Occup. Mod. Vol. 47, No. 1, pp.

25-32,1997
Copyright O 1997 Rapid Science Pubishars for SOM
Printed In Great Britain. Allrightsreserved
0982-748OCB7

Is ergonomic intervention
alone sufficient to limit
musculoskeletal problems in
nurses?

Downloaded from https://academic.oup.com/occmed/article-abstract/47/1/25/1414172 by guest on 21 September 2019

A. K. Burton,* T. L. Symonds,* E. Zinzen,*
K. M. Tillotson,* D. Caboor,* P. Van Royf and J. P. Clarys1
*Spinal Research Unit, The University of Huddersfield, Huddersfield,
HD1 3DH, UK; ^Department of Exp. Anatomy, Vrije Universiteit
Brussel, Laarbeeklaan 103, 1090 Brussels, Belgium

This study retrospectively surveyed 1,216 nurses at hospitals in Belgium and The
Netherlands. Data concerning workloads, musculoskeletal symptoms, work loss and
psychosocial factors were collected by questionnaire. Lifetime prevalence rates for
musculoskeletal problems and low back trouble were significantly lower in the Dutch
hospitals than the Belgian hospitals, but a significantly higher proportion of Dutch
nurses had 'heavy' workloads. Overall, symptoms and work loss in the previous 12
months were not related to workload, nor was the perception that work was causative;
a change of duties because of symptoms was rare (< 3%). The Dutch nurses differed
strikingly from Belgian nurses on the psychosocial variables; they were less depressed
and significantly more positive about pain, work and activity. It is proposed that
ergonomic interventions alone may be sub-optimal in controlling musculoskeletal
problems among nurses The additional provision of psychosocoial information to
challenge misconceptions and encourage self-management is proposed.

Keywords: Attitudes; back pain; beliefs; coping strategies; ergonomics; musculoskeletal disorders;
nursing; occupation; work.

Occup. Mod. VoL 47, 25-32, 1997

Rtawtd 13 November 1995; accepud in final Jbrm 27 Stpumber 1996.

INTRODUCTION tively, with an annual prevalence of 36% and 38%

respectively.7 It has been estimated that approximately
Musculoskeletal problems in general, and back pain 40,000 nurses take a period of sick leave per year due
in particular, are major problems in industrialized to back pain in Britain, which amounts to 750,000
society and are associated with significant costs;1 working days lost annually.4 The health services in
Leamon2 has stated that low back trouble (LBT) is England and Wales received 1,170 permanent injury
often the most expensive work-related injury. Nursing benefit claims and paid out over £5.4 million, as well
has one of the highest prevalence rates for back pain as £1 million in temporary injury allowance during
in comparison to other professions.3"6 Stubbs and 1988/89.8 Venning et al.9 found, in a group of American
Buckle4 found in a retrospective study of 3,912 British nurses, that back injuries accounted for 40% of all
nurses that the annual prevalence was 43%, the point time-loss claims. Stappaerts10 reported a life-time
prevalence was 17% and the annual incidence was prevalence of low back problems among 1,775 nurses
7.7%. These rates are higher than found in the general in a large Belgian Catholic hospital of 66%. In terms
British population; point prevalence for males and of absence in the previous 12 months due to a spell
females has been reported as 12% and 15% respec- of back pain, 6% of the females and 5% of males took
a period of absence. Back pain is, then, not only highly
prevalent in the nursing profession but is also associ-
Correspondence and reprint requests tcr. Dr K. Burton, Spinal ated with considerable costs due to lost working time.
Research Unit, University of Huddersfield, c/o 30 Queen Street,
Huddersfield, HD1 2SP, UK. Tel: (+44) 1484 424329; Fax: (+44)1484
Many epidemiological studies have been carried out
435744; email: kburtonOcJx.compullnk.co.uk. in the nursing profession to identify risk factors asso-
26 Occup. Med. Vol. 47. 1997

ciated with back pain. The most consistently proposed MATERIALS AND METHODS
risk factor has been lifting,5''1" and much of the reported
back injuries in the British health service have been This study forms a discrete part of a larger multi-
attributed to manual handling.8 In all occupations in disciplinary study of Belgian nurses. In addition to the
Britain, 25% of workplace accidents are associated with parameters concerned here, the studies involved the
manual handling, but in the health service the proportion measurement of physical, biomechanical and anthropo-
is considerably higher with almost 50% of the accidents metric characteristics which will be reported elsewhere.
in 1985 reported as associated with manual handling.8 The study was retrospective and involved nurses
Ergonomic interventions have been suggested as a from hospitals in Belgium (two academic and two
means of improving the working conditions of nurses Catholic hospitals) and The Netherlands (three general
in order to prevent (or reduce) back injuries.4'12 hospitals); 1,783 nurses were given a booklet of
Ergonomics involves looking at how to best fit the questionnaires by the senior nurse in each department
employee to the work tasks required, by critically evalu- in 1992. The instructions stressed that the survey was

