160 Occupational and Environmental Medicine 1995;52:160-163

Manual handling activities and risk of low back

pain in nurses
Julia Smedley, Peter Egger, Cyrus Cooper, David Coggon
 it, 1616 nificant associations non- tasks see table 4).
Univ women, were found with fre- musculos Low back pain
ersi
ty the quency of manually keletal was defined as pain
of lifetime moving patients symptom lasting for longer than
Sout Abstr prevalen
handling that the job
s aimed a day in an area
hamp around on the bed, entails. Nurses are fre-
ton, act ce of manually quently required to to (indicated in a
Sout Objecti back investigat diagram) between the
hamp transferring patients undertake heavy lifting,
ve-To pain was e whether twelfth ribs and the
ton between bed and often with a bent or twisted
Gener investig 60% and psy- gluteal folds. Subjects
chair, and manually posture, and bio-mechanical
al ate the the one were asked to exclude
lifting patients from investigations chologic
Hosp risk year
have
pain that only occurred
ital the floor. confirmed that such tasks al factors
, factors period in association with
In contrast, no clear generate high spinal that
Sout for low prevalen pregnancy,
hamp increase in risk was stresses.45 influence
back ce 45%. menstruation, or
ton found in relation to Strategies to control the reporting febrile illness.
J pain in 10% had
Smed transfer of patients hazard include the of other
Sciatica was
ley hospital been
with canvas and elimination or symptom
P nurses, absent defined as pain
Egge with
poles, manually modification of activities s
radiating down the
from
r lifting patients in that carry the highest risk, (particula
particula work leg to below the
C Co and out of the bath, but first the levels of risk rly pain
op r because knee.
er or lifting patients associated with different and psy-
D Co
emphasi of back
with mechanical tasks must be assessed. chologica
The
gg s on the pain for
on role of a aids. Confir-mation Biomechanical evaluation l ill reproducib
Corr spe-cific cumulati of these findings is contributes to the health) ility of
nursing now being sought in assessment, but also
espo
nden
ve period
epidemiological evi-dence affect the
identical
activitie exceedi a prospective study
ce s. ng four of the same is also required. We report risk of questions
to:
Dr Method population. a survey in which the risk reported
J
weeks.
Smed low back
ley; s-A Rates in Conclusions-This of back symptoms in nurses
MRC cross men study confirms was examined in relation to pain.
Envi sectiona a range of lifting tasks. Subjects
ronm were that low back pain
enta l survey generally is highly prevalent were
l
Epidem of similar to among nurses and asked to
iology 2405 those in is associated with a Methods indicate
Unit,
Univer
nurses women. high level of The study how often
sity employe In sickness absence. population (never,
of
Southa d by a women People who often comprised the 2405 occasiona
mpton, group of back report non- hospital based lly, or
Sout nurses employed by
hamp teaching pain musculoskeletal frequentl
ton hospitals during Southampton y) they
Gener symptoms were
al was the University Hospitals experienc
significantly more Trust.
Hospi previous
tal, carried likely to report low ed
The Trust provides in
South out. Self 12 back pain. Specific headache
ampto patient facilities in most
n administ months manual handling s, period
was clinical specialties other
S016 ered tasks were pains,
6YD. than psychiatry.
question weakly associated with an constant
Accept
associate It is also a centre for
ed 24 naires increased risk of tiredness,
Novemb d with undergraduate
er were back pain; low
1994 height, medical and nursing
used to however, no such mood,
and was education, but nurses
M collect asso-ciation was and
significa in training are not
R informat found with feeling
C ntly employed by the Trust
ion mechanised patient tense or
E more and were excluded
about transfers. under
n common from the study. stress.
occupati
v in those (Occup Environ Each member of the
Occup
onal
i who Med study population was sent ational
r activities
reported 1995;52:160-163) a postal questionnaire details
o , non-
frequent asking about vari-ous included a
n occupati
non- Keywords: back pain; non-occupational risk self
m onal risk nurses; epidemiology
musculo- factors for back reported
e factors complaints including estimate of
skeletal
n for back Several studies have the number
symptom age, height, reproduc-
t sympto of times
s such as indicated an unusually tive history, and
a ms, and the subject
l headache high prevalence of back smoking habits; about
history and low fre-quency of certain non- performed
E disorders in nurses,'"
p of low mood. musculoskeletal each of a
i back and this is widely series of
After symptoms; about details
d specified
pain. adjustme attributed to the manual of his or her current job,
e patient
m Results- nt for and about lifetime and
handling
i The height recent history of low
o overall tasks per
l and non- back pain. Non-
average
o
respons musculo responders were sent a
e rate shift (for
g skeletal single reminder after a
was full list of
y
symptom month. patient
U 69%.
n s, sig- The questions about handling
Among
Manual handling activities and risk oflow back pain in nurses 161
about lifetime back pain and associated fea-tures has been assessed
in an earlier study in the general population. Subjects were asked to
complete the questions a second time, one year after their first response.
The level of agreement was 91% for reported lifetime low back pain,
82% (minimum) for reported sciat-ica, and 87% for reported sickness
absence.6
The validity of information on reported fre-quency of tasks handling
patients was mea- sured indirectly by comparing the responses of
nurses of the same grade working on the same ward. Good agreement
was found for task fre-quency, only 16% of reports differed from the
modal reported frequency for each job.7
Table 1 shows the prevalence of reported low back pain and
associated disability.
Among women the lifetime prevalence of back pain was 60%,
and 10% had been absent from work because of back pain for
a cumula-tive period exceeding four weeks. The one year
period prevalence of low back pain in women was 45%. Rates
in men were generally similar to those in women, although
fewer men had consulted a general practitioner about back
problems.
Because of the small number of male responders, further
analyses were restricted to women. Table 2 shows the associations of

