Heart (Supplement 1) 1997;78:23-28

Back pain among echocardiographers

Mina Solanki, Douglas Carr, Michael Martin

AbstractObjective-By having to adopt unnaturalpostures, echocardiographers place them-selves at risk of back injury. The presentstudy sought to document the incidenceand severity ofback pain among echocar-diographers, and to identify personal,lifestyle, and occupational characteristicsthat may predict risk ofback pain.Design-A questionnaire concerningoccupational, personal, and lifestyle char-acteristics, as well as the extent andseverity of back pain was completed andreturned by 183 echocardiographers.Results-The standard echocardiographyexamination position of "machine on theleft, patient on the right, and transducerheld in the right hand" (used by 66% ofrespondents) greatly increased the risk ofback pain (odds ratio (OR) = 4-9; 95%confidence interval (95%CI) = 1.49-16.4).Echocardiographers with more than 11years experience in the field were also atrisk of back pain (OR = 3-4; 95%CI =0-97-11.6). The intensity of back pain wasstrongly associated with the amount oftime spent on echocardiography exami-nations (P = 0.035), limitation of lifestyle(P < 0.001), and interference with work(P < 0.001). There was a significantlyhigher relative risk of back pain for therespondents whose job involved frequentor occasional lifting compared with non-lifters (P = 0-01; OR = 4.8, 95%CI =1.25-18.7). At least some limitation oflifestyle from back pain was reported by56% ofthe respondents.Conclusions-There is a high incidence ofback pain among echocardiographers,and specific tasks and activities increasethe risk of back pain. Further researchshould focus on preventative measuresand optimum work station designs thatmay serve as a guide to echocardiographydepartments.

(Heart 1997;78:23-28)

Keywords: back pain; echocardiographers; echocardio-graphy examinations; posture; occupational back pain

Back injury among nursing personnel hasbecome a topic of great interest in occupa-tional health. Individual risk factors includeage, physical fitness," psychological factors,and previous back pain.3 Number of yearsspent on units where frequent lifting was

required,45 heaviness of lifting,6 and frequencyof lifting,' are some of the lifting related causesidentified by the studies.

Echocardiographers perform tasks thatplace them at risk of back pain. Because ofhaving to adopt unnatural postures to performechocardiography, high and low back, neck,shoulder, and wrist pain are frequent com-plaints among echocardiographers. Back painmay be a chronic and sometimes disablingcondition.7 Poor postures sustained for pro-longed periods can result in tissue damage anddisc protrusion, causing longlasting symp-toms.8 Successive .injuries tend to be moresevere, eventually resulting in financial impli-cations from medical costs, decreased pro-ductivity, and absenteeism.910 To avoidoccurrence and recurrence of back pain,echocardiographers should identify activitiesin the work environment that increase theirrisk of back pain. Lack of research in this areahas hampered adequate risk assessment andimprovement in the work environment toensure safer working practices. To our knowl-edge, this is the first study to document theincidence of back pain among echocardiogra-phers and to identify risk factors associatedwith performing echocardiography tasks.The purposes of this study were to: (1) doc-

ument the incidence of back pain among asample of members of the British Society ofEchocardiography; (2) identify demographicand lifestyle characteristics that best forecastback injury; (3) identify occupational charac-teristics related to back pain; and (4) docu-ment the extent and severity of back injurysustained by echocardiographers.

MethodsA survey was performed to assess the fre-quency of back pain among echocardiogra-phers. The membership list of the BritishSociety of Echocardiography contains 1300members of whom 600 are echocardiogra-phers (cardiac technicians). The remainingmembers are medical staff who are involved inechocardiography as part of their cardiologytraining and are relatively infrequent users ofechocardiography equipment. They were con-sidered unlikely to be at risk of back pain andwere excluded from the study. Optimumsample size was calculated and 266 of the ran-domly selected members were sent a question-naire.The questionnaire was divided into four

sections, which covered preferred echocardio-graphy examination methods, personal charac-

CardiologyDepartment,Stockport AcuteServices (NHS) Trust,Stepping Hill Hospital,Poplar Grove,Stockport SK2 7JB,Cheshire, UKM SolankiM MartinSouth West School ofRadiography,University of Derby,Devonshire House,Derbyshire RoyalInfirmary, LondonRoad, DerbyDEl 2QY, UKD CarrCorrespondence to:Dr Solanli.