Downloaded from https://academic.oup.com/occmed/article-abstract/47/1/25/1414172 by guest on 21 September 2019

ating existing work practices and providing a rationale initiated by the government and was of general interest
for interventions to effect safer systems of work.12 If, to the nursing profession, but did not mention any
however, the working environment alone is considered special interest in back pain. These booklets contained
then there is a risk of neglecting the individual's a variety of questions related to musculoskeletal
psychosocial well-being. The importance of psycho- problems (via questions relating symptoms to general
social factors in back pain has been shown by Symonds anatomical areas, perceived causation and need for
et a/.13 who found that industrial workers with a history absence), as well as five psychosocial instruments
of LBT had significantly more negative psychometrics measuring different psychosocial dimensions related
{e.g. pain locus of control, inevitability beliefs and to activity, pain coping strategies, pain control, depres-
activity beliefs) than workers who did not have a sion, and work.
history of LBT. A recent intervention study, which The five instruments used were: (1) the Fear-Avoid-
used an educational pamphlet to encourage positive ance Beliefs Questionnaire (FABQ)17 which has two
attitudes and beliefs about pain, work and activity, subscales measuring beliefs about work activities and
achieved a significant reduction in sickness absence in beliefs about general physical activities related to back
an industrial environment.14 pain; (2) the Coping Strategies Questionnaire (CSQ)18
Few studies, though, have investigated the psycho- which has eight subscales measuring a number of
social component of musculoskeletal complaints and strategies for coping with pain such as, praying/ hoping,
back pain in nurses, which seems surprising when catastrophizing, or diverting attention; (3) the Pain
Berg et a/.,15 as early as 1976, identified that Swedish Locus of Control (PLC)19 which has two subscales
nurses perceived their job to be more stressful than which measure beliefs about pain control and about
did industrial workers and also stated that they received pain responsibility; (4) the Modified Zung Depression
less support from colleagues and bosses at work. Others16 Inventory (MZ)20 which is a measure of depression
have found that psychosocial factors (as opposed to and (5) the Psychosocial Aspects of Work (PAW)13
physical factors) were of primary importance when which measures three different work attitudes: job
differentiating nurses who had experienced a spell of satisfaction, mental stress and social support.
LBT for at least three consecutive weeks from nurses Because the FABQ, PLC, and CSQ instruments are
who had no history of LBT or who had experienced concerned specifically with back pain, they were
LBT for only three days. They measured various completed only by nurses with a history of low back
psychosocial parameters of which anxiety, neuroticism pain; all nurses answered the PAW and MZ instru-
and general health beliefs were most helpful. ments. When the booklets were completed they were
The evidence suggests that the nursing profession returned to the senior nurse of each department within
is particularly at risk of musculoskeletal problems, the seven collaborating hospitals, and then to a central
notably affecting the low back, and that there is a location for processing.
considerable loss of work time. The understandable Workload was categorized on the basis of the hospital
focus on ergonomic interventions may not be the department in which the nurses worked. Estimates of
optimal solution to the problem; the negative attitudes workload were made from the nurses' responses to
to work shown by nurses15'16 suggests that psychosocial questions concerning the amount of time (minutes per
factors may be important. day) spent on various nursing activities which
The aim of this study was to explore the relationship previously have been described in an ergonomic
between psychosocial parameters, workload and study21 and a job evaluation study.22 Three workload
musculoskeletal problems in nurses. The following categories were defined from the number of minutes
hypotheses were proposed: nurses in departments with per day spent on patient handling tasks. For the whole
a 'heavy* workload will have relatively high rates of sample, the average time per day spent on strenuous
musculoskeletal problems, back pain and absenteeism; tasks (turning, lifting, helping, etc.) was found to be
nurses with 'positive' psychometrics will have relatively 160 minutes (SD = 100); a calculation based on half
low rates of musculoskeletal problems, back pain and of the standard deviation was used to categorize levels
absenteeism. of workload. This resulted in a 'heavy' category for
 A K. Burton et a/.: Back and muscutoskeletal problems in nurses 27

