SPECIAL ARTICLE

Recommended Measures for the

Assessment of Behavioral Disturbances
Associated With Dementia
Yun-Hee Jeon, R.N., B.H.Sc. (Nursing), M.N., Ph.D.,
Jan Sansoni, B.A., Dip. Ed., M.Sc., M.A.P.S., Lee-Fay Low, B.Sc. (Psych, Hons), Ph.D.,
Lynn Chenoweth, R.N., B.A., M.A. (Hons), M.Ad.Ed., Ph.D.,
Siggi Zapart, B.Psych. (Hons), Emily Sansoni, B.Psych. (Hons),
Nicholas Marosszeky, B.A. (Hons), M.A.P.S.

Objective: The study reviewed all the published instruments used for the assess-
ment, diagnosis, screening, and outcomes monitoring/evaluation of behavioral dis-
turbances associated with dementia (BDAD) to recommend a set of psychometrically
valid measures for clinicians and researchers to use, across a range of different prac-
tice settings. Methods: The study involved a broad scoping search, followed by a series
of in-depth literature reviews on 29 instruments using scientific literature databases
(MEDLINE, PsycINFO, CINAHL, and the Cochrane Library) and various national,
international government, and government agency websites and professional orga-
nization websites. External consultations from measurement, clinical and research
experts in dementia care, consumer representatives, and policy/decision makers, were
sought in selecting the best instruments and in making the final recommendations.
Findings: Key attributes and psychometric properties of a short list of five instruments
were measured against prespecified criteria. The Neuropsychiatry Inventory (NPI)
and the Behavioral Pathology in Alzheimer’s Disease Rating Scale (BEHAVE-AD)
were rated as the best measures for assessment of behavioral disturbances, followed
by the Consortium to Establish a Registry for Alzheimer’s Disease-Behavior Rating
Scale for Dementia (CERAD-BRSD), the Dementia Behavior Disturbance Scale, and
the Neurobehavioral Rating Scale. Conclusion: The use of valid and standardized
outcome measures for the assessment of BDAD is critical for epidemiological studies,
prevention, early intervention and treatment of dementia conditions, and funding
for relevant healthcare services. The review recommends the NPI and BEHAVE-AD as

Received June 24, 2009; revised May 18, 2010; accepted May 21, 2010. From the Sydney Nursing School, the University of Sydney (Y-HJ); the
Australian Primary Health Care Research Institute, the Australian National University (Y-HJ); Australian Health Outcomes Collaboration (JS); Centre
for Health Service Development, University of Wollongong (JS); the Dementia Collaborative Research Centre, School of Psychiatry, University of
New South Wales (L-FL); University of Technology Sydney (LC); Health and Ageing Research Unit, South Eastern Sydney-Illawarra Area Health
Service (LC); Centre for Health Equity Training Research and Evaluation, Part of the Centre for Primary Health Care and Equity, University of
NSW and a Unit of the Division of Population Health, Sydney South West Area Health Service (SZ); the Australian Primary Health Care Research
Institute, the Australian National University (ES); and Centre for Health Service Development, the University of Wollongong (NM), Australia. Send
correspondence and reprint requests to Yun-Hee Jeon, R.N., B.H.Sc. (Nursing), M.N., Ph.D., Faculty of Nursing & Midwifery, the University of
Sydney, 88 Mallett Street, Camperdown, NSW 2050, Australia. e-mail: [email protected]
c 2011 American Association for Geriatric Psychiatry
DOI: 10.1097/JGP.0b013e3181ef7a0d

Am J Geriatr Psychiatry 19:5, May 2011 403

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Assessment of BDAD

the most appropriate measures for both clinical and research, whereas the CERAD-
BRSD is suited better for research. The review was designed for the Australian context;
however, the findings are applicable in other developed countries. (Am J Geriatr
Psychiatry 2011; 19:403–415)
Key Words: Behavioral disturbances associated with dementia, dementia, outcome
measures

S ixty-one percentage to 88% of community-

dwelling people with dementia experience some
form of behavioral disturbances associated with de-
weaker association with anxiety and mood disorders,
delusions, and hallucinations.12
Observed patterns of BDAD reveal the prevalence
mentia aq: 4 (BDAD).1 BDAD is considered a lead- of different symptoms at different stages of the dis-
ing cause of institutionalization for the person with ease,7,13 with some symptoms worsening and others
dementia,2 and the prevalence of BDAD is particu- lessening with disease progression.14 This has led to
larly high in nursing homes at approximately 80%- inconsistent findings linking BDAD and cognitive de-
91%.3,4 However, it has been just over 2 decades cline in dementia. There is some evidence that behav-
since BDAD were recognized as defining noncog- ioral, rather than psychological, symptoms are asso-
nitive features of dementia.5 It is generally agreed ciated with cognitive deterioration.13,15 There is also
that BDAD include physical aggression, screaming, evidence that premorbid intelligence level and per-
restlessness, agitation, wandering, culturally inap- sonality can influence the type and course of BDAD.14
propriate behaviors, sexual disinhibition, hoarding, Inconsistent research findings have led to a lack of
cursing, apathy and shadowing, anxiety, depressive general agreement about which BDADs are likely to
mood, hallucinations, delusions, and psychosis.2 be present at various stages of dementia.13
The presence of BDAD signals a sense of unease Early detection and accurate identification of
or distress in the person with dementia and con- BDAD are essential for diagnosis and management
tributes to carer stress and burden, and often requires of persons with dementia, and these can be achieved
treatment to reduce the severity of the behavior.6,7 by routine assessment of persons with changes
Because of the care-intensive nature of BDAD, the is- in behavior thought to be related to dementia.7
sues of identification and management are of partic- Increased identification and recognition of BDAD
ular importance to families and carers, as well as to may ultimately improve the person’s quality of life,
the person with dementia. BDAD can impact on the reduce the person’s and caregiver’s distress, and re-
quality of life of the person with dementia and their duce caregiver burden, if education, treatment, and
family/carers, cause stress to those who provide for- support are provided.9 To achieve these outcomes,
mal and informal care, increase financial costs and the first step is accurate identification of BDAD to
early nursing home admission, and worsen progno- rule out treatable causes. One of the difficulties for
sis and the person’s capacity to function in everyday health professionals is the complex range of avail-
life.8–11 able dementia assessment tools and procedures. The
The presentation of BDAD can vary according to availability of more than 100 assessment measures
the type of dementia. For example, depressive mood makes the selection of global BDAD and BDAD-
may be more common in vascular dementia, apa- specific tools difficult for caregivers and health
thy in Alzheimer disease (AD), and hallucinations in staff.11
dementia with Lewy bodies.4,6,7 Furthermore, some This article provides a review of measures/tools
types of dementia (e.g., vascular) may have a lower that assess BDAD and makes recommendation about
frequency of BDAD compared with other types.8 some of the best available instruments suitable for
Other less common forms of dementia, such as fron- routine clinical and research use. These recommen-
totemporal dementia, are also thought to carry their dations are the product of the Dementia Outcome
own distinct patterns of BDAD such as increased ap- Measurement Suite (DOMS) project,16 an Australian
athy, disinhibition, and loss of insight and have a Government Dementia Initiative. The DOMS project

404 Am J Geriatr Psychiatry 19:5, May 2011

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
 Jeon et al.

