historical research report

0 downloads 0 Views 4MB Size Report
use the systems available to them at each colliery (manriding belts and locomotives). A different study design would be required to examine this phenomenon ...
HISTORICAL RESEARCH REPORT Research Report TM/93/05 1993

Case control study of the relations between risk of back pain sickness absence and the nature of tasks carried out by coalminers. Final report on CEC research contract 7280/04/022 Waclawski ER, Hagen S, Symes AM, Graveling RA, Scott AJ, Miller BG

HISTORICAL RESEARCH REPORT Research Report TM/93/05 1993

Case control study of the relations between risk of back pain sickness absence and the nature of tasks carried out by coalminers. Final report on CEC research contract 7280/04/022 Waclawski ER, Hagen S, Symes AM, Graveling RA, Scott AJ, Miller BG

This document is a facsimile of an original copy of the report, which has been scanned as an image, with searchable text. Because the quality of this scanned image is determined by the clarity of the original text pages, there may be variations in the overall appearance of pages within the report. The scanning of this and the other historical reports in the Research Reports series was funded by a grant from the Wellcome Trust. The IOM’s research reports are freely available for download as PDF files from our web site: http://www.iom-world.org/research/libraryentry.php

Copyright © 2006 Institute of Occupational Medicine. No part of this publication may be reproduced, stored or transmitted in any form or by any means without written permission from the IOM

INSTITUTE OF OCCUPATIONAL MEDICINE Research Avenue North, Riccarton, Edinburgh, EH14 4AP Tel: +44 (0)870 850 5131 Fax: +44 (0)870 850 5132 e-mail [email protected]

ii

Research Report TM/93/05

A case control study of the relations between risk of back pain sickness absence and the nature of tasks carried out by coalminers

ER Waclawski, S Hagen, AM Symes, RA Graveling, AJ Scott, BG Miller

November 1993 IOM Report TM / 93/05

Report No. TM/93/05 CEC Contract 7280/04/022

INSTITUTE OF OCCUPATIONAL MEDICINE

A CASE CONTROL STUDY OF THE RELATIONS BETWEEN RISK OF BACK PAIN SICKNESS ABSENCE AND THE NATURE OF TASKS CARRIED OUT BY COALMINERS Final Report on CEC Contract 7280/04/022

by ER Waclawski, S Hagen, AM Symes, RA Graveling, AJ Scott, BG Miller

Institute of Occupational Medicine 8 Roxburgh Place Edinburgh EH8 9SU

Tel 031 667 5131 Fax 031 667 0136

November 1993

This report is one of a series of Technical Memoranda (TM) distributed by the Institute of Occupational Medicine. Current and earlier lists of these reports and of other Institute publications, are available from the Technical Information Officer/Librarian.

Report No. TM/93/05 INSTITUTE OF OCCUPATIONAL MEDICINE A CASE CONTROL STUDY OF THE RELATIONS BETWEEN RISK OF BACK PAIN SICKNESS ABSENCE AND THE NATURE OF TASKS CARRIED OUT BY COALMINERS Final Report on CEC Contract 7280/04/022 by

ER Waclawski, S Hagen, AM Symes, RA Graveling, AJ Scott, BG Miller

SUMMARY This case control study investigated the differences between tasks performed by underground coalminers at two collieries in the United Kingdom and their history of back pain sickness absence during a six month period. Sickness absence data was provided by British Coal Medical Services and used for case selection. A case was defined as a miner with one or more spells of at least seven days absence due to a defined back complaint during the period 1.8.91 to 31.1.92. The control miners were selected in a random manner from colliery staff lists separated into ten year age bands in a ratio of two controls to one case. A questionnaire was developed as the method of identifying the tasks which each miner had performed. Ergonomists visited both collieries and observed the work performed underground. These observations were used to identify the tasks which occurred underground. These tasks were then incorporated into a questionnaire which asked about lifting, carrying, holding, pushing, pulling and use of transport systems underground. In addition, a shortened version of a back pain questionnaire previously used by the Institute in research on coalminers was included. The task component questionnaire was the subject of a small validation exercise at one colliery. This exercise indicated that cases reported completing tasks that they were observed performing more commonly than controls. In addition, the estimates of frequency of performing a task, total duration that a task was performed and the longest duration without a break that a task was performed were more accurate for cases than controls. A questionnaire administrator was trained in the method of administration of the questionnaire. The questionnaire administration took place in the medical centre at each colliery.