which more than 210 minutes (average + % SD) was variance (ANOVA) including Duncan's multiple range
spent on the strenuous tasks, a 'moderate' category in test, and chi-square tests were used where appropriate.
which between 110-210 minutes were spent on strenuous The level of statistical significance was set at 5%.
tasks, and a 'light' category in which less than 110
minutes (average-1/^ SD) were spent on strenuous
tasks. Thus each nurse could be allocated to a workload RESULTS
category. In order to validate the resultant classifica-
tions, the nurse's perception of workload was compared Description of study population
with their hospital departments; it transpired, for
example, that orthopaedics and geriatrics fell into the One thousand two hundred and sixteen nurses (68.2%)
heavy category; paediatrics and gynaecology were returned the-booklet. The respective responses from
moderate; outpatients and administration were classed Catholic, Academic and Dutch hospitals were: 72.6%,
as light. This classification of departments was then 66.8% and 68.7%; non-responders were almost exclu-

Downloaded from https://academic.oup.com/occmed/article-abstract/47/1/25/1414172 by guest on 21 September 2019

independently compared with departmental workload sively confined to those who were off-duty at the time
as classified in a standard Belgian 'minimal nursing of questioning. The ages ranged from 19-59 years with
plan'. The classification from the present data was a mean of 32 (SD = 7.9) years. Across the three types
agreed to be representative of nursing workloads, and of hospital the nurses working in the Catholic hospitals
reflected departmental workloads previously reported.23 were on average significantly (p < 0.04), but not sub-
Absence taken, due to musculoskeletal trouble, in stantially, older (mean = 33.5; SD = 7.7 years) than
the previous 12 months, was requested on the the nurses working in either the Academic (mean =
questionnaire quantified in days; the nurses were also 32.1; SD = 7.5 years) or Dutch hospitals (mean =
asked to state if they had had to change duties because 31.9; SD = 9.2 years). The lowest proportion of male
of their trouble. The nurses' perception of work being nurses was in the Catholic hospitals where only 14%
the cause of their symptoms was noted both for recent were male, compared to 24% in the Academic hospitals
trouble (past 12 months) and for past trouble. and 22% in the Dutch hospital (p < 0.05).

Statistical analysis Musculoskeletal symptoms

The seven hospitals were reduced to three categories A history of any musculoskeletal symptoms was
(Academic, Catholic or Dutch) reflecting their social reported by 812 nurses (66.8%), whilst 648 (53.3%)
and administrative structures. The t-test, analysis of had experienced low back pain specifically — some

Table 1 a. Prevalence of a history of musculoskeletal symptoms, history of low back trouble, and the point-prevalence for low back
pain broken down by type of hospital and type of duty; figures given as percentage and number of hospital group

Musculoskeletal symptoms History of low back trouble Current back pain (point prevalence)
Yes No Yes No Yes No

% (n) % (n) % (n) % (n) % (n) % N

Hospital
Academic 67.9 (540) 32.1 (255) 54.3 (432) 45.7 (363) 15.5 (123) 84.5 (672)
Catholic 74.0 (159) 26.0 (56) 58.1 (125) 41.9 (90) 17.7 (38) 82.3 (177)
Dutch 54.9 (113) 45.1 (93) 44.2 (91) 55.8 (115) 15.5 (32) 84.5 (174)
X 2 ;P 18.66; -e 0.001 9.26; 0.009 0.66; ns*
Duties
Heavy 66.6 (353) 33.4 (177) 54.7 (290) 45.3 (240) 40.4 (76) 59.6 (112)
Moderate 66.4 (255) 33.6 (129) 51.6 (198) 48.4 (186) 49.3 (69) 50.7 (71)
Light 69.3 (194) 30.7 (86) 53.9 (151) 46.1 (151) 46.0 (46) 54.0 (54)
Z2:P 0.75; ns* 0.92;; ns* 2.65; ns*

Figures In boW denote significantly lower symptom reporting by the Dutch nurses
•NS - non-slgnlflcant