Associations between low back pain and risk factors low back pain in the past 12 months with height and with report

were assessed by logistic regres-sion. Risk estimates are ofvarious non-musculoskeletal symptoms. Risk of back pain increased

quoted with 95% con-fidence intervals (95% Cls). with stature, but the trend was weak and failed to reach significance.
More striking was the rela-tion to other symptoms. There was a strong
association between reported frequent headache, period pains,
Results fatigue, low mood, or stress and reported low back pain.
Questionnaires were completed by 1659 nurses (a response Table 3 shows the relation of low back pain to certain other non-
rate of 69%), of whom occupational risk factors.
97% were women and 50% worked full time. After adjustment for age, height, and report of non-
The ages of responders ranged from 19 to 65 with mean 38 musculoskeletal symptoms, all of the associations
and median 36 years. All levels of seniority were were weak and none was signifi-cant.
represented from nursing assistant to senior clinical and The occupational risk factors studied included hours worked
administrative grades. (full time or part time), shifts worked (days, nights, or both), and
grade (nursing assistant or higher).
Analysis, with logistic regression, showed no association between any
of these factors and the risk of back pain. Back symptoms in the past
Table 1 Prevalence (%) of reported low back pain and year were reported less frequently than the average by paediatric nurses,
associated disability but other-wise there were no clear patterns by depart- ment. Associations
Men Women were found, however, in relation to the frequency of certain lifting
(n = 43*) (n = 1616*) activities (table 4). In particular, risk increased with frequency of
Low back pain, ever 56 60 manually moving patients around on the bed, and was signifi-cantly
Sciatica, ever 23 25
Low back pain lasting raised in nurses who manually trans-ferred patients between bed and chair
> 1 year in total 5 9 and
Cumulative lifetime sickness
absence for low back pain
> 4 weeks 9 10
Consulted a general practitioner
because of low back pain, ever 29 40
Low back pain in past 12 months 39 45
*Answers to specific questions were missing for up to 36 sub-jects
(one man and 35 women).
Table 2 Association of low back pain in the past 12 months with height and report of non-
musculoskeletal symptoms in female nurses
Each risk factor Risk estimates
With Without examined mutually
back pain* back pain* separately adjusted
Risk factor n n OR (95% CI) OR (95% CI)
Height (cm):
<157 160 223
158-163 187 231 1-1 (0-9-1-5) 1-2 (0-9-1-6)
164-166 119 124 1-3 (1-0-1-8) 1-4 (1-0-2-0)
167-170 161 180 1-3 (0-9-1-7) 1-3 (1-0-1-8)
) 171 104 113 1-3 (0-9-1-8) 1-3 (0-9-1-9)
Headache:
Never or occasionally 566 760
Frequently 156 100 2-1 (1-6-2-7) 1-7 (1-3-2-3)
Period pain:
Never or occasionally 521 686
Frequently 181 149 1-5 (1-2-2-0) 1-4 (1-0-1-8)
Fatigue:
Never or occasionally 480 682
Frequently 244 171 2-0 (1-6-2-5) 1-4 (1-1-1-9)
Low mood:
Never or occasionally 629 818
Frequently 94 38 3-2 (2-2-4-8) 2-0 (1-2-3-1)
Stress:
Never or occasionally 545 740 1
Frequently 176 115 2-1 (1-6-2-7) 1-3 (1-0-1-8)
for in