23

on April 4, 2020 by guest. P

rotected by copyright.http://heart.bm

j.com/

Heart: first published as 10.1136/hrt.78.S

uppl_1.23 on 1 August 1997. D

ownloaded from

Solanki, Carr, Martin

teristics, lifestyle, and prevalence and intensityof back pain (fig 1). The questionnaire wasprecoded for analysis. The questions on inten-sity of back pain were based on Likert-typescale and a few exploratory questions wereopen ended.

Information on echocardiography experi-ence was derived from the question "how longhave you been involved in echocardiogra-phy?", which provided a list of options: lessthan 2 years; between 2 and 4 years; between5 and 10 years; between 11 and 20 years; andmore than 20 years. The amount of time spentscanning (per week) was indicated by therespondents by selecting from a list of nineoptions ranging from "less than 3 hours perweek" to "more than 50 hours per week".Information on work postures or ergonomicswas evaluated by a series of closed questionsreferring to the position of the operator in rela-tion to the echocardiography machine and thepatient, such as "where do you sit to conductthe examination?" and "where is the echomachine situated in relation to you?". Liftingwas recorded as either the job involved liftingor not. Frequency of lifting was addressed bythe question "how often do you lift?" (occa-sionally, frequently, rarely). Work satisfactionwas evaluated by the following options: good,average, poor, and don't know. The abovequestions were derived from personal experi-ence of the author in this field and fromrelated studies of occupations involving similarwork activities, such as radiography and nurs-ing.5' 1

Questions on lifestyle parameters werebased on review of studies on social and occu-pational history in the general population andthe nursing population.2-'2 The question onlevel of exercise had the following options:

Figure 1 Questionnaire design.

once a week, twice a week, three times a week,more than three times a week, hardly ever, andI don't exercise. The influence of preschoolaged children on low back pain has beenrecognised and included in surveys by severalresearchers.'2 13

Questions on pain intensity were derived inpart from a Million Visual Analogue, 14 theOwestry Scaled Back Pain Questionnaire,'5and from various instruments used in similarsettings.'6 The questionnaire contained scaledLikert-type response options pertaining toback pain at work and non-work activities.Close correlation between the disability scoreobtained from the Likert scale and clinicallyobserved disability and symptoms have beenreported. 14-16The questionnaire was pretested on 10 col-

leagues to determine clarity and ease of com-pletion. Consistency questions were used toevaluate whether responses were consistentthus providing support for the internal consis-tency and reliability of the questionnaire.

STATISTICAL ANALYSISCategorical variables were analysed using X2contingency table techniques. X2 was chosenfor analysis of nominal data and Mann-Whitney U test for analysis of ordinal datasuch as height, weight, stress levels, and jobsatisfaction.

Univariant logistic regression analysis wasundertaken to examine associations betweenthe potential predictors and dependent vari-ables, namely incidence of back pain, wherethe dependent variable was dichotomous (0 =no pain, 1 = pain). Based on the results ofunivariant logistic regression and correlationanalysis, variables were selected to fit a multi-variant logistic model. Relative risks were esti-mated as odds ratios and presented with their95% confidence intervals (95%CI). A variablewith an odds ratio of less than 1 was consid-ered not to be a significant risk factor.

ResultsThe fully completed questionnaires wereobtained from 183 of the 266 potentialrespondents, a response rate of 69%. Thesample was 30 5% of the target population(cardiac technicians) of which 80% of respon-dents reported back pain. Thirty four malesand 149 females responded, of whom 25males (74%) and 121 females (81%) experi-enced back pain.