Table 1 b. Percentage and number of nurses at each type of hospital engaged in the three categories of duty a n d w o r k l o a d . A signifi-
cantly higher proportion of Dutch nurses worked in heavy departments

Academic Catholic Dutch

% % cX, (n)
(n) (n)
Heavy 35.4 (277) 55.2 (117) 6E1.3 (136)
Moderate 35.2 (276) 29.2 (62) 2Z1.1 (46).
Ught 29.4 (230) 15.6 (33) 8 .5 (17)
X2 = 89.8; p < 0.001
28 Occup. Med. VcJ 47. 1997

239 (36.9%) of whom stated they were experiencing a Considering the perceived cause of the complaints
current spell of LBT at the time of completing the at any time; for those with musculoskeletal problems
booklet. the trouble was attributed to a 'work incident' by a
The nurses in the Dutch hospital had a significantly little over 50%; these proportions did not differ sig-
lower prevalence of a history both of musculoskeletal nificantly between the hospitals or between workload
symptoms and low back symptoms compared with the categories (Table 2). When looking at low back trouble
Belgian Academic and Catholic hospitals, but the point alone, the figures were virtually identical. For work
prevalence of LBT symptoms did not significantly incidents occurring just in the last 12 months, the
differ across the three types of hospital (Table la). perception of work being a cause was somewhat less
When looking at the occurrence of symptoms across (42%). The nurses in the Catholic hospitals were more
workload categories, no significant differences were likely to have had a work incident in the previous 12
found (Table la). However, the distribution of work- months compared with the other two types of hospital.
load categories differed significantly across the hospital Only 31 of those nurses with musculoskeletal symp-
toms (2.9%) had had to change department because

Downloaded from https://academic.oup.com/occmed/article-abstract/47/1/25/1414172 by guest on 21 September 2019

categories; the Dutch nurses were more likely to work
in heavy departments (Table lb). of their problems during the previous 12 months.

Table 2. Perceived causation of musculoskeletal symptoms at any time, together with related need for absence or change of duties,
broken down by type of hospital and type of duty; figures given as percentage and number of responses

Attributed to work Needed absence Needed change of duties

Yes No Yes responses Yes responses

% (n) % (n) % (n) % (n)

Hospital
Academic 50.4 (348) 49.6 (343) 18.3 (129) 2.9 (20)
Catholic 54.3 (100) 45.7 (84) 19.5 (37) 3.6 (7)
Dutch 54.6 (95) 45.4 (79) 18.8 (35) 2.1 (4)
X2;P 1.59; ns* 0 .14; ns* 0.75; ns*
Workload
Heavy 50.9 (236) 49.1 (228) 16.7 (81) 3.1 (15)
Moderate 52.7 (176) 47.3 (158) 18.6 (63) 2.9 (10)
Light 51.9 (121) 48.1 (112) 21.2 (51) 2.6 (6)
Z2;P 0.27; ns* 2.22; ns* 0.16; ns*

ns = no significant difference between the hospital categories or between the workload categories.

Tabl* 3. Mean scores and standard deviations (SD) from responses to the psychosodal instruments broken down by type of hospital:
Academic, n = 795; Catholic, n = 215; or Dutch, n = 206.

Instrument Subscale Academic Catholic Dutch

Mean (SD) Mean (SD) Mean (SD)
MZ Depression 12.8 (7.8) 13.8 (7.8) 11.0 (6.1)
PAW Job satisfaction 25.1 (5.1) 24.5 (5.4) 26.1 (4.2)
Mental stress 15.9 (3.2) 15.4 (3.3) 16.4 (2.7)
Social support 14.2 (3.1) 14.3 (3.1) 15.5 (2.7)
CSQ Catastrophizing 4.7 (5.2) 5.6 (5.1) 3.8 (5.1)
Self statements 20.9 (7.8) 21.5 (7.1) 19.2 (7.5)
Diverting attention 10.0 (7.2) 11.3 (7.3) 10.9 (7.3)
Increasing activity 13.8 (7.3) 14.5 (6.9) 14.3 (7-5)
Ignoring pain 13.6 (7.1) 14.3 (7.2) 14.6 (7.5)
Pain behaviours 16.4 (6.1) 17.1 (5.7) 15.8 (6.1)
Praying/hoping 6.0 (5.6) 7.1 (6.1) 6.0 (5.8)
Pain sensation 6.7 (5.8) 7.1 (5.8) 6.4 (6.2)
FABQ Physical activity 14.6 (5.6) 15.1 (4.8) 12.3 (5.7)
Work activity 18.4 (9.3) 20.1 (9.0) 17.2 (9.1)
PLC Pain control 10.1 (4.4) 11.0 (4.3) 10.6 (4.7)
Pain responsibility 9.6 (2.4) 9.1 (2.4) 9.7 (2.5)

A high score on these Instruments represents abnormal psychometrics, with the exception of PAW scores, where low scores represent negative attitudes.