*Answers to specific questions were missing for up to 70 women. All risk estimates were adjusted age quintiles.
162
Smedley, Egger, Cooper, Coggon
 with a hoist: 20% of women
0 611
who had sometimes to 1+ 58 compared with the
lift patients up from the
Manually lift a patient in or out Discussion shortest
of the bath: The prevalence of back symptoms 20%. '7 If there is a
floor. These associations 0 569 recorded in
1+ 102 relation between height
persisted when risk Lift a patient in or out of the this survey cannot be compared directly and back pain in women,
estimates were adjusted bath with a hoist:
with those found in most other studies then it does not
0 426
for height and report of 1-4 because of seem to be as strong as in
222
non-musculoskeletal differences in the method of men, and would not
5+ 24 investigation, the
symptoms. In contrast, no *Information was missing for up to 213 women; t(a) justify selective
adjusted for age (in quintiles) and height
clear increase in risk was in table definition of back pain, and the age exclusion of tall women
(classified as 2) (b) adjusted for structures
found in relation to age (in quintiles) and for height and of from
report of non-musculoskeletal symptoms the populations studied. In a postal to
transfer of patients with (classified as in table 2). survey of entry the nursing
canvas and poles, 1495 women from eight areas of Britain profession.
that used identical question to In contrast with other
manually lifting patients an
studies, we found no
into and out of ascertain back pain, lifetime
relation ofback
prevalence varied from 45-2% at pain to
the bath, or reproductive hiStory,3
age 20-29 years to 63-7% at ages 50-59
lifting patients years,
10 18
smoking,"' 1920 or
with whereas one year prevalences were
severe trauma.2'
Of the
mechanical in the occupational
aids. range 27-0-43-7%.8 The activities studied,
prevalences that we manually moving
recorded in our patients around on the
study were higher
a
bed, transferring
(despite higher response rate), patients between bed and
Table 3 Association oflow back pain in the past which supports the view that nurses are
12 months with non-occupational risk factors in at an increased risk of chair, and lifting
female nurses back disorders. Particularly notable patients up from the
floor were associated with
was the high proportion of nurses who
With Without increased risk of back
had taken
back pain back pain
Risk factor n n
more than four weeks off work pain, and the
Current participation in because of back trouble (9%
regular sport or exercise: of men and 10% of associations
No 211 276 women), which
Yes 511 572
emphasises the burden persisted after
Number of children:
0 297 293 that the problem places adjustment
1-2 294 374 both on the nurses and
3+ 128 183
Cigarette smoking: employers. for height and
Never 409 509 Of the non-occupational risk
Ex 203 219
Current 109 131 factors for report of non-
Previous fall or road back pain that we examined, non-
traffic accident leading
to hospital admission: musculo- musculoskelet
No 654 786 skeletal symptoms showed the
Yes 68 73 strongest asso-ciations with back al symptoms.
*Information was missing for up to 122 women; t(a) adjusted for age (in quintiles) pain. This may reflect
and height