Lifestyle variables such as smoking (P=0 36), exercise level (P = 0-51), and care ofpreschool children (P = 0-21) did not showany association with risk of back pain. At leastsome limitation of lifestyle from back pain wasreported by 56% of the respondents.

Seventy five per cent of the respondents(137) reported having been involved inechocardiography for more than five years, ofthis group 38% reported more than 11 years ofexperience in the field. The number of years inechocardiography did not appear to be a signif-icant risk factor when X2 analysis was applied(P = 0 27). However, logistic linear regression

24

on April 4, 2020 by guest. P

rotected by copyright.http://heart.bm

j.com/

Heart: first published as 10.1136/hrt.78.S

uppl_1.23 on 1 August 1997. D

ownloaded from

Back pain among echocardiographers

Table 1 Relative risk of back pain with echo experience(n = 183)Echo experience Odds ratio(years) Significance (95%CI)

2-4 0-33 1-8 (0-5-7-02)5-10 0-11 2-5 (0-7-8-6)11-20 0-05 3-2 (0-97-11-5)

analysis showed that compared with respon-dents with less than two years echocardiogra-phy experience, there was a trend towardsincreased risk of back pain with increase inyears of experience. Respondents in the 11-20years group were at three times greater risk (P= 0 05) of back pain than those with less thantwo years' experience (table 1). Similarly, theamount of time spent scanning (echotime) didnot show significant (P = 0-55) associationwith the risk of back pain, although the inten-sity of back pain for sufferers showed a signifi-cant correlation with echotime (P = 0 035)(fig 2), interference with work (P < 0.01), andlimitation of lifestyle (P < 0-01), even whencontrolled for back pain history, age, reducedjob satisfaction, and number of years inechocardiography.For the question regarding job satisfaction,

68% replied good, 31% replied average, andonly 1% replied poor. Mann-Whitney U testsshowed an inverse relation between back painand job satisfaction (P = 0 04). Logisticregression analysis confirmed that for therespondents who rated their job satisfaction asaverage or poor there was a greater risk of backpain (OR = 2-4; 95%CI = 0 96-5-9).

3-6 7-10 11-15 16-20 20-30How much time do you spend on echo examinations?

Frequently

_ e! 8 SiBl | :31 1 ELBIISI s_ #ltli_ E ; , _ l , , rRarely

Figure 3 Frequency of lifting involvedfor the respondents whose jobs required lifting(n = 144). (Lifting, P = 0 07; occasional andfrequent lifting, P = 0 04).

When ergonomic variables were examined,76% replied yes to the question "does your jobrequire lifting?" and 24% replied no.Association between lifting and back painshowed x2 borderline significance (P = 007)but logistic regression analysis showed a signifi-cantly greater risk for respondents who liftedoccasionally, 62% (P = 0-01; OR = 4-8;95%CI = 1-25-18&7), and frequently, 31%(P = 0 04; OR = 4-4; 95%CI = 1.02-18-9),compared with non-lifters (fig 3). Eight percent reported that their jobs rarely requiredthem to lift. X2 analyses did not show any signif-icant relation between back pain and type oflifting (lift patient P = 0-22, lift equipmentP = 04).

Figure 4 shows the frequency distribution foreach chosen method for echocardiographyexamination as derived from the questionnairedata. Method 1 emerged to be the most popular,used by 66% of the respondents. The respon-dents in the method 6 group (machine on theright, transducer in the left hand, and patienton the left), used by 7% of the respondents,emerged to be the group least at risk from backpain (P = 0-008, 95%CI = 0 06-0-67) (table2).When method 1 was compared with method

6, logistic regression analyses showed thatrespondents in the method 1 group were fivetimes at greater risk from back pain thanrespondents in the method 6 group (P =0-008; 95%CI = 1-5-16.4) (table 2). Method5 was not included in the analysis because onlyone respondent indicated the use of thismethod. Analysis on method 4 was incompletebecause it was not clear from the questionnairereplies whether this group required more thanone echocardiographer to conduct the exami-nation. Logistic regression analysis based on asingle operator, however, showed an increase inrisk of back pain (OR = 24-4; P = 0-006;95%CI = 6 5-247).

