Schaefer J, Bonomi A. Improving chronic illness Europe a model for quality management in chronic
care: translating evidence into action. Health Aff diseases. Int Clin Psychopharmacol. 2001;16 Suppl
(Millwood). 2001;20(6):6478. 3:S513.
20. Wagner EH, Glasgow RE, Davis C, Bonomi AE, 36. US Department of Health & Human Services. EHR
Provost L, McCulloch D, et al. Quality improvement Incentives & Certification [05/24/2016]. Available
from: https://www.healthit.gov/providers-professionals/ 40. OECD OECD. Stat. Income distribution and pov-
meaningful-use-definition-objectives erty: median disposable income at current prices
37. Kings Fund. Consultancy report: partnership for sus- (n.d.) [05/12/2016]. Available from: http://stats.oecd.
tainable health care development: options paper for org/Index.aspx?DataSetCode=IDD
approaches to governance and clinical governance for 41. Simmons D, Hartnell S, Watts J, Ward C, Davenport
an integrated care approach to diabetes. Final report. K, Gunn E, et al. Effectiveness of a multidisciplinary
London; October 2010. team approach to the prevention of readmission for
38. Simmons D, Peng A, Cecil A, Gatland B. The per- acute glycaemic events. Diabet Med.
sonal costs of diabetes: a significant barrier to care in 2015;32(10):13617.
South Auckland. N Z Med 42. Murray NJ, Gasper AV, Irvine L, Scarpello TJ,
J. 1999;112(1097):383285. Sampson MJ. A motivational peer support program
39. International Diabetes Federation Europe. Access to for type 2 diabetes prevention delivered by people
quality medicines and medical devices for diabetes with type 2 diabetes: the UEA-IFG feasibility study.
care in Europe; 2013. Diabetes Educ. 2012;38(3):36676.
Index
C G
Cardiovascular outcomes, 77 German healthcare system, 173
Care delivery, 5, 12, 13, 16, 1820, 25, 28, 29, 32, 34, Gesundes Kinzigtal, 169184
35, 51, 65, 67, 71, 76, 81, 198, 207, 217, 220, 242 Glycemic control, 37, 56, 69, 71, 74, 78, 81, 96, 103
China, 65, 81
Community-based, 16, 107, 109, 133136, 150, 159, 228
Community work, 228, 230 H
Consultants, 107, 109, 116, 124, 131, 133, 134, 138, 151, HbA1c, 19, 20, 26, 3134, 3744, 5157, 60, 61,
160, 217, 220, 222, 224, 227, 228, 230, 231 67, 69, 70, 76, 78, 79, 81, 89, 93, 94, 99,
Consultations, 16, 22, 40, 43, 66, 88, 91, 99, 100, 112, 102, 107, 109111, 116, 117, 134, 139, 140,
116, 134, 135, 139, 155, 156, 163, 178, 179, 147, 149151, 153155, 160, 163, 164, 166,
190192, 195, 215, 217, 218, 223 177, 193196, 211, 212, 217, 234, 235,
Continuous improvement, 17, 35, 79, 81, 101, 132, 135, 237239, 241, 242
136, 190, 193, 224, 234 Health coaching, 38, 44, 45
Cost-effectiveness, 7, 44, 45, 54, 77, 88, 89, 99, 103, Health information technology, 34, 44
148150, 156159, 165, 206, 231 Healthcare performance, 203205
Hong Kong, 6582, 236, 241, 244246
Hospitalisation, 7, 109, 111, 112, 116120, 126128,
D 189, 203, 204, 234238, 241, 245
Diabetes, 18, 17, 19, 20, 22, 29, 33, 5162, 6577,
7981, 87105, 107129, 133137, 140142,
147, 148, 150167, 175180, 182, 186, 189193, I
206212, 215225, 227231, 233, 234, 237, Individualized Multidisciplinary Team-care Model,
240246 17, 26, 29
Diabetes care, 3, 8, 1118, 20, 26, 29, 3145, 51, 52, 56, Innovations, 31, 91, 149, 159, 160, 164, 165,
59, 62, 65, 68, 71, 72, 76, 77, 80, 81, 88, 89, 91, 167, 171, 193
92, 97, 100, 101, 103, 108, 109, 113114, 116, Integrated care, 18, 12, 13, 18, 3134, 37, 38, 40,
118, 129, 131, 133, 136, 140, 144, 149, 159, 4245, 68, 90, 97, 102, 114, 118, 128, 129, 132,
163166, 179, 180, 189195, 206208, 211, 212, 135, 138, 140, 141, 143, 148, 170, 171, 174,
215225, 228, 231, 233235, 237, 242, 244246 178180, 182, 190, 193, 198, 205, 206, 218221,
Diabetes education, 20, 37, 40, 54, 55, 57, 59, 66, 76, 78, 223, 225, 227231, 233244, 246
80, 88, 93, 103, 109, 110, 115, 118, 215, 244 Integrated diabetes care, 18, 26, 3145, 6582, 87105,
Diabetes educators, 8, 51, 5460, 98, 107, 233 107121, 131144, 147167, 169198, 201212,
Diabetes expenditures, 147, 171, 186 215225, 227231
Diabetes mellitus, 11, 88, 97, 170, 177, Integrated healthcare, 15, 16, 170, 171
190192, 206, 209 Interdisciplinary working groups, 5
Digital patient data recording, 66 Interventions, 1, 6, 7, 12, 14, 25, 29, 3234, 3742, 44,
45, 51, 5356, 6769, 74, 7681, 88, 89, 101,
108, 110, 115, 117119, 126129, 148151,
E 154157, 159, 160, 163, 165, 175, 176, 178180,
Education, 3, 6, 7, 12, 15, 18, 21, 29, 33, 34, 38, 4043, 217, 234, 238, 246
45, 5155, 59, 66, 68, 76, 79, 80, 89, 99, 101,
113, 118, 129, 134, 135, 139, 143, 148, 149, 159,
163, 165, 166, 178, 208, 211, 212, 215, 219, 221, J
224, 227229, 231, 240, 244, 245 Joint Venture Organisations, 133