The study recruited 104 cases and 215 controls. The mean age of the population was 37 years. The mean duration of employment in the coal industry was 17 years. Participants at Colliery 1 were older with more years experience than participants at Colliery 2. A large number of redundancies occurred at Colliery 2 during this study with the loss of older workers. The average length of back pain absence was 16.7 days. The most common reason for absence was back pain or strain, followed by back injury. A smaller number were certified as having lumbago or sciatica. Self-reported back pain in the preceding twelve months was experienced by all cases and 78% of controls. Eighty percent of cases and 72% of controls had had attacks of back pain in the period more than 12 months before the study. Chronic back ache or pain (occurring on most days of every month) was present in 46% of cases and 24% of controls. Forty one percent of cases had more than one attack of back pain in the preceding year compared to 42% of controls. The questionnaire responses were analysed by logistic regression techniques. Individual positive associations were found between increased risk of back-related sickness absence and increases in (1) lifting more than 50 kg with a frequency of more than 20 times in a shift; (2) holding weights above shoulder height more than 50 times in a shift; (3) carrying more than 50 kg; and (4) twisting the body while pushing more than 50 times in a shift. A negative association was found with time spent driving free steer vehicles underground. In statistically derived models fitted to assess the effect of multiple risk factors, the only risk factors which showed significant effects were driving free steer vehicles (negative) and lifting more than 50 kg (positive). This study has identified that frequency of heavy lifting by underground miners is associated with an increased risk of absence due to a back problem. A variety of risk reduction methods can be implemented to attempt to improve this situation. Research is recommended to validate the task questionnaire further and also to study a more active medical management of back pain among coalminers.

CONTENTS Page No SUMMARY 1.

INTRODUCTION

1

1.2

Review of Relevant Literature

2

1.2.1 1.2.2 1.2.3 1.2.4

2 2 3 4

Back pain and coalmining Back pain and heavy manual work Other risk factors for back pain Summary

2.

STUDY AIMS

5

3.

METHODS

7

3.1

Study Design

7

3.1.1

7

4.

Definition of cases and controls

3.2

Questionnaire Development

7

3.3

Training of Questionnaire Administrator

8

3.4

Questionnaire Administration

8

3.5

Questionnaire Validation

8

3.6

Statistical Analysis

9

3.6.1 3.6.2

9 9

Descriptions Multiple logistic regression

RESULTS

13

4.1

Response Rate

13

4.2

Population Characteristics

13

4.2.1 4.2.2 4.2.3 4.2.4 4.2.5

13 13 14 14 15

4.3

4.4

Age distribution Occupation Sickness absence Responses to questions on back pain Non participation

Results of Main Logistic Regression

16

4.3.1 4.3.2 4.3.3

16 16

Introduction Individual tasks Correlation between tasks, and a proposal model

Summary

18 19

Page No 5.

DISCUSSION

21

5.1

Introduction

21

5.2

The Effects of Chance, Bias and Confounding

22

5.3

Risk Reduction

22

5.3.1 5.3.2

22

5.3.3 5.4

Introduction The reduction of injury risk from specific factors General risk reduction

Further Research

23 25 28

6.

CONCLUSIONS

29

7.

ACKNOWLEDGEMENTS

31

8.