team undertook a comprehensive review of mea- ments for detailed psychometric review using the
surement instruments in routine use for assess- initial selection criteria outlined in Table 1. These
ing, diagnosing, screening, and monitoring demen- criteria were developed based on clinical experience
tia and evaluating treatments that are applicable of specialist clinicians and on advice from local and
for the Australian healthcare context and for other international dementia experts in both clinical and
English-speaking developed countries.16 research arenas. One important criterion was that
the instruments be applicable for use in a range of
healthcare settings by staff and caregivers with dif-
METHODS
ferent levels of experience with BDAD assessment.
An initial broad literature search was undertaken Although some expert clinicians were more famil-
by the DOMS review team in 2007 on the MED- iar with dementia instruments not selected for the
LINE and PsyclNFO databases using terms associ- final review, they nevertheless agreed that the top
ated with dementia assessment. The major texts in the five selected instruments met all the review criteria
field were also examined. They included psychomet- (Stage IV-Step 1).
ric texts containing instrument reviews17,18 and texts Table 2 provides summary information for the five
containing instrument reviews applicable for demen- selected BDAD instruments.
tia and assessment of the elderly.19–21 Altogether, The top five BDAD instruments selected for de-
138 instruments that assessed BDAD were identi- tailed review were the NPI, BEHAVE-AD, CERAD
fied (Stage I). The review team then identified 29 BRSD, NRS, and DBDS. Columbia University
instruments that addressed global or overall BDAD Scale for Psychopathology in Alzheimer’s Disease
(Stage II). An impact sheet was then developed was not selected for detailed review because of its
for consideration by the review team including overly detailed emphasis on delusional aspects of
MEDLINE citations, major textbooks and web im- BDAD and its insufficient coverage of behavioral
pacts, presence in instrument databases (e.g., Patient- issues; the Manchester and Oxford Universities Scale
Reported Outcome and Quality of Life Instrument
Database), and its use in clinical practice (by a field
survey). Using this framework to examine and select TABLE 1. Selection Criteria
high-impact instruments, a short list of nine preferred
Selection Criteria
global BDAD instruments was produced (Stage III),
including the following: • Whether there is a copy of the instrument and the original article
concerning its development available.
r Behavioral Pathology in Alzheimer’s Disease • The number of citations found. In the case of new instruments,
some care was taken to assess this criterion as it was considered
Rating Scale (BEHAVE-AD)22 ;
r Consortium to Establish a Registry for that recently developed instruments may not have a high citation
rate. However, for instruments developed more than 5 years
Alzheimer’s Disease-Behavior Rating Scale previously a low citation rate might indicate limited adoption by
the field.
for Dementia (CERAD-BRSD)23–25 ; • The amount and range of the published psychometric evidence.
r Columbia University Scale for Psychopathol- • Whether the instrument is used in clinical practice (evidence
ogy in Alzheimer’s Disease26 from the literature and data from the National Expert Panel and
r Dementia Behavior Disturbance Scale (DBDS)27
field surveys).
• The availability of normative and clinical reference data.
r Manchester and Oxford Universities Scale for the • Administration time (generally 30 minutes or less) where a
shorter administration time would be preferred. It was noted that
Psychopathological Assessment of Dementia28
r Neurobehavioral Rating Scale (NRS)29
as a number of instruments assessing different aspects (e.g.,
symptoms, cognition, and HRQOL) will need to be utilized,
r Nursing Home Behavior Problem Scale30 lengthy instruments that may be more appropriate for detailed
r Neuropsychiatric Inventory (NPI)31
follow-up assessment may not be appropriate for use in routine
r Revised Memory and Behavior Problem Check-
assessment and across the range of practice settings.
• Whether the instrument is applicable for people with varying
list (RMBPC).32 levels of dementia severity.
• Proprietary considerations (e.g., prohibitive cost).
Assisted by a team of experts in psychological • Applicability for use in routine care. Instruments would be
preferred if they did not require specialist skills for administration
measurement and dementia specialist clinicians, the or if extensive training in their use was not required (e.g., as for
review team then selected the top five BDAD instru- many neuropsychological/medical assessments).

Am J Geriatr Psychiatry 19:5, May 2011 405

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
 406
TABLE 2. Summary Information for the Five Selected BPSD Instruments

Instruments Applicable for All Training/Manual

Domains/Subdomains Stages (Proxy) Score Range (Availability/Cost) Time Psychometrics Use in Practice to Date

BEHAVE AD
Assessment of BDAD

Twenty-five items grouped All stages () 0–45 or 0–297 No formal training, 20 minutes Good Assessment of BDAD in the
into seven major categories. for BEHAVE- although psychiatric community, outpatients,
Paranoid and delusional AD-FW language in the scale and residential care.
ideation, hallucinations, means that it should be Measurement of change in
activity disturbance, used by a person with pharmacological and
aggressiveness, diurnal some health training. nonpharmacological trials
rhythm disturbances, (Free)
affective disturbances, and
anxieties and phobias
CERAD-BRSD
Forty-six items grouped into All stages () 0–164 Provided with instrument as 20–30 minutes Very good Assesses/evaluates behavioral
six domains. Depressive per availability/cost disturbance in people with
features, inertia, psychotic section. (Cost of dementia or cognitive
features, vegetative features, US85.00 for instrument impairment and
Irritability/aggression, and plus instruction manual) effectiveness of drug
behavioral dysregulation treatments or other
nonpharmacological
interventions
DBDS
Twenty-eight items. All stages () 0–112 No formal training required. 15 minutes Good Assessment of behavioral
Subdomains not defined (Free) disturbance in clinical
settings and residential care
NRS
Twenty-seven items consisting All stages () 0–168 for NRS or No formal training required, 15–20 minutes Good Used extensively to assess
of cognitive deficits, 0–81 for NRS-R though originally psychiatric symptoms in
psychiatric symptoms, and completed after a head injury patients. Less
behavioral disturbances structured interview and frequently used for people
clinical observation. with dementia, although it
(Free) has been used in a few
pharmaceutical trials
NPI
Twelve domains with five to All stages () 0–144 Specific training not 10–15 minutes Very good to Assesses psychopathology in
eight items per domain: required, but training excellent dementia to distinguish
delusions, hallucinations, pack available to assist between different causes
agitation, depression, staff recording caregivers’ and include symptoms
anxiety, euphoria, apathy, responses. (Free) rare in Alzheimer disease
disinhibition, irritability, but characteristic of
aberrant behaviors, night frontotemporal dementias
time behaviors, and and caregiver distress
appetite changes. One item associated with behavioral
on carer stress symptoms

Am J Geriatr Psychiatry 19:5, May 2011

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
 Jeon et al.