Descriptive data showed that of the respon-dents with back pain (n = 146), 92% reportedmultiple incidences of back pain, and 63%reported pain occurring once a month or more.Spearman correlation tests showed significantassociation (P < 0-001) between the intensityof back pain and multiple incidences of pain(P < 0.001) and frequency of back pain(P < 0-001). When asked "did you report yourback pain?", 57% replied no, and to the ques-tion "have you ever claimed compensationfrom your employer", 77% replied no, and2-7% replied yes. Previous history of back painwas reported by 21% of the respondents, whichshowed significant (P = 0.001) associationwith back pain when X2 analysis was applied.

For the purpose of this study, upper backand neck pain were combined and included inthe variable "upper back". Low back pain wasreported by 54 1%, upper back pain by 17-8%,and 28- 1% reported both upper and lower backpain. The respondents indicated similaranatomical distribution of back pain in malesand females; upper back, 9% males and 15%females; lower back, 44% males and 43%females; upper and lower back, 20% males and23% females.

3.2

3.0

2.8 -

2.6 -

2.4

:I,.a)c

c

a1)

2.2 -< 3

Figure 2 Relation between the amount of time (hours per week) spent onechocardiography examinations and intensity of back pain.

100 r

C/c

0Q0.C,,U)

0U).0Ez

80 H

60 H

40 -

20 H

o0Occasionally

25

on April 4, 2020 by guest. P

rotected by copyright.http://heart.bm

j.com/

Heart: first published as 10.1136/hrt.78.S

uppl_1.23 on 1 August 1997. D

ownloaded from

Solanki, Carr, Martin

Pain

M YesM No

10

Method 1 Method 2

12 WMethod 3 Method 4 Method 6

Figure 4 Influence of examination methods used on back pain (n = 183).

Table 2 Relative risk for methods usedfor echo examinations (n = 183)

Method Descnption Odds ratio (95%CI)1 machine left, right handed, patient on the right 4 9 (1-49-16-4)2 machine in front, right handed, patient on the right 3-9 (0-69-21-7)3 machine in front, left handed, patient on the right 7 0 (1-07-46-4)4 machine right, right handed, patient on the right 24-4 (6-5-249)5 machine on the left, left handed, patient on the left not computed6 machine on the right, left handed, patient on the left 0-20 (0 06-0 67)

Spearman correlation tests showed highlysignificant correlation between multiple inci-dences of back pain and the following vari-ables: how would you rate your back painwhile performing echocardiography? (P =

0 001); in the last month has your back paininterfered with your work? (P = 0-03); andlocation of back pain (P = 0 026). The inten-sity of back pain showed significant correlationwith the variables: amount of time spent scan-ning per week (P = 0 035); interference withwork (P < 0.001); and limitation of lifestyle(P <s 0 001). Of the 56% of the back pain suf-ferers who reported limitation of lifestyle, 24%reported moderately severe limitation. Whenintensity of back pain while performingechocardiography was examined, 69%reported some interference with work becauseof back pain. Intensity of back pain scores ofmoderately severe and severe while performingechocardiography were reported by 19% and8%, respectively.

Unfortunately, the pattern for time lostbecause of sickness was difficult to establishowing to the complexity of the question, whichwas included in the question "did you reportyour back pain?". A significant number ofrespondents reported taking sick leave but didnot report the number of days.

DiscussionBecause back pain is common in the generalpopulation and the aetiology of back pain ismultifactorial, indication of the risk factors inback pain surveys -can be difficult. When plan-ning for prevention, however, a distinction hasto be made between occupational and non-occupational origins of back pain. Th¢ presentstudy has identified several risk factors affect-ing the echocardiographer, of which the tradi-tional echocardiography examination methodand frequent lifting were the main risks forback pain.