REFERENCES

33

FIGURES TABLES APPENDICES

1. INTRODUCTION Low back pain is a common complaint. Within the population up to 80% of all people will experience back pain at some time. Among workers it is one of the most frequent and disabling conditions. Low back pain can develop gradually or suddenly, with or without association to an initiating event. A variety of classifications exist based on aetiological factors, or symptoms and clinical findings. Low back pain can be classified according to the duration of the symptom as either acute, where the condition resolves, or as chronic, where the complaint persists. Individuals who have acute low back pain do risk a recurrence of their problem. In addition, a small percentage of individuals who have an acute onset have symptoms which continue for more than a year (Riihamaki, 1991). The past history of back pain can be used to determine the risk of recurrence of back pain in the future. Important aspects of this history include previous sickness absence and its duration, the number of previous episodes and residual symptoms, particularly in the leg, after the acute phase of the attack has subsided (Troup, 1984). A definite structural diagnosis can only be reached in up to half of cases. Common diagnostic terms used include lumbosacral strain and sprain, postural low back pain, muscular insufficiency, sacroiliitis, annular tears, prolapsed disc, degenerative spinal disease or nonspecific low back pain (Garg and Moore, 1992). Low back pain is a result of the stimulation of pain receptors which exist in the skin, subcutaneous and adipose tissues; the fascial, aponeuroses and ligaments; the vertebral periosteum and marrow; the adventitial sheaths of blood vessels and the fibrous capsules of the lumbar facet and sacroiliac joints. These receptors can be activated by mechanical stress or exposure to irritating chemical substances in the surrounding tissues that are released from traumatised, inflamed or metabolically abnormal tissues. Low back pain can originate from nearly all the tissues in the lumbosacral region (Riihimaki, 1991). Mechanical and traumatic forces during work can produce stimulation of pain receptors leading to back pain. Back pain is a major cause of illness and sickness absence at work in the coalmining industry. It is believed within the industry that the nature of the work is a major contributory factor to this high incidence. However, it is not clear which aspects of the work are mainly responsible. While some features of conditions underground which may predispose to back pain, such as roof height, can be dictated by engineering considerations, other task elements could be associated with increased risk, and, once identified, might be amenable to work redesign. A previous questionnaire development loads on the

investigation by the IOM has enabled the development of a on back pain, suitable for epidemiological studies, and the preliminary of, methods of task analysis suitable for quantifying movements and back (Agius et al. 1988).

Ideally, identification and recording of task components should be carried out by direct observation and recording by trained observers. However, although this may be effective in a highly repetitive production cycle, the previous work (Agius et al. 1988) demonstrated that the variation in underground activities made it difficult to observe an adequate sample of activities without incurring prohibitive

costs. Comparisons on the basis of small scale sampling identified considerable between and within subject variability at a single work location (Agius et al. 1988). Earlier IOM research into other musculoskeletal disorders (English et al. 1988) adopted an approach of formulating a detailed questionnaire on individual elements of range and frequency of movement, loading etc. supported by exemplary material to aid recognition of the element in question, and possible recall. A case-control study of people with and without various upper-limb strain injuries, showed that such an approach was successful in identifying high risk task elements (English et al. 1989). This led us to believe that such an approach might have a good chance of success in identifying high risk task elements related to back pain. 1.2

Review of Relevant Literature on Occupational Risk Factors for Low Back Pain

1.2.1

Back pain and coalmining

Low back pain is recognised to be a particular problem In a study of sickness absence due to back lesions in found that 14.8% of miners were absent due to back period. Back pain accounted for 11.9% of working sickness absence.

in the coalmining industry. coalminers, Afacan (1982) lesions during a 12 month days lost due to certified

Lawrence and Aitken-Swan (1952) studied rheumatism in miners and non-miners and showed that disc disorders were twice as common in miners aged 30-60 years as in non-miners. The study indicated that disc disorders were more common in miners than non-miners as a cause of absence of three months or more duration in 1945-49. The study demonstrated an earlier onset of, and greater incapacity from, lower dorsal and lumbar disc degeneration as the main feature of rheumatism in miners. A study of miners aged 41-50 years (Kellgren and Lawrence, 1952) demonstrated significantly more radiological changes of lumbar disc degeneration in miners, compared with office and manual workers. The radiological changes observed correlated with complaints of low back pain in the previous 5 years. Lawrence (1955) performed a study to identify factors in mining which are responsible for the observed increase in disc degeneration. Incapacity for work became progressively greater as the duration of heavy lifting increased. Those who lifted for more than 30 years had four times the incapacity of those who lifted for less than 25 years. Those working in wet conditions for more than five years had twice the incidence of pain compared to those who had always worked in dry conditions. Miners who had worked at the coalface for over 30 years had four times the prolonged disability (3 months or more absence) of other miners. The severity and frequency of disc changes were also related to prolonged stooping. Agius et al. (1988) compared the prevalence of back pain in miners with clerical workers and found that the prevalence was similar up to the age of 40 years. However, in workers aged over 40 years, the prevalence of back pain was greater in miners than clerical workers. 1.2.2