Figures In bold are significantly different from the other hospitals at the 5% level.
 A. K. Burton eta/.: Back and musculoskeletal problems in nurses 29

Psychosocial profiles Work loss

Statistically significant differences were found between The proportion of nurses having to stay at home due
the hospital types for the psychosocial score means, to musculoskeletal injury in the preceding 12 months
and there was a consistent pattern to the data (Table was 18.6%; this proportion was not significantly
3). The nurses in the Dutch hospitals were less different across either the hospitals or workload
depressed (MZ), had more positive attitudes about job categories.
satisfaction, social support and mental stress (PAW), Work incidents occurring in the 12 months prior to
catastrophized less, used positive coping strategies interview, and perceived as a cause of symptoms, were
(CSQ), and had less fear avoidance beliefs for both more frequent in the Catholic hospitals, but there was
work and physical activity (FABQ). When looking at no significant difference in the average number of days
workload, there were no more significant differences lost for these incidents between the hospital types.
in psychosocial scores across the categories than might Neither were symptomatic work incidents or average
be expected by chance alone (Table 4), but heavy days lost significantly associated with workload, though

Downloaded from https://academic.oup.com/occmed/article-abstract/47/1/25/1414172 by guest on 21 September 2019

departments did show lower job satisfaction (PAW) interestingly the heavy departments had the lowest
and greater fear of work activity (FABQ). absence rate (Table 5).

Table 4. Mean scores and standard deviations (SD) from responses to the psychosocial instruments broken down by categories of work-
load: Heavy, n - 530; Moderate, n » 384; Light, n = 280

Instrument Subscale Heavy Moderate Ught

Mean (SD) Mean (SD) Mean (SD)
MZ Depression 12.9 (7.7) 13.0 (7.9) 11.9 (6.9)
PAW Job satisfaction 24.7 (4.9) 25.5 (5.0) 25.7 (5.1)
Mental stress 15.8 (3.1) 15.9 (3.2) 16.0 (3.4)
Social support 14.4 (3.2) 14.5 (2.7) 14.2 (3.3)
CSQ Catastrophizing 4.8 (5.3) 4.7 (5.3) 4.7 (5.2)
Self statements 20.9 (7.3) 20.8 (8.1) 20.3 (7.7)
Diverting attention 10.7 (7.4) 10.6 (7.5) 9.6 (6.6)
Increasing activity 14.2 (7.6) 14.4 (7.2) 13.1 (6.7)
Ignoring pain 14.4 (7.1) 14.3 (7.3) 12.6 (72)
Pain behaviours 16.9 (5.7) 16.2 (6.0) 16.0 (6.4)
Praying/hoping 6.3 (5.8) 5.6 (5.7) 6.7 (5.4)
Pain sensation 6.8 (6.0) 72 (6.0) 6.0 (5.6)
FABQ Physical activity 14.9 (5.4) 13.7 (5.1) 14.3 (6.4)
Work activity 20.2 (8.9) 16.6 (9.6) 18.1 (9.1)
PLC Pain control 10.7 (4.1) 10.1 (4.8) 10.0 (4.5)
Pain responsibility 9.3 (2.4) 9.9 (2.4) 9.3 (2.5)

A high score on these Instruments represents abnormal psychometrtcs, with the exception of PAW scores, where low scores represent negative attitudes.

Figures In bold are significantly different from the other workload categories at the 5% level.

Table 5. Muscuioskeleta] symptoms attributed to work In the 12 months prior to Interview and the absence taken, broken down by
type of hospital and category of workload

Attributed to work Absence taken (days)

% yes responses Yes responses (n) Mean (SD)

Hospital
Academic 55.4 (320) 23.4 (44.6)
Catholic 66.2 (102) 20.4 (36.5)
Dutch 59.9 (88) 32.7 (42.3)
X 2 :P X2 = 6.14; p < 0.05 ns"
Workload
Heavy 59.7, (230) 20.3 (32.5)
Moderate 57.1 (157) 31.1 (51.3)
Ught 55.7 (113) 26.2 (53.7)
X Z ;P X =1.02; ns* ns1
• ns » no significant difference between the hospital categories or between the workload categories.