for the Psychopathological Assessment of Dementia Once the detailed reviews were completed using
for the lengthy time required for administration; the Instrument Review Sheet and the requisite test
Nursing Home Behavior Problem Scale for its focus information found, the instruments were scored on
on difficulties perceived by nursing staff; and RMBPC a 10-item assessment system, which combined psy-
for its limited coverage on the various aspects of chometric evidence identified in the scientific litera-
BDAD and limited applicability in the nursing home ture with practical information on each test’s feasi-
setting (e.g., staff may not have sufficient knowledge bility and applicability in healthcare (Table 3). Each
about residents with dementia to answer some of instrument was given a score against each criterion,
the questions). Despite the relatively small number and this was multiplied by the weight for this crite-
of citations identified compared with the other four rion. The resulting weighted score for each criterion
selected instruments, the DBDS was chosen for its was then added to form a total score for each instru-
proven applicability in both community and nursing ment as shown in Table 3. This comparative table of
home settings, easy availability and implementation, scores formed the basis for the final recommenda-
and short time required for administration. tions concerning BDAD instruments (Stage VI). The
A National Expert Panel consisting of experienced criteria and weights for instrument rating were de-
clinicians, researchers, policy makers, and consumer cided by the Expert Measurement Group (DOMS),
representatives in dementia then provided feedback consisting of experienced dementia clinicians and
on the suitability of each of the five selected instru- experts in psychological measurement. These crite-
ments for use in a range of health and care settings ria reflect those identified in the Instrument Review
by staff with varying levels of expertise in dementia Sheet. Criteria given the higher weight (i.e., 3) were
and by caregivers (Stage IV-Step 2). considered essential, and the most important criteria
These five instruments then underwent detailed in terms of its psychometric properties and clinical
psychometric review into the instrument’s avail- and research relevance.
ability, applicability, requirements for administra- As the original review was conducted in 2007,
tion, psychometric properties (reliability, validity, re- an update of the dementia measurement literature
sponsiveness, sensitivity, and specificity), and the was conducted in April 2009. This included a cita-
availability of normative and clinical reference data tion search on key articles from the initial DOMS
(Stage V). For the detailed reviews, further literature project literature review. Peak body websites rel-
searches were undertaken on the chosen instruments, evant to dementia including those of Alzheimer’s
including searches on the instrument’s name and Australia, the International Psychogeriatric As-
common abbreviation(s), as well as a citation search sociation, and Australian Institute of Health and
of the original article(s). This searching included Welfare were also searched for new publica-
additional databases (CINAHL and the Co-chrane tions. This identified an additional 34 relevant
Library) and web searches (e.g., Google). publications.

TABLE 3. Summary of Ratings for BPSD Global Instruments

Criteria (Scoring System: 1–3,

the Higher the Better)Instrument Weight NPI BEHAVE-AD CERAD-BRSD DBDS NRS

Availability of comparison data 3 2 2 2.5 1 2

Length/feasibility of instrument for 2 3 2 1 2 2
inclusion in battery
Complexity of administration (for clinician 2 3 3 2 2 1
use) and cognitive burden
Cultural appropriateness 1 3 3 2 2 2
Ease of obtaining score 2 3 3 3 3 3
Sensitivity to dementia 3 3 3 3 2 2
Reliability evidence 3 3 3 3 3 3
Validity evidence 3 3 3 3 2 2
Cost of the instrument 2 3 3 1 2 2
Cost of instrument administration 2 2 2 2 2 1
weighted total 64 62 54.5 50 49

Am J Geriatr Psychiatry 19:5, May 2011 407

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Assessment of BDAD

nighttime behaviors and eating behaviors. Currently,

FINDINGS the NPI itself is regarded to be the gold standard in
the measurement of BDAD and is widely used in clin-
Tables 2 and 3 provide a summary of the key at-
ical trials, community investigations, and large- scale
tributes and psychometric properties of the final five
population studies.4,6,43,44
instruments measured against the weighted criteria.
Internal consistency was established (0.75–0.89) for
Detailed findings for each of the three most highly
each item/subscale of the NPI.31 Other studies have
rated contenders for BDAD global measures are de-
also found high levels of internal consistency for the
scribed below.
NPI subscales.45,46 Internal consistency of the NPI-
NH in nursing home patients with dementia was
found to be 0.67.47
Neuropsychiatric Inventory
Test-retest reliability of NPI and NPI-NH is high,
The NPI was developed and validated by Cum- with overall correlations of 0.79 for frequency and
mings et al.31 to assess psychopathology in people 0.86 for severity of symptoms within 3 weeks
with dementia and to help distinguish between dif- of the first test and more than 0.79 for fre-
ferent causes of dementia. It includes items pertain- quency and severity of all items 3 weeks later.31
ing to symptoms known to be rare in AD but char- Interrater reliability has been established for the
acteristic for frontotemporal dementias. The NPI has NPI, NPI-NH, and NPI-Q.31,45,46,48 Between-rater re-
been used in large-scale studies of patients with de- liability is reported to vary from 71% to 100% for
mentia in Special Care Units,33 a homogenous sam- different items and from 80% to 100% for the to-
ple of people with AD6 and mixed samples of people tal score.31,45,49 However, test-retest reliability when
with vascular dementia, dementia with Lewy bodies, conducted with staff of different levels of expertise
or AD.8,34 Initially developed for use in the inpatient found that Certified Nursing Aids’ ratings corre-
clinical setting, the NPI is based on a structured in- lated only moderately well, especially for residents
terview with a health worker or caregiver who is fa- with high levels of neuropsychiatric disturbance.49
miliar with the person and has direct knowledge of It is suggested that a trained observer or the pa-
the person’s behavior over the recent past. There is tient’s primary nurse/care staff, with direct knowl-
a nursing home version (NPI-NH) and a shortened edge of the patient’s behavior over the recent past,
version, the NPI-Q. is a reliable source of information for the NPI-NH
The NPI has been used to measure behavioral version.31,45,49
distinctions associated with frontotemporal demen- Medical and other health clinicians agree that the
tia and AD35 ; measure behavioral correlates of cere- items on the NPI, NPI-NH, and NPI-Q compare
bral blood flow in AD36 ; and identify the link be- favorably with their own clinical assessment pro-
tween behavioral disturbance in dementia and body cesses. These use a combination of health and social
mass index.37 The NPI is effective at measuring history, psychiatric examination, Mini-Mental State
change in relationship to drugs, nutrition, and treat- Examination (MMSE) test scores,48 neurologic exam-
ments/therapies. Scores on the NPI and NPI-NH ination, and rating of behavioral disturbance with
have been shown to be sensitive to drug treatments other validated measures such as the Brief Psychiatric
(e.g., rivastigmine, olanzapine, tacrine, donepezil, Rating Scale (BPRS)46 and the Cohen-Mansfield Agi-
memantine, and dimebon) with scores improving tation Inventory (CMAI).50
significantly.38–42 However, the NPI-Q is not suit- The NPI subscales also correlate well (p = 0.001)
able for use in pharmacological trials because of with the BEHAVE-AD, Hamilton Depression Rat-
its brevity and reliance on caregiver report. Only a ing Scale, Clinical Dementia Rating Scale,47 and
few of the NPI categories show a minimal response BPRS.46 All five factors identified in the NPI are
in nondemented controls, indicating that the older similar to the factors identified in other vali-
person without cognitive impairment/dementia has dated measures of behavioral disturbance in de-
hardly any of the symptoms identified by the NPI.31 mentia, including the agitation factor in the CMAI;
Two domains not covered in similar instruments of sleep/aberrant motor activity factor in the Geri-
psychopathology in dementia were added as result: Sub-Acute and Non-Acute Patient Classification

408 Am J Geriatr Psychiatry 19:5, May 2011

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
 Jeon et al.