Back disorders and heavy manual work

Many studies have described a higher prevalence of back disorders in jobs involving heavy manual work. The risk of hospitalisation for lumbar disc herniation was higher in blue collar workers in industry and motor vehicle drivers than in professional groups in a study by Heliovaara (1987). Riihimaki et al. (1989)

found that machine operators had more sciatic pain than office workers. Independent effects of occupation, age, back accidents and twisted or bent postures at work on the occurrence of sciatic pain were identified. Wickstrflm, Hanninen, Lehtinen and Riihima'ki (1978) studied back problems in concrete reinforcement workers and compared the results to computer technicians. A history of sciatica was more common in reinforcement workers than computer technicians. Though restricted lumbar flexion was equally common in both groups, the occurrence of pain during forward flexion was more common among reinforcement workers (WickstrOm. Nummi and Nurminen, 1978). Riihima'ki et al. (1990) compared concrete reinforcement workers to housepainters and found a significantly greater prevalence of degenerative changes to the spine in the former. Disc space narrowing occurred on average about 10 years earlier in concrete reinforcement workers compared to housepainters. In multivariate logistic regression analysis age and occupation had significant effects on disc space narrowing. A history of earlier back accidents was not a significant predictor. Earlier work (Riihima'ki 1985) indicated that back accidents and age were related to sciatic pain but occupation was not so related. Riihima'ki et al. (1989) showed that the risk of having sciatic pain was increased in concrete reinforcement workers. Wickstrom et al. (1985) identified that reinforcement workers work in stooped positions and perform heavy lifting more often than painters. 1.2.3

Other risk factors for back disorders

Brendstrup and Biering-Sorensen (1987) studied fork-lift truck (FLT) drivers. In the year preceding the study 65% of FLT drivers had low-back trouble compared to only 47% of non-FLT working men. The frequency of absence due to low back trouble was also higher. Walsh, Cruddas and Coggon (1991) identified heavy lifting and digging as being associated with onset of low back pain in men in a recent population based study. A multitude of other risk factors for low back pain have been identified. These include smoking (Deyo, 1989; Biering-Sorenson and Thomsen, 1986; Frymoyer, 1983; Kelsey et al. 1984); obesity (Deyo, 1989); history of back injury (Daltroy et al. 1991; Riihimaki et al. 1989); emotionally stressful occupation (Frymoyer et al. 1983); sporting activities (Frymoyer et al. 1983) and marital status (Reisbord and Greenland, 1985). Studies have been carried out to investigate physical workplace risk factors for back pain. Damkot et al. (1984) surveyed a population of men aged 18-55 years and identified men with severe low back pain, moderate low back pain and those with no pain. The three groups were compared for physical workplace factors by self-report. Significant differences were reported for lifting technique but frequency of lifting and amount of weight lifted were not significantly different. Low back pain sufferers were more likely to reach with arms fully extended, spend more time standing and push objects at work. Videman et al. (1984) compared qualified nurses and nursing assistants for prevalence of low back pain and physical work load. Nursing assistants had a higher prevalence of low back pain and reported twice as much lifting, bending and rotation at work as qualified nurses. Svensson and Andersson, (1983) found that a high degree of lifting correlated with the occurrence of low back pain in a

population based study, and in a similar study by Frymoyer et al. (1980) lifting, carrying, pulling, pushing and twisting were significantly related to reported low back pain. Maeda et al. (1980) found back pain to be more frequent in strawberry farmers than egg plant farmers, relating this to measured differences in time spent bending. In a case control study Punnett et al. (1991) measured individual exposures to non-neutral postures (forward trunk flexion, trunk twist and lateral bend) in workers in a car assembly plant and found that the risk of back disorders increased with duration of exposure and exposure to multiple postures. Magora (1973) implicated 'sudden maximal physical effort' in the aetiology of back pain. Driving motor vehicles has also received attention. Time spent driving cars (Kelsey and Hardy, 1975; Biering-Sdrensen and Thomson, 1986); truck driving (Frymoyer, 1980; Kelsey and Hardy, 1975); driving earth movers (Dupuis, 1987); driving cranes (Bongers, 1988) and driving subway trains (Johanning, 1991) have all been associated with an increased occurrence of back problems. Whether this is an effect of whole body vibration or duration of sitting posture has not been determined. Agius et al. (1988) compared the prevalence of low back pain in miners in two coalmines which used different transport systems but did not uncover a significant difference. Trauma produced by accidental injuries can lead to back pain. The majority of cases of low back pain are not preceded by an accident. Where an accident occurs the prognosis for the individual is worse than for others with low back pain. Such injuries include those caused by handling and underfoot accidents (Riihimaki, 1991). ^ 1.2.4 Summary Lumbar disc degeneration in miners has been related to the amount of heavy lifting performed and to working with stooped postures. In other industries manual workers have increased back problems which are associated with work factors such as heavy lifting, twisting or bent postures. Other factors noted from population studies include carrying, pulling, pushing and digging. Driving in a variety of vehicles has also been associated with back pain. Accidents can precipitate low back pain.