Figure In bold denotes a significantly higher proportion of Catholic nurses attributing their symptoms to work.
30 Occup Mod. Vol. 47, 1997

DISCUSSION psychosocial domain. The responses from the Dutch

nurses revealed that, on average, they were significantly
This study concerned a large sample of nurses from more positive on a range of psychosocial measures,
different types of hospital in two countries, and has notably those concerned with attitudes towards work
concentrated on relationships between musculoskeletal and activity; they also showed less depressive tendencies
symptoms, in particular low back trouble, and work- and had more positive pain coping strategies. It is
load with the additional consideration of psychosocial possible that the Dutch nurses, perhaps because of
factors. their positive attitudes and beliefs, simply did not
Some comments about the methodology are perti- report problems to the same extent as the Belgian
nent in order to estimate the extent to which the nurses. It has been found that reporting of back pain
findings can be generalized. The sample was of by British nurses increased between 1983 and 1993,25
volunteers; it involved largely anamnestic data and the following an increased awareness of back trouble by
workloads were not measured directly. There was no virtue of the adoption of the European Directive
90/269/EEC on the manual handling of loads sometime

Downloaded from https://academic.oup.com/occmed/article-abstract/47/1/25/1414172 by guest on 21 September 2019

evidence that individuals with a history of back pain
were over-represented in the sample, in fact the preva- after 1990. The Belgian nurses could have been
lence was slightly lower than a previous report on displaying a similar reporting effect, which may be
Belgian nurses.10 We believe that the sample is repre- related to their work perceptions. It is also possible
sentative of nurses working in a variety of hospitals that the inter-country differences could be explained
in Belgium and The Netherlands; the numbers are by differences in staffing levels and/or local implemen-
high and the volunteers came from a typical mix of tation of legislation. Although compliance with the
departments with a typical sex distribution. Admit- European Directive is doubtless incomplete, it might
tedly, anamnestic data is subject to error but we have be reasoned that the hospitals concerned would have
concentrated on a global picture which should be little implemented the principles of reduced manual
influenced by forgetfulness or over-estimations. It handling operations to some (similar) extent. There
would clearly have been preferable to have had formal was no reason, so far as we could discern, to indicate
task assessments to estimate workload but this was not that implementation, staffing levels or use of mechanical
practicable in such a large study; that the classifications aids was substantially different between the countries.
from self-assessed workload were in agreement with a The workloads at the time of the study are probably
standard protocol suggests that they are adequately a reasonable representation of contemporary expo-
representative of the work for the purposes of the sures, so it was a little surprising that the point
present investigation. Although previous reports have prevalence figure for back pain was no better than
found differences between males and females in respect reported some 10 years previously in nurses.26
of symptoms and associated work loss, those differ- That the point prevalence of LBT is not dependent
ences have not been substantial. Therefore, we chose on workload has been noted previously among nurses.26
not to analyze males and females separately (even with However, the lower prevalence of a history of LBT with
such a large sample, the numbers of males would have heavier workloads found here contradicts previous
left statistical power inadequate); our intention at this reports (e.g. Videman et a/."). Because of methodo-
stage was to examine nursing work in general terms. logical differences and definitions of symptoms,
The prevalence of musculoskeletal problems reported comparisons between occupational prevalence studies
from this survey includes low back trouble. The fact are difficult. It is possible that our measure of workload
that the low back figures were only 10-15% lower than was insensitive, but the fact that the 'heavy' group
for all musculoskeletal complaints supports the here had a typical exposure (in hours per week) some-
contention that the low back is an important site for what higher than the equivalent group in the Videman
possible injury from nursing activities. et al. study renders this unlikely. Their study compared
The Dutch nurses reported a significantly lower nursing aides with qualified nurses (where the aides
occurrence both of back trouble and musculoskeletal had the higher workload and the higher prevalence),
problems in general than did the Belgian nurses, but when their whole sample was combined the rela-
despite the fact that they more frequently worked in tionship between 'heavy' work and previous back pain
departments with a higher workload. In fact, workload was less clear cut — only for those aged less than 30
was not generally associated with a higher occurrence years was the relationship statistically significant."
of symptoms, a higher rate of absence, a greater The present study found no significant relationship
perception that work had caused injury, or a greater need between present workload and either work loss in the
for change of duties due to injury. The proportion of previous 12 months, or the belief that work had been
symptomatic nurses who changed duties was surpris- a cause of symptoms in the previous 12 months. It
ingly low; this may be a reflection of their purported has been commented that simply reducing loads on
stoicism,24 or it could reflect a limited opportunity for the spine through reductions in manual handling is
redeployment due to specific needs for nursing unlikely to significantly improve matters when only
specialities in the different hospital departments. manual handling training is implemented.27 Alternative
Apart from the workloads, the striking difference preventive strategies which involve an ergonomic
between the Dutch and Belgian nurses was in the approach, proposing job redesign through task analysis
 A. K. Burton ef a/.: Back and muscutosketetal problems In nurses 31