In the present study, 80% of the respon-dents reported back pain. These findings were

similar to the incidence rates from the study ofback pain among radiographers,'7 but signifi-cantly higher than the incidence rates reportedby a study on back pain in the general popula-tion, which reported 22% for males and 26%for females,'8 and even the nursing profession(69%).19

Identification of job satisfaction was clearlyassociated with back pain in the present study.These findings were consistent with otherinvestigations.' 20 Logistic regression analysisshowed a relation between back pain and aver-age or poor job satisfaction, however, the rea-son for this association is less clear cut. It ispossible that a dissatisfied worker would bemore likely to complain of back pain to avoidwork; however, it is quite likely that theechocardiographer may feel under stress tryingto keep pace with the increasing workload.This stress could be further increased if theechocardiographer suffers back pain whilescanning and experiences increased intensityof pain with prolonged periods of scanning (asthe data from this study has shown to be thecase).The most striking findings of this study

have been the influence of occupational andergonomic aspects of echocardiography onback pain. In the present study, the risk ofback pain was shown to increase with thelength of service. For those in the field for fiveor more years the risk was more than twotimes greater, and for those in the field for 11years or more the risk was more than threetimes greater than the group with less thanfour years' experience. These results supportthe findings of a study by Seccombe and Ball5that reported that the proportion of nurses suf-fering from back pain rose from 5% for thosewith five or fewer years' service to 10% forthose with 10 or more years' service. This mayin part be explained by the fact that the greaterthe length of time spent in a high risk profes-sion, such as nursing, the greater the numberof opportunities for injuries to have occurred.These findings were supported by Feldstien etal9 who found that new nurses were injuredmore frequently than veteran nurses, but therewas a higher incidence of re-injury and moredisabling injury among veteran nurses. Thepresent study found a similar pattern to theresults of Feldstein et al,'9 as indicated by thestrong relation between the frequency and theintensity of back pain among the more experi-enced echocardiographers.

Lifting as a predictor of back injury is con-sistent with current thought. The hypothesishas biological sense and is well supported bybasic science research.2' The present study didnot identify lifting as a risk factor per se butfrequency of lifting was significantly associatedwith back pain. These results are consistentwith other investigations where frequent lift-igjbcategory requiring frequent lifting,"l

and lack of ergonomic training in patienthandling3 were identified as the main risk fac-tors for back pain in the nursing profession.Clearly, the provision of lifting training forechocardiographers is indicated by the presentstudy.

120 r

100 [-

80 ;u'

C0)

0)

60 K-

40 H

20 H

o

26

on April 4, 2020 by guest. P

rotected by copyright.http://heart.bm

j.com/

Heart: first published as 10.1136/hrt.78.S

uppl_1.23 on 1 August 1997. D

ownloaded from

Back pain among echocardiographers

Low back pain is widely documentedamong the general and nursing population.The incidence rate of 54% of echocardiogra-phers reporting pain in the lower back onlymay be explained by the link between liftingand lower back pain. The finding, however, of46% of echocardiographers reporting upper

back problems suggests factors other than lift-ing may be involved. Research on repetitivestrain injury suggests that neck strain resultsfrom activities in which the neck is moved up

and down or rotated, and upper back strainresults from improper static postures.8The present study found that operators

using the traditional echocardiography posi-tion (method 1) of machine on their left,transducer in the right hand, and patient on

their right were five times more likely to be atrisk from back pain (table 2). Although theother positions were used by a relativelysmaller number of respondents, the resultsindicate that an examination position is leastlikely to present risk of back pain when theoperator is able to reach the patient and theequipment without having to sit or stand withthe spine out of the neutral position, such as

method 6. The influence of other examinationpositions, although having significant associa-tion with back pain, was less clear cut becauseit was not clear from the questionnaire replieswhether this group required more than one