(SNAP); elevated behavior factor in the Mania Rat- then decreasing for severe dementia.54 BEHAVE-AD
ing Scale; aggression and compliance factor in the scores were related to cognitive decline in a sam-
Geri-SNAP; mood in the Cornell Depression Scale ple of 142 Taiwanese patients.56 The scale correlates
and the BPRS; and the psychosis factor in the with other measures of behavioral disturbance such
Geri-SNAP.39,46–48 as the CMAI and the NPI.57,58 In 62 community-
dwelling people with dementia, increased behavioral
Behavioral Pathology in Alzheimer’s Disease disturbance as measured by the BEHAVE-AD was as-
Rating Scale sociated with increased recognition of dementia by
family informants.59
The BEHAVE-AD22,51 is a clinician rated scale de- The BEHAVE-AD has been used in BDAD preva-
veloped to measure change in behavioral distur- lence studies, for example, comparing the preva-
bance in people with AD. It comprises 25 symptoms lence of BDAD between institutionalized residents
grouped into seven categories (refer to Table 2) rated and memory clinic outpatients with AD,60 as well
over the previous 2 weeks based on an informant in- as the outcome measure in several clinical trials that
terview. It was one of the first scales developed for the have evaluated the effects of different models of nurs-
purpose of measuring behavioral disturbance in peo- ing home care on behavioral disturbance,61 and phar-
ple with AD and was developed by clinical experts macological effects.15,57,62,63 Both symptom category
based on retrospective chart review. The original ver- and total scores can be calculated, although measur-
sion of the scale rated items on severity only, but ing change using the latter may obscure changes on
the scale has been revised to become the BEHAVE- individual categories. Data have been published on
AD-Frequency Weighted (BEHAVE-AD-FW) version scores across different levels of dementia severity and
where the items are rated on both severity and fre- MMSE scores,54 which would assist clinicians in in-
quency. Although originally validated in community- terpreting scores, although there are no published
dwelling people with dementia, the scale has also normative data.
been extensively used in nursing home residents with The BEHAVE-AD is suitable for use in people with
dementia.3,52 It has also been validated for adminis- AD of any severity living in the community or resi-
tration by telephone.53 dential care and may also be suitable for people with
The BEHAVE-AD and BEHAVE-AD-FW, overall, vascular dementia, dementia with Lewy bodies, and
have good psychometric properties. However, it is frontotemporal dementia.64 However, the scale may
noted that no information was located in the research not adequately measure behavioral changes observed
literature concerning the internal consistency and in frontotemporal dementia such as apathy, disinhibi-
test-retest reliability rates. Interrater reliability was tion, and emotional inappropriateness.
established by the instrument developers at >0.70
for all subscales except anxiety and phobias on the
Consortium to Establish a Registry for
BEHAVE-AD54 and >0.75 for all subscales except di- Alzheimer’s Disease-Behavior Rating Scale for
urnal rhythm disturbance on the BEHAVE-AD- FW.51
Dementia
The scale has good content validity. Factor analyses
have revealed differing numbers of factors, which The CERAD-BRSD23–25 is a standardized instru-
may be because of differences in the sample type ment designed to measure behavioral abnormalities
or size or instability in the internal structure of the in people with dementia or cognitive impairment. It
scale.55 Ito et al.15 used the BEHAVE-AD-FW to assess was developed by a team of experts in the field and is
40 patients with AD and found support for a simple one of the assessment instruments that make up the
separation between behavioral symptoms and psy- CERAD battery. A 17-item abbreviated version is also
chological symptoms where only behavioral symp- available, and there has been some demand for this
toms were found to be related to cognitive perfor- version, but limited evidence on this version is cur-
mance, specifically the attention domain. rently available.
Scores on the scale show a curvilinear relationship The CERAD-BRSD has been used in both clini-
with severity, increasing with dementia severity and cal and research settings and in numerous clinical
the MMSE up to moderately severe dementia and and intervention studies. It has been used as an

Am J Geriatr Psychiatry 19:5, May 2011 409

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Assessment of BDAD

outcome measure in studies evaluating the effective- terms of correlations with other well-known instru-
ness of drug treatment65,66 and in a study assessing an ments measuring agitation and/or aggression is also
activities-based adult dementia care program.67 The available.74,77
instrument has also been used in studies investigat- Discriminant validity of the CERAD-BRSD has
ing comorbidity in community-dwelling people with been confirmed through several studies. The in-
AD,68 predicting psychosis onset,69 investigating strument has been shown to discriminate between
subtypes of psychosis,70 and in a longitudinal study different levels of dementia severity and between de-
examining the effects of depressive symptoms in mented and nondemented people.23,24,74,75
people with AD on the incidence and levels of Evidence relating to responsiveness is limited and
depression in family caregivers.71 Other clinical stud- mixed. In a study evaluating the effectiveness of a
ies include investigations of the relationship between weekly activity-based program, scores were sensi-
nursing home placement and measures of change,72 tive to the intervention, changing significantly, but
and a pilot study of a potential new outcome, emer- not in the expected direction, that is, behavior did
gent symptomatology.73 not improve as a result of the intervention.67 There
Evidence from numerous studies indicates that the is moderate evidence for sensitivity over time with
CERAD-BRSD has good to excellent psychometric one study24 reporting significant change over time
properties. Most studies have provided considerable but only for people with mild to moderate dementia.
evidence to ensure the findings can be appropriately
interpreted. The internal structure of the instrument
has generally been confirmed through factor analysis
DISCUSSION AND CONCLUSION
supporting the domains proposed by the authors,23
and interitem consistency for the subscales was re- Table 3 indicates that the NPI and the BEHAVE-
ported to range from 0.48 to 0.80. AD were the most highly rated measures for
Reliability of the instrument has been confirmed assessment of global BDAD, followed by the
through a number of studies.23,24,74 Test-retest re- CERAD-BRSD, the NRS, and the DBDS. All the five
liability has been shown to be good with Pear- instruments are proxy rated by interviewing clini-
son correlations and intraclass correlation coefficients cians/carers/informants who are deemed to have the
of 0.70–0.89, and interrater reliability excellent with best knowledge about the behavioral and psycholog-
kappa ranging from 0.77 to 1.00. Internal consis- ical conditions of the person with dementia in every-
tency for the total scale and for depressive symptoms, day life.
irritability/aggression, and psychotic symptoms sub- The NPI and the BEHAVE-AD are two of the most
scales are reported to be very good with Cronbach’s a popular, widely utilized instruments to assess the
of 0.87, 0.77, 0.75, and 0.80. For the inertia, vegetative presence and severity of BDAD in people with de-
symptoms, and behavioral dysregulation subscales, mentia. They have proven to have good psycho-
however, Cronbach’s a was only 0.48, 0.56, and 0.51, metric properties, are sensitive to pharmacological
respectively. and nonpharmacological interventions, and are ap-
Evidence for construct validity comes from the plicable to various institutional, outpatient, and com-
extent to which scores on the CERAD-BRSD relate munity settings. Both have been translated in nu-
to other measures in a manner that is consistent merous languages and shown to be reliable and
with theoretically derived hypotheses concerning the valid measures of neuropsychiatric disturbances in
domains measured. Studies have reported expected persons with dementia from non-English speaking
correlations with several measures of physical and countries. The NPI takes a shorter time to complete
cognitive functioning, including the Functional As- (10–15 minutes) than the BEHAVE-AD (20 minutes)
sessment Staging, Activities of Daily although both can be completed in a shorter time
Living (Functional status), and the RMBPC.75 when self-administered by carers/informants. Both
Scores have also been shown to be associated with in- tools assess the impact of BDAD on carers, which
dicators of depression.76 Expected correlations were can be rated separately from the symptom severity
found with the CMAI and Agitated Behavior in measure of the tool. The NPI comes with a training
Dementia scale.77 Evidence of construct validity in manual, whereas the BEHAVE-AD does not, which