(classified as in table 2) (b) adjusted for age (in quintiles) and for
differences in perception of back pain In contrast,
rather
height and report of non-mus-culoskeletal symptoms (classified as
in table 2). than an
increased risk of physical no association
the
damage to spine, but without good
objective mea-sures of spinal pathology, was found
the distinction can-

Table 4 Association oflow back pain in the not be tested. The finding accords with lifting or
past 12 months with occupational activities with those
in female nurses of other studies that have shown that transfer-ring
are
people who depressed or
dissatisfied with their patients with
Activity: With Without work are more likely to complain of back
frequency in an back pain * back pain* symptoms.9-12 In particular, one a hoist or
average working shift n n longitudinal study has shown that
Transfer a patient on depression predicts the onset of with canvas
canvas and poles:
subsequent back pain, and that the
0
1-4
450
197
547
228 association therefore is not explained
and poles. In
5+ 43 41
Manually transfer a patient
simply
by the occurrence of depression
general, this
between bed and chair:
0 168 274 secondary to
1-4 227 257 back
trouble."I
pattern of risk
5-9 128 123
10+
Transfer a patient between
152 157 In female nurses back pain was also
is consis-tent
bed and chair with a hoist: weakly associated with height. Other
0 531 669 studies of back with
1-4 120 109 pain and stature in women have
5+ 21 26
Manually move a patient produced inconsistent results. biomechanica
around on the bed: Surveys in Sweden,'4
0 84 137 l evaluation,4
1-4 169 246 Denmark,'5 and Britain'6 have failed to

5-9 167 175 show a relation, but an analysis of back

10+
Manually lift a patient up off
270 260 symptoms in a but
the floor: British longitudinal study did suggest an
0 438 567 asso-ciation, with a relative risk of 1-5 in the
1+ 230 231
Lift a patient from the floor tallest
Manual handling activities and risk of low back pain in nurses 163
 in the
Meanwhile our findings development of low-back
pain. Scand J Work Environ
interpretation must take have important implications the use of turning sheets and Health 1991;
for the assessment and reduction other such aids 17:420-4.
into account several 17 Kuh DJL, Coggon D, Mann S,
of risk to nurses from manual is being encouraged. Also, Cooper C,
Yusuf E.
limitations of the study Height, occupation and back
handling activi- is
greater use being in
design. pain a national prospec-tive
ties. Particular attention should be made of hoists to assist lifting and study. BrjRheumatol
Firstly, the response to the given to transfer of 1993;32:911-6.
18 Svensson H-O, Andersson GBJ,
questionnaire was only 69%. techniques for
around on
moving patients. Hagstad A, Jansson P-O.
The relationship of
low back
Although comparable with the patients hospital beds and
This study was
supported
in
part by a pain to pregnancy
response rate in other surveys of this and for bed to chair transfers. gynaecologic factors.
from Wessex
type, this leaves open the The results of this study grant Spine 1990;15:371-5.
Regional Health Authority. We thank 19 Biering-Sorenson F, Thomsen C.
possibility that the responders support the use of hoists as a Graham Wield for his Medical, social and occupational
were unrepresentative. In means of risk reduction. In with the history as risk indicators for low

Southampton
help back trou-ble in a general
particular, nurses whose work was nurses have for population. Spine 1986;11:720-
computing. 5.
most physically demanding may some years been instructed to use
1 Stubbs DA, Buckle P, Hudson 20 Svensson H-0, Viden A,
Wilhelmsson C, Andersson GBJ.
have been more handicapped by the shoulder (Australian) lift to MP, Rivers PM,
Low back pain in relation to other diseases
back disorders and therefore more Worringham CQ. Back pain in the and cardio-
move patients around on the nursing profession. vascular risk factors. Spine
inclined to par-ticipate in the study. Part I-Epidemiology
and
pilot 1983;8:277-85.
bed.22 Now, methodology. Ergonomics 21 Walsh K, Cruddas M, Coggon D. Risk of low
This would tend to exag-gerate
1983;26:755-65. back pain in people admitted to
associations with occupational 2 Harber P, Billet E, Gutowski M, hospital for traffic accidents and falls.
activity. SooHoo D, Lew M, _JEpidemiol Community
low Health 1992;46:231-3.
Roman A. Occupational
Also, may
errors have back in
pain hospital nurses.J7
22 Corlett EN, Lloyd PV, Tarling C,
Troup JDG, Wright B.
occurred from biased recall of Occup Med 1985;27:518- The guide to the handling of
24.
symptoms or activities. Nurses 3 Videman T, Nurminen T, Tola S,
patients. Teddington:

with Kuorinka I, Vanharanta National Back Pain

H, Troup JDG. Low back Association, 1992.
back pain may have been pain in nurses and some load-
more aware of the ing factors at work. Spine
1984;9:400-4.
lifting requirements of their job 4 Garg A, Owen BD, Carlson B. An
and reported tasks differently ergonomic evaluation of nursing
home.
assistants' job in a nursing
from those who were pain free, Ergonomics
and nurses with heavier jobs may 1992;35:979-95.
5 Gagnon M, Sicard C, Sirois JP. Evaluation
have been more conscious of
of forces on the lumbo-sacral joint
their symptoms and had a lower and assessment of work and energy
transfers in nursing aides lifting
threshold for reporting back patients. Ergonomics
pain. 1986;29:407-21.
6 Walsh K. An epidemiological study
low back
Again, this would inflate risk of pain.
of
estimates. On the other hand, Southampton: University
1992.
Southampton (MD
people with back symptoms thesis.)
may have been selected out of the 7 Campbell L, Pannett B, Egger P,
D.
Cooper C, Coggon
most physi-cally demanding Validity of self-reported
activities. Proceed-
occupational ings
jobs. This bias would tend to ofthe 10th International
in
Symposium on Epidemiology
obscure associations between Health
Occupational
back pain and [abstract].
Como: 1994:196.

lifting. 8 Walsh K, Cruddas M, Coggon D.

Low back pain in eight areas of
Because of the potential Britain. J7 Epidemiol
for such biases, the findings Community Health
1992;46:227-30.
require further testing with a 9 Bigos SJ, Battie MC, Spengler
Fisher
DM, LD, Fordyce
prospective study design where al. A
WE, Hansson TH, et
are of work
activities ascertained in prospective study
perceptions and psychosocial
people who are pain free, and factors
affecting the report of
back injury. Spine 1991;16:1-6.
they are then followed up to 10 Frymoyer JW, Pope MH,
Rosen
assess the subse- Constanza MC, JC,
quent incidence of symptoms. It Goggin JE, Wilder DG.
Epidemiologic studies of low-
may also be important to back pain. Spine
1980;5:419-23.
distinguish chronic or recurrent
11 Nagi SZ, Riley LE, Newby
LG. A social

back symptoms from new epidemiology of back pain in a

episodes of low back pain. The general population. J7 Chron Dis
1973;26:
quality of information about 769-79.
12 Bergquist-Ullman M, Larsson U. Acute
recurrence of symptoms obtained
low back pain in industry. Acta
in a cross Orthop Scand 1977;170(suppl):
to 1-117.
sectional study is likely be
13 Croft P, Ferry S, Papageorgiou A,
compromised by poor recall. Jayson M, Silman A.
We are currently following up Depression and back pain in the
general population. BrJ
this
Rheumatol 1994;33(abstract
105.
suppl I):
population of nurses, with the 14 Hirsch C, Jonsson B, Lewin T.
aim of over-coming some of these Low-back symptoms in a
Swedish female population. Clin
methodological difficul- Orthop 1969;63: 171-6.
15 Biering-Sorensen F. Physical
ties. The question of recurrence measurements as risk indica-tors for
rate will be low-back trouble over a one-year
considered period. Spine

specifically in the 1984;9: 106-19.

16 Walsh K, Cruddas M, Coggon D.
ongoing Interaction
mechanical
of height and
loading of the spine
prospective study.