and training, have been proposed27 and have met with REFERENCES
some success,28 yet nurses apparently continue to see
patient handling as a major factor in the development 1. Hrudey WH Overdiagnosis and overtxeatment of low back pain:
of back trouble.25 Long term effects. J Occup Rehab 1991; 1: 303-311.
2. Leamon TB. Research to reality: a critical review of the validity
The potential for ergonomics intervention to reduce of various criteria for the prevention of occupationalry induced
the incidence of injury to nurses24 is not in question, low back pain disability. Ergonomics 1994; 37: 1959-1974.
but the epidemiological profile of musculoskeletal 3. Cust G, Pearson JCG, Mair A. The prevalence of low back
disorders (in all working groups) is confounded with pain in nurses. Int Nursing Rev 1972; 19: 169-179.
symptoms that are not directly work-related. What 4. Stubbs D, Buckle P. The epidemiology of back pain in nurses.
MedEdu 1984; 32: 935-938.
becomes of the individual after developing a disorder
5. Ferguson D. Strain injuries in hospital employees. MedJAust
(whether by work-injury or otherwise), how quickly 1970; 1: 376-379.
they recover or how much absence they take, probably 6. Harber P, Billet E, Gutowski M, Soo-Hoo K, Lew M, Roman A.
has more to do with psychosocial than physical Occupational low back pain in hospital nurses. J Occup Med
issues.13*29 The results of the present study suggest that

Downloaded from https://academic.oup.com/occmed/article-abstract/47/1/25/1414172 by guest on 21 September 2019