410 Am J Geriatr Psychiatry 19:5, May 2011

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
 Jeon et al.

makes it unsuitable for the completion by people Notably, all raters need to be systematically trained
without some clinical training given the use of psy- to use these instruments in a consistent manner.
chiatric language in the tool. The NPI was originally Having good knowledge of the person is particu-
developed to assess psychopathology in people with larly important when determining the presence and
dementia and to help distinguish between different severity of the behavior(s) occurring. Lyketsos78 and
causes of dementia. It includes items pertaining to Rosenberg et al.79 argue that NPI ratings are most
symptoms known to be rare in AD but characteris- vulnerable to the effect of caregiver variables when
tic of frontotemporal dementias. On the other hand, there is a reliance on caregiver assessment with-
the BEHAVE-AD was developed to measure change out reference to the clinical judgment of experi-
in behavioral disturbance in people with AD and may enced clinicians and the person with dementia. These
not adequately measure behavioral changes observed researchers/clinicians recommend developing a re-
in frontotemporal dementia such as apathy, disinhibi- vised NPI to include these additional inputs in the
tion, and emotional inappropriateness. assessment. Connor et al.80 also discussed evidence
The CERAD-BRSD was designed to measure for variability in how raters are being trained to ad-
behavioral abnormalities in demented or cogni- minister and score the NPI in clinical trials. For an
tively impaired people. Similar to the NPI and the overview of this issue and recommendations see the
BEHAVE-AD, the CERAD-BRSD has been used in study by Connor et al.80 Because of the psychiatric
numerous studies in both clinical and research set- language in the BEHAVE-AD, it is unsuitable for
tings, and its psychometric properties are reported completion by people without appropriate training.
to be good to excellent. The main reasons for lower A trained interviewer is also required to adminis-
scores, according to the DOMS weighting criteria ter the CERAD-BRSD, because inconsistent admin-
shown in Table 2, relate to the cost and longer ad- istration could have adverse effects on psychometric
ministration time. The shorter version of the CERAD- properties and the ability to accurately measure treat-
BRSD may better address administrative require- ment effects.80
ments, but currently, there is little literature available The DOMS report16 provides recommendations
on the psychometric properties of this instrument for concerning the use of proxies, which should be con-
it to be recommended in this review. sidered when using these scales:
Two recent reviews of clinical trials of 1. Proxy reports should be examined for three
cholinesterase inhibitors suggest that although potential biases: 1) the cognitive status of the
the effects of such medications show modest clinical proxy (as many elderly people are cared for by
benefits in reducing BDAD measured by NPI or an elderly spouse carer, who may themselves
BEHAVE-AD, these global measures have limitations be impaired or unwell but to a lesser degree); 2)
in capturing a true picture of the drug effects on the health status of the proxy; and 3) the level of
BDAD as they are a complex collection of signs carer burden and stress.81
and symptoms that vary between patients in terms 2. There is usually a trade-off between those “with
of presentation and pattern.43,44 It is thought that the greatest amount of contact and those with
BDAD may be measured more accurately when more training” (81, p 488). However, generally,
symptoms are grouped into clusters that can provide where a proxy report is used, information
a framework for assessment and management.5,6 should be collected from the family mem-
Despite the widespread, international use of the ber/carer or care staff member who is closest
NPI and NPI-NH, Lange et al.47 recommend scoring to the patient and has the greatest degree of
and interpreting the individual items or factors as interaction with the patient.
opposed to total scores when using the NPI-NH with 3. Proxy reports should be based on usual behavior
a heterogeneous population. Separate symptom cate- rather than extreme or rare behaviors.81
gory scores (as opposed to a total score) are available 4. Given the noted discrepancy between patient
for the BEHAVE-AD22 and BEHAVE-AD-FW.51 The and proxy reports, where it is possible and fea-
CERAD-BRSD contains six separate domains that sible, subjective phenomena should be assessed
can be measured.74 by patient self-report rather than by proxy

Am J Geriatr Psychiatry 19:5, May 2011 411

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Assessment of BDAD

report. Proxy reports can be best used when as- measures can often be associated with underlying
sessing observable phenomena such as physi- mental conditions of delusions, hallucinations, de-
cal symptoms and functioning; however, they pression, and/or anxiety. It may be more appropriate
should be used with a great caution when as- and meaningful to measure specific symptom groups
sessing subjective phenomena such as depres- such as affect, apathy, and psychosis rather than mea-
sion, social isolation, and quality of life.82 suring global scores of BDAD.84 More research is
needed to confirm this classification method.
Obtaining an accurate picture of BDAD using It is beyond the scope of this review to provide an
proxy reports is challenging, and some may argue for in-depth exploration of diverse definitions of each in-
the use of direct observation instead. However, ob- dividual BDAD or an analysis of diverse methods of
taining good direct observations is time and resource grouping; however, it is important to acknowledge
intensive. Furthermore, if the observer is not famil- consensus on the groupings of BDAD has yet to be
iar with the person being observed, they will have no established, and there may need to be new develop-
insight into the person’s variable responses to their ment of better outcome measures to assess BDAD if
environment nor of the unique context within which different approaches of grouping are introduced. The
the person is living and receiving care. In this circum- reviews of BDAD global in this article are based on
stance, a combination of direct clinical observation currently available instruments.
and proxy reports would render best representation Cummings85 appropriately argues that choice of
of the individual’s BDAD. This is an issue that war- tools depends largely on the type(s) of behaviors to
rants further research. be measured while considering various characteris-
No diagnostic nomenclature has provided clear de- tics of instruments available. He states both BDAD
scriptions or information on the severity, course, or global measures and measures focusing on specific
types of BDAD. A major advance was made when aspects of BDAD are useful in different ways, in that
Diagnostic and Statistical Manual of Mental Disorders, global measures provide information on the overall
4th version, was released in which the “specifier” condition of the person with dementia, whereas spec-
phrase “with behavioral disturbance” was added to ified measures can provide more explicit information
the 4th revision; but without any guidance how one about particular attributes of BDAD that clinicians or
might interpret “behavioral disturbance,”83 and as- researchers aim to evaluate. The choice of the tools
sess its types, causes, prognoses, and outcomes ap- also depends on the best available sources of infor-
propriately. This poses a difficult situation when val- mation.
idating an instrument for BDAD, as there is no gold Despite shortcomings of the BDAD global mea-
standard to measure against. sures such as training issues, the reliability and valid-
Although most symptoms such as hallucinations, ity of all five instruments have been established, and
delusions, depression, apathy, and anxiety denote they have demonstrated applicability for use in the
common definitions across the literature, there seem community, healthcare settings, research, and a num-
mixed methods of categorizing BDAD and defin- ber of different cultures. However, it is recommended
ing some of the individual symptoms. For example, the NPI and the BEHAVE-AD be used in both clinical
Cohen-Mansfield and Billing (2, p 10) define agita- and research settings for assessment of global assess-
tion as “inappropriate verbal, vocal or motor activ- ment of BDAD. The CERAD-BRSD is recommended
ity that is not judged by an outside observer to result for research, but its limitations are its cost and time
directly from the needs or confusion of the person,” required for administration. The 17- item abbrevi-
and propose four types of agitation including phys- ated version may be more appropriate for clinical
ically and verbally nonaggressive behaviors, and settings, but limited evidence on this version is cur-
physically and verbally aggressive behaviors, which rently available. If specific attributes of BDAD need
clearly assign aggression under agitation. Lyketsos78 to be assessed, separate symptom categories of the
argues that classifying BDAD based on observa- BDAD global measures or other specified measures
tional information with a focus on explicit behav- for individual attributes of BDAD should be consid-
iors may not provide accurate assessment of BDAD. ered instead of relying on the total scores of the BPSD
For example, agitation defined in Cohen-Mansfield’s global measures. The DOMS was designed to provide