D. Simmons et al. (eds.), Integrated Diabetes Care, DOI 10.1007/978-3-319-13389-8
250 Index
K Q
Knowledge translation, 81 Quality improvement, 5, 16, 35, 5253, 59, 65,
68, 71, 76, 79, 81, 136, 144, 190, 193,
234, 242
L Quality of care, 4, 6, 34, 37, 39, 42, 5162, 88,
Logistics, 76, 82 132, 147, 160, 182, 185, 203, 204, 206,
215, 222, 236
M
Managed care, 34, 89, 95, 96 R
Medical curriculum, 230 Randomised trial, 158
Medical training, 227230 Registries, 29, 34, 37, 43, 62
Risk factors, 7, 32, 37, 44, 51, 54, 55, 67, 6972,
74, 76, 78, 79, 94, 96, 99, 101, 147, 149,
O 150, 154, 155, 159, 160, 163, 164, 166,
Organizations, 24, 6, 12, 14, 32, 34, 39, 52, 71, 81, 99, 175, 206, 211
101, 103, 128, 170172
S
P Secondary/tertiary care, 3, 16
Patient Centered Medical Home (PCMH), 13, 34, 35, Single budget, 133, 220, 245
3739, 4245, 6061 Swedish Healthcare System, 205
Patient perspective of integrated and coordinated
care, 3, 171
Peer-review, 68 T
Person-centered, 94, 99, 100, 103 Toyota Production System, 17, 29
Pillars of integration, 132, 136 Triple aim, 6
Policy, 2, 5, 8, 16, 34, 39, 6582, 102, 103, 115, 127, Type-2-diabetes mellitus, 170
170, 188, 201, 205, 206, 216, 227, 228, 234, 235,
237, 243
Postgraduate, 68, 227229, 231 U
PPCSM, 18, 29 Unambiguous work flow
Predefined pathways, 18 connections, 20
Primary care, 1, 3, 4, 8, 16, 18, 3235, 3740, 43, 44, Undergraduate, 228, 229
5160, 62, 73, 76, 87, 91, 102, 103, 108111,
118, 119, 124, 126, 127, 131, 132, 134, 135, 139,
140, 147149, 155, 158160, 163, 165, 186, 188, W
189, 203205, 208212, 216, 221, 222, 230, 231, Whole system integration, 132, 218, 237
234, 235, 238, 241, 245, 246
Primary care Chronic Care Model, 32