1985; 27: 518-524.
the nurses' reporting of back or musculoskeletal prob- 7. Hickman M, Mason V. The Prevalence of Back Pain. A Report
Prepared for the Department of Health. London, UK: Office of
lems, as well as the subsequent course, is influenced Population Censuses Surveys, 1993.
by psychosocial phenomena. Thus, if the longer-term 8. Health Services Advisory Commission. Guidance on Manual
effects of back and musculoskeletal disorders in nurses Handling of Loads in the Health Services. Health Services Ad-
is to be reduced, some form of psychosocial interven- visory Commission, 1992.
tion, before symptoms develop, would appear to be 9. Venning PJ, Walter SD, Stitt LW. Personal and job-related
factors as determinants of incidence of back injuries among
an appropriate addition to conventional ergonomic
nursing personnel. J Occup Med 1987; 29: 820-825.
intervention. Indeed, a recent intervention study in 10. Stappaerts K Lage rugpijn: een onderzoek bij verplegenden en
industry has shown that broadcasting a simple richtUjnenvoorpreventie. Lcuven, Belgium: Hermes, Kul, 1989.
educational pamphlet (stressing positive attitudes and 11. Videman T, Numinen T, Tola S, Kuorinka I, Vanharranta H,
beliefs) can reduce absenteeism due to low back Troup JDG. Low back pain in nurses and some loading factors
trouble.14 of work. Spine 1984; 9: 400-404.
12. PheasantS. Back pain at work. In: Ergonomics,\fbrk and Health.
In conclusion, the relatively high workload of Dutch London, UK: Macmillan Press, 1991. 3: 57-76.
nurses was not associated with an increased occurrence 13. SymondsTL, Burton AK.Tillotson KM, Main CJ. Do attitudes
of back and musculoskeletal complaints, statistically and beliefs influence work loss due to low back trouble? Occup
significant work loss or the perception that work is Med 1996; 46: 25-32
harmful. This may be because of a more positive 14. Symonds TL, Burton AK, Tillotson KM, Main CJ. Absence
due to low back trouble can be reduced by psychosocial inter-
outlook on their health and job (which could be an vention at the workplace. Spine 1995; 20: 2738-2745.
inherent social characteristic or due to some, uniden- 15. Berg ST, Dahl L, Dehlin O, Hedenrud B. Psychological
tified, positive and supportive approach by hospitals perception of nursing aides' work. ScandJ Rehab Med 1976;
in The Netherlands). Whether the relationship between 8:79-84.
low prevalence and positive psychometrics is cause or 16. Klaber-Moffett JA, Hughes GI, Griffiths P. A longitudinal study
effect can only be answered by longitudinal studies, of low back pain in student nurses. IntJ Nursing Studies 1993;
30: 197-212.
but there is clinical evidence that the psychological 17. Waddell G, Newton M, Henderson I, Somerville D, Main CJ.
factors associated with chronic back pain may be A fear avoidance belief questionnaire (FABQ) and the role of
present before the onset of trouble.29 The title of this fear-avoidance beliefs in chronic low back pain and disability.
paper questions the value of ergonomic intervention. Pain 1993; 52: 157-168.
The results presented here suggest that 'minimal nursing' 18. Rosenstiel AK, Keefe FJ. The use of coping strategies in chronic
' low back pain patients: relationship to patient characteristics
plans alone, whilst having the potential to reduce injury
and current adjustment. Pain 1983; 17: 33-44.
rates, will have a sub-optimal influence on symptoma- 19. Main CJ, Waddell G. A comparison of cognitive measures in
tology. This becomes important when considering low back pain: statistical structure and clinical validity at initial
either the detrimental effect that musculoskeletal assessment. Pain 1991; 46: 287-298.
symptoms have on morale and work loss, or the risk 20. Main CJ, Waddell G. The detection of psychological abnor-
of chronic sequelae. We believe there is justification mality in chronic low back pain using four simple scales. Curr
Con Pain 1984; 2: 10-15.
for adopting an additional programme of psychosocial
21. Garg A, Owen PD, Carlson B. An ergonomic evaluation of
intervention which challenges misconceptions about nursing assistants' job in a nursing home. Ergonomics 1992; 35:
the detrimental effects of work on health, and encour- 979-995.
ages an element of self-management of symptoms. 22. Caboor D, Claryis JP, Zinzen E, et aL A method to identify the
attitude of 'Supernurses' towards their daily tasks. London,
U K Society for Back Pain Research, 15 October 1993.
23. Stubbs DA, Rivers PM, Hudson MP, et al. Patient handling
ACKNOWLEDGEMENTS and truncal stress in nursing. Proceedings of conference
organised by the Nursing Practice Research Unit Northwick
The authors would like to acknowledge the support Park Hospital: Back Pain Association and DHSS, 1980.
24. Pheasant S, Stubbs D. Back pain in nurses: epidemiology and
of the Belgian Federal Services for scientific, technical risk assessment Appl Erg 1992; 23: 226-232.
and cultural matters, and the British Council for 25. Leighton DJ, Reilly T. Epidemiological aspects of back pain:
assistance with travel costs. the incidence and prevalence of back pain in nurses compared
32 Occup. Med. Vol. 47, 1997

to the general population. Occup Med 1995; 45: 263-267. 28. Aird JW, Nyran P, Roberts G, Aghazadeh F, eds. Trends in
26. Stubbs DA, Buckle PW, Hudson MP, Rivers PM, Worringham Ergonomics/Human Factors V. North-Holland; Elsevicr Science
CJ. Back pain in the nursing profession. I. Epidemiology and Publishers BV, 1988.
pilot methodology. Ergonomics 1983; 26: 755-765. 29. Burton AK, Tillotson KM, Main CJ, Hollis S. Psychosocial
27. Stubbs DA, Buckle P, Hudson M, Rivers P. Back pain in the predictors of outcome in acute and subchronic low back trouble.
nursing profession II: The effectiveness of training. Ergonomics Spine 1995; 20: 722-728.
1983; 26: 767-779.

Downloaded from https://academic.oup.com/occmed/article-abstract/47/1/25/1414172 by guest on 21 September 2019