echocardiographer to conduct examinations.For a single operator, however, there would begreater requirement to adopt awkward pos-tures, thus increasing the risk of back pain by a

factor of 24.Isernhagen8 suggests that postures, espe-

cially those where the spine is out of neutralposition, result in damage of tissue and pro-longed flexion may cause disc protrusion. Headds that although flexion itself is not harmful,unequal stresses on the disc and spine cause

longlasting symptoms. The present study sup-ports Isernhagen's hypothesis because theresults show that at least 73% of echocardio-graphers adopt postures that involve some

degree of flexion of the spine, and 55% ofechocardiographers report having back painfor more than five years.

Intensity of back pain and limitation ofwork or lifestyle has been shown to be an indi-cator of total work disability by previousresearch.22 The results of the present studyshow strong statistical association in sufferersbetween the intensity of back pain and theamount of time spent on echocardiographyexamination, interference with work, and limi-tation of lifestyle. These indicators can bevaluable in targeting interventions aimed atpreventing the worsening of the early stages ofback pain trouble. In the authors' department,back pain has led to a change of examinationmethods, from method 1 to method 6 or avariation of method 6 for staff who were

uncomfortable with a left handed scanningposition. This has resulted in symptomaticrelief from back pain.

STUDY LIMITATIONSThis study design contains some limitations,

including lack of availability of certain infor-mation that could help to focus more specifi-cally on selected risk factors. For example,knowledge of actual age rather than age rangemay have been valuable, as would more spe-cific information on body build and fitnesslevel, rather than using height, weight, and selfreported exercise levels.As with any voluntary survey, the respon-

dents are self selected. The study has thepotential flaw of representing only those indi-viduals with a possible vested interest in thetopic, and it may have overestimated theprevalence of back pain among the target pop-ulation. The questionnaire design has resultedin some ambiguity in interpretation of some ofthe ergonomic data. For example, questionson echocardiography examination methods orpositions could have been ambiguous, andcould have been clarified by the use ofdiagrams instead of closed questions. Never-theless, the design used here has decidedadvantages in terms of expense, time, and flex-ibility, and allowed a complete analysis of alarge group of risk factors.

CONCLUSIONSThis study examined four research questionsto reach the following conclusions:

* 80% of echocardiographers in the sam-ple reported back pain

* Lifestyle and personal characteristics didnot show significant association withback pain

* The traditional echocardiography exami-nation method (method 1; transducerheld in the right hand, patient on theright, machine on the left of the echocar-diographer) was shown to be the mostsignificant risk factor. Frequent liftingwas also a contributory factor to backpain

* 11 or more years in echocardiographywas positively associated with back painand the amount of time spent inechocardiography examination was lin-early related to the intensity ofback pain;56% of echocardiographers reportedsome limitation of lifestyle due to backpain; 69% of respondents reported someinterference with work due to back pain;intensity of back pain scores of moder-ately severe and severe while performingechocardiography were reported by 19%and 8% of echocardiographers, respec-tively.

Any form of prevention depends on reliablesources of information. There is usually noformal source of such information unless asurvey, such as this questionnaire, is deliber-ately made. The present study has focusedonly on back pain. Future studies should covera wider range of basic health, safety, and wel-fare issues such as: risk of injury to other partsof the body (Wrist, shoulders, legs); spacerequirements; and requirements for work withvisual display units. Further research shouldalso focus on preventative measures and opti-mal workstation designs that may serve as aguide to all echocardiography departments.

27

on April 4, 2020 by guest. P

rotected by copyright.http://heart.bm

j.com/

Heart: first published as 10.1136/hrt.78.S

uppl_1.23 on 1 August 1997. D

ownloaded from

Solanki, Carr, Martin

The authors acknowledge the support of the British Society ofEchocardiography for this study, and echocardiographers,physiotherapists, and academic staff for their assistance andfeedback during the pilot study.