Integrated DM Care - Multidisciplinary Approach

Uploaded by

Copyright:

Available Formats

Integrated DM Care - Multidisciplinary Approach

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Integrated DM Care - Multidisciplinary Approach

Uploaded by

Copyright:

Available Formats

Integrated

ISBN 978-3-319-13388-1 ISBN 978-3-319-13389-8 (eBook)

Library of Congress Control Number: 2016955045

Springer International Publishing Switzerland 2017

Printed on acid-free paper

This Springer imprint is published by Springer Nature

If one wished to choose a health condition that requires an integrated multi-

Professor Paul Zimmet, AO MD PhD FRACP FRCP (London) FTSE

Sydney, Australia David Simmons

1 An Introduction to Integrated Care and Diabetes

10 Integrated Diabetes Care in Germany: Triple Aim

Index ..................................................................................................... 249

Heather Bird Diabetes UK, London, UK

Mona Landin-Olsson Department of Endocrinology, Lund University

Roseanne O. Yeung Department of Medicine, Division of Endocrinology,

Background delivery is now fragmented [2]. Such fragmenta-

Negative coordination is characterised by

Springer International Publishing Switzerland 2017 1

High Multidisciplinary teams manage all care

Identify population at risk

Identify emergent need

Linkage Coordination Integration

However, disease management programmes

R. Harsha Rao and Peter Perreiah

Introduction pleting all the mandated process measures for

Springer International Publishing Switzerland 2017 11

Basic eligibility: VA healthcare benets are cinations, or screening for hypertension,

Fig. 2.2 Likers 4P Tenets adapted to an Individualized

More advanced skills for patients In patients on a Multiple Daily Insulin

Fig. 2.3 Changes made to

Provider Patient Interactions a net of eight patient appointments of 30 min

Conclusions lary care providers to set and attain care

David H. Thom and Thomas Bodenheimer

Introduction our approach was developed within the larger

Springer International Publishing Switzerland 2017 31

Table 3.1 Key characteristics of the Patient-Centered Medical Home

Fig. 3.3 Ten building blocks of high-performing primary care

Telemedicine and Digital Health Another promising use of digital technology

Janice C. Zgibor, Francis X. Solano Jr.,

Introduction cant improvement in diabetes outcomes on a

Springer International Publishing Switzerland 2017 51

While several models of care are reported in Provider-Based Strategies

to dilution by the inux of patients with better Patient-Centred Strategies-Diabetes

Association (ADA), Institute of Medicine and Demonstration Projects

Fig. 4.1 HbA1c over time in UPMC Glucose to Goal programme

Roseanne O. Yeung, Junmei Yin,

Introduction public hospitals and institutions, 47 Specialist

R.O. Yeung (*)

Springer International Publishing Switzerland 2017 65

Fig. 5.1 Integrated diabetes

entered by clerical staff to be mailed to the refer-

Facilitating Knowledge Transfer

The process of knowledge transfer in clinical

(i) The patient conveys information to one or

Each health-care system has its unique chal-

In addressing the first two steps, the team

Fig. 5.3 (a, b) KaplanMeier

vention of complications. A biobank was estab- cuhk.edu.hk/Programmes/MScinEndocrinolo

Community Engagement: OPAL comparisons were made between those with

Fig. 5.5 The JADE

Introduction emphasis on reducing this cost and improving

I nefficiency DuetoLack improved glycaemic control, lower blood pres-

Aims andObjectives 1 . Severe cognitive impairment

Primary Outcome is normally distributed. Equivalently, a separate

willingness to pay (WTP) for one quality- Cost-Effectiveness

introduction of QoF in 2004 enabled cardiovas- Objectives