412 Am J Geriatr Psychiatry 19:5, May 2011

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
 Jeon et al.

a useful guide to address those particularly in the and policy/decision makers; those personnel most
Australian context; however, the findings indicate its impacted by BDAD assessment. We have considered
applicability in other English-speaking countries. the utility, feasibility, costs, accessibility, cultural di-
In this regard, the DOMS review provides rec- versity, and psychometrics of each tool selected, so
ommendations for the best available measures/tools that clinicians and researchers can make informed de-
suitable for routine clinical and research use, cisions for their routine use.
which are critical for epidemiological studies, early
intervention and treatment of dementia conditions, The DOMS project was supported by the Australian
and funding allocations for relevant healthcare ser- Government Department of Health and Ageing. This
vices. Unlike other reviews of BDAD instruments that article is based on Chapter 9 of the Final Report:
focus solely on psychometric properties of the tools, Dementia Outcomes Measurement Suite Project (Sansoni
we have adopted a rigorously designed, six-staged et al, 2007), an Australian Government Initiative.
process of identifying, selecting, and critiquing sci- The views expressed in this work are those of its au-
entific evidence. Each instrument selected for com- thors and are not necessarily those of the Commonwealth of
prehensive review is evaluated against explicit and Australia. The reader needs to be aware that the informa-
standardized criteria. A novel component of the re- tion contained in this work is not necessarily endorsed,
view included consultations with key Australian ex- and its contents may not have been approved or reviewed,
perts in the field consisting of clinicians, consumers, by the Australian Government Department of Health and
researchers, biostatisticians, health service providers, Ageing.

References
1. Lyketsos CG, Steinberg M, Tschanz JT, et al: Mental and behavioral 11. O’Rourke N, Bedard M, Bachner YG: Measurement and analysis
disturbances in dementia: findings from the Cache County study of behavioral disturbance among community-dwelling and insti-
on memory in aging. Am J Psychiatry 2000; 157:708–714 tutionalized persons with dementia. Aging Mental Health 2007;
2. International Psychogeriatric Association: Behavioral and Psycho- 11:256–265
logical Symptoms of Dementia, 2003. Available at: http://www. 12. Mendez MF, Lauterbach EC, Sampson SM: An evidence-based re-
ipa-online.net/pdfs/bpsd.asp. Accessed May 1, 2009 view of the psychopathology of frontotemporal dementia: a re-
3. Brodaty H, Draper B, Saab D, et al: Psychosis, depression and be- port of the ANPA Committee on Research. J Neuropsychiatry Clin
havioural disturbances in Sydney nursing home residents: preva- Neurosci 2008; 20:130–149
lence and predictors. Int J Geriatr Psychiatry 2001; 16: 504–512 13. Lovheim H, Sandman P-O, Karlsson S, et al: Behavioral and psy-
4. Saz P, Lopez-Anton R, Dewey ME, et al: Prevalence and implica- chological symptoms of dementia in relation to level of cognitive
tions of psychopathological non-cognitive symptoms in dementia. impairment. Int Psychogeriatr 2008; 20:777–789
Acta Psychiatr Scand 2009; 119:107–116 14. Starr JM, Lonie J: Relationship between behavioural and psycho-
5. Robert PH, Verhey FRJ, Byrne EJ, et al: Grouping for behavioral logical symptoms of dementia and cognition in Alzheimer’s dis-
and psychological symptoms in dementia: clinical and biologi- ease. Dement Geriatr Cogn Disord 2007; 24:343–347
cal aspects. Consensus paper of the European Alzheimer disease 15. Ito T, Meguro K, Akanuma K, et al: Behavioral and psychological
consortium. Eur Psychiatry 2005; 20:490–496 symptoms assessed with the BEHAVE-AD-FW are differentially
6. Aalten P, Verhey FRJ, Boziki M, et al: Neuropsychiatry syn- associated with cognitive dysfunction in Alzheimer’s disease. J
dromes in dementia: results from the European Alzheimer Disease Clin Neurosci 2007; 14:850–855
Consortium: part I. Dement Geriatr Cogn Disord 2007; 24: 16. Sansoni J, Marosszeky N, Jeon Y-H, et al: Final Report: Demen-
457–463 tia Outcomes Measurement Suite Project. Centre for Health Ser-
7. Waldemar G, Dubois B, Emre M, et al: Recommendations for vice Development, University of Wollongong, Wollongong, NSW,
the diagnosis and management of Alzheimer’s disease and other Australia; 2007
disorders associated with dementia: EFNS guideline. Eur J Neurol 17. Bowling A: Measuring Health: A Review of Quality of Life
2007; 14:e1-e26 Measurement Scales. 3rd ed. Berkshire, Open University Press,
8. Caputo M, Monastero R, Mariani E, et al: Neuropsychiatry symp- 2005
toms in 921 elderly subjects with dementia: a comparison be- 18. McDowell I: Measuring Health: A Guide to Rating Scales and
tween vascular and neurodegenerative types. Acta Psychiatr Questionnaires. 3rd ed. Oxford, England, Oxford University Press,
Scand 2008; 117:455–464 2006
9. Kverno KS, Rabins PV, Blass DM, et al: Prevalence and treatment 19. Burns A, Lawlor B, Craig S: Assessment Scales in Old Age Psychi-
of neuropsychiatric symptoms in advanced dementia. J Gerontol atry. 2nd ed. London, Taylor & Francis, 2004
Nurs 2008; 34:8–15 20. Lezak MD: Neuropsychological Assessment. 4th ed. New York,
10. Neil W, Bowie P: Carer burden in dementia—assessing the im- Oxford University Press, 2004
pact of behavioural and psychological symptoms via self-report 21. McKeith I, Cummings J, Lovestone S, et al: Outcome Measures in
questionnaire. Int J Geriatr Psychiatry 2008; 23:60–64 Alzheimer’s Disease. London, Martin Dunitz, 1999