1 Owen B. Personal characteristics important to back injury.Rehab Nursing 1986;11: 14-16.

2 Ready A, Boreskie S, Law S, Russell R. Fitness and lifestyleparameters fail to predict back injury in nurses. Can JAppi Physiol 1993;18:80-90.

3 Ryden A, Molgaard C, Bobbitt S, Conway J. Occupationallow back injury in a hospital employee population: anepidemic analysis of multiple risk factors of a high riskoccupational group. Spine 1989;4:315-20.

4 Owen B, Damron C. Personal characteristics and backinjury amongst nurses. Res Nursing Health 1984;7:305-13.

5 Seccombe I, Ball J. Back injured nurses: a profile. A discus-sion paper for the Royal College of Nursing. London:Institute of Manpower Studies, 1992.

6 Videman T, Nurminen T, Tola S, Kuornika I, VanharantaV, Troup J. Low back pain in nurses and some loadingfactors of work. Spine 1984;9:400-4.

7 Deyo R, Tsui-Wu Y. Descriptive epidemiology of low backpain and its related medical care in the United States.Spine 1987;12:264-8.

8 Iserhagen S. Principles of prevention for cumulative trau-mas. In: White L, ed. Occupational medicine; back schoolprogrammes. Vol 7. Philadelphia: Hanley & Belfus, Inc,1992:147-54.

9 McKenzie R. The lumber spine: mechanical diagnosis andtherapy. New Zealand: Spine Publications Limited,1981:1-3.

10 Mandell P, Lipton M, Bernstein G. Low back pain. NewJersey: Slack Incorporated, 1989:1-37.

11 May J, Gale A, Haslegrave C. Investigation of posturesadopted by breast screening radiographers. In: Robertson

S, ed. Contemporary ergonomics. London: Taylor andFrancis, 1995:422-7.

12 Biering-Sorensen F, Thomsen C. Medical, social and occu-pational history as risk indicators for low-back trouble in ageneral population. Spine 1986;11:720-5.

13 Venning P, Walter D, Lawrence W. Personal and jobrelated factors as determinants of incidence of backinjuries among nursing personnel. J Occup Med 1987;29:820-5.

14 Capra P, Mayer T, Gatchell R. Adding psychological scalesto your back pain assessment. J Musculoskelet Med1985;2:41-52.

15 Fairbanks J, Couper J, Davies J, O'Brien D. Owestry lowback pain disability questionnaire. Physiotherapy 1980;60:270-3.

16 Miller S, O'Brien J. Prevalence of back pain among radia-tion therapists. Radiology Technology 1991;62:460-6.

17 Wright D, Witt P. Initial study of back pain among radiog-raphers. Radiology Technology 1993;64:283-9.

18 Wright S, Fisher A, Horsley S, Jayson M. Influence ofphysical, psychological and behavioural factors on con-sultations for back pain. BrJ Rheumatol 1995;34: 156-61.

19 Feldstein A, Valanis B, Voller W, Stevens N, Overton C.The back injury prevention project-pilot study. J OccupMed 1993;35:114-19.

20 Dehlin 0, Stig B, Gunnar B. Effects of physical trainingand ergonomic counselling on the psychological percep-tion ofwork and on the subjective assessment of low backinsufficiency. Scand J Rehab Med 198 1;13: 1-9.

21 Pope M, Lehmann T, Frymore J. Structure and function ofthe lower spine. In: Pope M, Frymore J, Anderson G,eds. Occupational low back pain. New York: Praeger,1984:chapter 7.

22 Rossingnol M, Lortie M, Ledoux E. Comparison of spinalhealth indicators in predicting spinal status in a 1 yearlongitudinal study. Spine 1993;18:54-60.

28

on April 4, 2020 by guest. P

rotected by copyright.http://heart.bm

j.com/

Heart: first published as 10.1136/hrt.78.S

uppl_1.23 on 1 August 1997. D

ownloaded from