Am J Geriatr Psychiatry 19:5, May 2011 413

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Assessment of BDAD

22. Reisberg B, Borenstein J, Salob SP, et al: Behavioral symptoms in receiving donepezil: a randomized controlled trial. JAMA 2004;
Alzheimer’s disease: phenomenology and treatment. J Clin Psy- 291:317–324
chiatry 1987; 48(suppl 1):9–15 41. Holmes C, Wilkinson D, Dean C, et al: The efficacy of donepezil
23. Mack JL, Patterson MB, Tariot PN: Behavior Rating Scale for in the treatment of neuropsychiatric symptoms in Alzheimer dis-
Dementia: development of test scales and presentation of data ease. Neurology 2004; 63:214–219
for 555 individuals with Alzheimer’s disease. J Geriatr Psychiatry 42. Doody RS, Gavrilova SI, Sano M, et al: Effect of dimebon on cogni-
Neurol 1999; 12:211–223 tion, activities of daily living, behaviour, and global function in pa-
24. Patterson MB, Mack JL, Mackell JA, et al: A longitudinal study tients with mild-to-moderate Alzheimer’s disease: a randomised,
of behavioral pathology across five levels of dementia severity double-blind, placebo-controlled study. Lancet 2008; 372:
in Alzheimer’s disease: the CERAD Behavior Rating Scale for De- 207–215
mentia. The Alzheimer’s Disease Cooperative Study. Alzheimer 43. Rodda J, Morgan S, Walker Z: Are cholinesterase inhibitors effec-
Dis Assoc Disord 1997; 11(suppl 2):S40-S44 tive in the management of the behavioral and psychological symp-
25. Tariot PN: CERAD Behavior Rating Scale for dementia. Int Psy- toms of dementia in Alzheimer’s disease? A systematic review of
chogeriatr 1996; 8(suppl 3):514–515 randomized, placebo-controlled trials of donepezil, rivastigmine
26. Devanand DP, Miller L, Richards M, et al: Columbia University and galantamine. Int Psychogeriatr 2009; 21:813–824
Scale for psychopathology in Alzheimer’s disease. Arch Neurol 44. Winblad B, Black SE, Homma A, et al: Donepezil treatment in
1992; 49:371–376 severe Alzheimer’s disease: a pooled analysis of three clinical
27. Baumgarten M, Becker R, Gauthier S: Validity and reliability of the trials. Curr Med Res Opin 2009; 25:2577–2587
Dementia Behavior Disturbance Scale. J Am Geriatr Assoc 1990; 45. Cummings JL: The Neuropsychiatry Inventory: assessing psycho-
38:221–226 pathology in dementia patients. Neurology 1997; 48(suppl 6):
28. Allen NH, Gordon S, Hope T, et al: Manchester and Oxford Univer- S10-S16
sities Scale for the Psychopathological Assessment of Dementia 46. Politis AM, Mayer LS, Passa M, et al: Validity and reliability of
(MOUSEPAD). Br J Psychiatry 1996; 169:293–307 the newly translated Hellenic Neuropsychiatric Inventory (H-NPI)
29. Levin HS, High WM, Goethe KE, et al: The neurobehavioural applied to Greek outpatients with Alzheimer’s Disease: a study of
rating scale: assessment of the behavioural sequelae of head in- the disturbing behaviours among referrals to a memory clinic. Int
jury by the clinician. J Neurol Neurosurg Psychiatry 1987; 50: J Geriatr Psychiatry 2004; 19:203–208
183–193 47. Lange RT, Hopp GA, Kang N: Psychometric properties and fac-
30. Ray WA, Taylor JA, Lichtenstein MJ, et al: The Nursing Home tor structure of the Neuropsychiatric Inventory Nursing Home
Behavior Problem Scale. J Gerontol 1992; 47:9–16 version in an elderly neuropsychiatric population. Int J Geriatr
31. Cummings JL, Mega M, Gray K, et al: The Neuropsychiatry Psychiatry 2004; 19:440–448
Inventory: comprehensive assessment of psychopathology in de- 48. Ikeda M, Fukuhara R, Shigenobu K, et al: Dementia associated
mentia. Neurology 1994; 44:2308–2314 mental and behavioural disturbances in elderly people in the
32. Teri L, Truax P, Logsdon R, et al: Assessment of behavioral prob- community: findings from the first Nakayama study. J Neurol
lems in dementia: the Revised Memory and Behavior Problems Neurosurg Psychiatry 2004; 75:146–148
Checklist. Psychol Aging 1992; 7:622–631 49. Wood S, Cummings JL, Hsu MA, et al: The use of the neu-
33. Selbaek G, Kirkevold O, Engedal K: Psychiatric and behavioural ropsychiatric inventory in nursing home residents. Character-
symptoms and the use of psychotropic medication in Special ization and measurement. Am J Geriatr Psychiatry 2000; 8:
Care Units and Regular Units in Norwegian nursing homes. Scand 75–83
J Caring Sci 2008; 22:568–573 50. Cohen-Mansfield J, Marx M, Rosenthal A: A description of agita-
34. Aalten P, Verhey FRJ, Boziki M, et al: Consistency of neuropsy- tion in a nursing home. J Gerontol 1989; 44:77–84
chiatric syndromes across dementias: results from the European 51. Monteiro IM, Boksay I, Auer SR, et al: Addition of a frequency-
Alzheimer Disease Consortium. Part II. Dement Geriatr Cogn Dis- weighted score to the Behavioral Pathology in Alzheimer’s Dis-
ord 2008; 25:1–8 ease Rating Scale: the BEHAVE-AD-FW: methodology and reliabil-
35. Levy MH, Miller BL, Cummings JL, et al: Alzheimer’s disease and ity. Eur Psychiatry 2001; 16(suppl 1):5s-24s
frontotemporal dementias: behavioural distinctions. Arch Neurol 52. De Deyn PP, Katz IR, Brodaty H, et al: Management of agitation,
1996; 53:687–690 aggression, and psychosis associated with dementia: a pooled
36. Craig AH, Cummings JL, Fairbanks L, et al: Cerebral blood flow analysis including three randomized, placebo-controlled double-
correlates of apathy in Alzheimer disease. Arch Neurol 1996; blind trials in nursing home residents treated with risperidone.
53:1116–1120 Clin Neurol Neurosurg 2005; 107:497–508
37. White HK, McConnell ES, Bales CW, et al: The 6th month study of 53. Monteiro IM, Boksay I, Auer SR, et al: Reliability of routine clin-
the relationship between weight loss and behavioural symptoms ical instruments for the assessment of Alzheimer’s disease ad-
in institutionalised Alzheimer’s disease subjects. J Am Med Dir ministered by telephone. J Geriatr Psychiatry Neurol 1998; 11:
Assoc 2004; 5:89–97 18–24
38. Cummings JL, Street J, Masterman D, et al: Efficacy of olanzipine 54. Sclan SG, Saillon A, Franssen E, et al: The Behavior Pathology
in the treatment of psychosis in dementia with Lewy bodies. in Alzheimer’s Disease Rating Scale (BEHAVE-AD): reliability and
Dement Geriatr Cogn Disord 2002; 13:67–73 analysis of symptom category scores. Int J Geriatr Psychiatry 1996;
39. Hatoum HT, Lin S-J, Arcona S, et al: The use of the occupational 11:819–830
disruptiveness scale of the neuropsychiatric inventory-nursing 55. Schreinzer D, Ballaban T, Brannath W, et al: Components of be-
home version to measure the impact of rivastigmine on the disrup- havioral pathology in dementia. Int J Geriatr Psychiatry 2005;
tive behavior of nursing home residents with Alzheimer’s disease. 20:137–145
J Am Med Dir Assoc 2005; 6:238–245 56. Liu CY, Wang PN, Lin KN, et al: Behavioral and psychological
40. Tariot PN, Farlow MR, Grossberg GT, et al: Memantine treatment symptoms in Taiwanese patients with Alzheimer’s disease. Int
in patients with moderate to severe alzheimer disease already Psychogeriatr 2007; 19:605–613

414 Am J Geriatr Psychiatry 19:5, May 2011

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
 Jeon et al.

57. Cummings JL, McRae T, Zhang R, Donepezil-Sertraline Study 72. Knopman DS, Berg JD, Thomas R, et al: Nursing home placement
Group: Effects of donepezil on neuropsychiatric symptoms in is related to dementia progression: experience from a clinical
patients with dementia and severe behavioral disorders. Am J trial. Alzheimer’s Disease Cooperative Study. Neurology 1999;
Geriatr Psychiatry 2006; 14:605–612 52:714–718
58. Finkel SI, Lyons J, Anderson R: Reliability and validity of the 73. Tractenberg RE, Gamst A, Thomas RG, et al: Investigating emer-
Cohen-Mansfield Agitation Inventory in institutionalized elderly. gent symptomatology as an outcome measure in a behavioral
Int J Geriatr Psychiatry 1992; 7:486–490 study of Alzheimer’s disease. J Neuropsychiatry Clin Neurosci
59. Eustace A, Bruce I, Coen R, et al: Behavioural disturbance trig- 2002; 14:303–310
gers recognition of dementia by family informants. Int J Geriatr 74. Tariot PN, Mack JL, Patterson MB, et al: The Behav-
Psychiatry 2007; 22:574–579 ior Rating Scale for Dementia of the Consortium to Es-
60. Cheng T-W, Chen T-F, Yip P-K, et al: Comparison of behavioral tablish a Registry for Alzheimer’s Disease. The Behavioral
and psychological symptoms of Alzheimer’s disease among insti- Pathology Committee of the Consortium to Establish a Reg-
tution residents and memory clinic outpatients. Int Psychogeriatr istry for Alzheimer’s Disease. Am J Psychiatry 1995; 152:
2009; 21:1134–1141 1349–1357
61. Brodaty H, Draper BM, Millar J, et al: Randomized controlled trial 75. Tractenberg RE, Weiner MF, Patterson MB, et al: Emergent psy-
of different models of care for nursing home residents with de- chopathology in Alzheimer’s disease patients over 12 months
mentia complicated by depression or psychosis. J Clin Psychiatry associated with functional, not cognitive, changes. J Geriatr
2003; 64:63–72 Psychiatry Neurol 2002; 15:110–117
62. Tangwongchai S, Thavichachart N, Senanarong V, et al: 76. Jacobs MR, Strauss ME, Patterson MB, et al: Characterization of
Galantamine for the treatment of BPSD in Thai patients with possi- depression in Alzheimer’s disease by the CERAD Behavior Rat-
ble Alzheimer’s disease with or without cerebrovascular disease. ing Scale for Dementia (BRSD). Am J Geriatr Psychiatry 1998; 6:
Am J Alzheimer’s Dis Other Demen 2009; 23:593–601 53–58
63. Rabinowitz J, Katz I, De Deyn PP, et al: Treating behavioral 77. Weiner MF, Tractenberg R, Teri L, et al: Quantifying behavioral
and psychological symptoms in patients with psychosis of disturbance in Alzheimer’s disease patients. J Psychiatr Res 2000;
Alzheimer’s disease using risperidone. Int Psychogeriatr 2007; 19: 34:163–167
227240 78. Lyketsos CG: Neuropsychiatric symptoms (behavioral and
64. Chiu MJ, Chen TF, Yip PK, et al: Behavioral and psychologic psychological symptoms of dementia) and the develop-
symptoms in different types of dementia. J Formos Med Assoc ment of dementia treatments. Int Psychogeriatr 2007; 19:
2006; 105:556–562 409–420
65. Teri L, Logsdon RG, Peskind E, et al: Treatment of agitation in AD: 79. Rosenberg PR, Meikle MM, Lyketsos CG: Caregivger assessment of
a randomized, placebo-controlled clinical trial. Neurology 2000; patient’s depression in Alzheimer’s disease: longitudinal analysis
55:1271–1278 in a drug treatment study. Am J Geriatr Psychiatry 2005; 13:822–
66. Weiner MF, Martin-Cook K, Foster BM, et al: Effects of donepezil 826
on emotional/behavioral symptoms in Alzheimer’s disease pa- 80. Connor DJ, Sabbagh MN, Cummings JL: Comment on
tients. J Clin Psychiatry 2000; 61:487–492 administration and scoring of the Neuropsychiatric In-
67. Higgins M, Koch K, Hynan LS, et al: Impact of an activities- ventory in clinical trials. Alzheimers Dement 2008; 4:
based adult dementia care program. Neuropsychiatr Dis Treat- 390–394
ment 2005; 1:165–169 81. Harper G: Assessing older adults who cannot communicate,
68. Tractenberg RE, Weiner MF, Patterson MB, et al: Comorbidity in Assessing Older Persons: Measures, Meaning, and Practical
of psychopathological domains in community-dwelling persons Applications. Edited by Kane R, Kane R. Oxford, England,
with Alzheimer’s disease. J Geriatr Psychiatry Neurol 2003; 16: Oxford University Press; 2000. Ch. 17, pp 483–515
94–99 82. Snow A, Cook K, Lin P-S, et al: Proxies and other external
69. Wilkosz PA, Miyahara S, Lopez OL, et al: Prediction of psychosis raters: Methodological considerations. Health Serv Res 2005;
onset in Alzheimer disease: the role of cognitive impairment, de- 40:16761693
pressive symptoms, and further evidence for psychosis subtypes. 83. Caine E: Behavioural disturbances of dementia in our current
Am J Geriatr Psychiatry 2006; 14:352–360 nomenclature system: diagnostic classification of neuropsychiat-
70. Perez-Madrinan G, Cook SE, Saxton JA, et al: Alzheimer dis- ric signs and symptoms in patients with dementia. Int Psychoge-
ease with psychosis: excess cognitive impairment is restricted riatr 1996; 8(suppl 3):273–279
to the misidentification subtype. Am J Geriatr Psychiatry 2004; 84. Lyketsos C, Breitner J, Rabins P: An evidence based pro-
12:449456 posal for the classification of neuropsychiatric disturbance
71. Neundorfer MM, McClendon MJ, Smyth KA, et al: A longitudinal in Alzheimer’s disease. Int J Geriatr Psychiatry 2001; 16:
study of the relationship between levels of depression among 1037–1042
persons with Alzheimer’s disease and levels of depression among 85. Cummings J: Theories behind scales and measurements: theo-
their family caregivers. J Gerontol Seri B Psychol Sci Soc Sci 2001; ries behind existing scales for rating behaviour in dementia. Int
56B:301–313 Psychogeriatr 1996; 8(suppl 3): 293–300

Am J Geriatr Psychiatry 19:5, May 2011 415

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.