2. AIMS The aim was to identify defined components of tasks at work and to quantify how these were associated with risk of back pain in miners. The specific research questions were: 1. What components of tasks are associated with an increased risk of back pain? 2. What are the relative risks associated with performance of these task components? 3. In what way can selected tasks be modified to reduce these high risk components and reduce the risk of back pain?

3. METHODS 3.1

Study Design

It was known that the Medical Service of British Coal kept good records of sickness absence. It was therefore possible to design a case-control study, where recent absence due to back pain defined cases, and controls were chosen from men without such absence. Two modern collieries were chosen to provide a population from which to draw the study sample. The cooperation of medical staff at these collieries was recruited, to provide assistance in selecting the sample. 3.1.1.

Definition of cases and controls

Absences of seven days or more require medical certification and are associated with a medical diagnosis. We restricted attention to the diagnoses on medical certificates:

therefore following

( i) back pain/strain ( ii) back pain or ache (unspecified) ( i i i ) lumbago etc. (including sciatica) ( iv) displacement of (prolapsed) intervertebral disc ( v) arthritis of spine ( vi) injury to back

A case was defined as an underground worker from either of the selected collieries who had at least one spell of certificated absence in the period 1.8.91 to 31.1.92 with one of the above diagnoses. A person whose spell commenced prior to 1.8.91 but extended into the study period was not used as a case. A person whose spell commenced in the study period but did not end before 31.1.92 was used as a case. Two controls from the same colliery were selected for each case; they were matched by age (in ten year bands) from colliery staff lists. It was possible for a control to have experienced sickness absence due to one of the above diagnoses, in the appropriate period, but for less than 7 days duration. 3.2

Questionnaire Development

The information obtained from the ergonomic surveys (Appendix 1) was used to develop a questionnaire (Appendix II). Each section contained questions concerning specific task components developed from ergonomic observations and relevant previous literature. Components of tasks performed by underground coalminers were identified and grouped into sections on lifting, holding, carrying, pushing, pulling and use of transport systems underground. The first question in each task section asked if the task was performed during a typical shift. If a positive response was obtained supplementary questions were asked about frequency, about the longest time without a break that the task was performed, and about the total amount of time in a shift that the task was performed. In order to ensure that the nature of a task was considered to be understood equally by all subjects, photographs of tasks were shown (Appendix III). Answers to questions on frequency, longest duration and total time were obtained from specific ranges identified on cards by the

questionnaire administrator (Appendix IV). A shortened version of the IOM back pain questionnaire (Agius et al. 1988) was administered to all subjects after the task component questionnaire. The IOM back pain questionnaire contained questions relating to individuals' experience of back pain, both within the last year and before this. Questions tried to ascertain whether back pain was chronic, that is occurring on most days of every month or, if not, how many separate attacks occurred. The date of the first attack of back pain was also established. 3.3

Training of Questionnaire Administrator

Initial training was performed at the Institute of Occupational Medicine. The questionnaire was reviewed item by item. This included detailed question specifications, mock interviewing and role play. Assessment of interviewing techniques was made during initial pilot work in July 1991 and subsequently at regular intervals during the period of questionnaire administration. No changes in administration technique were observed.

3.4

Questionnaire Administration

The questionnaire was administered at the Medical Centres of each colliery between August 1991 and July 1992. An office, which was used for no other purpose during the administration of each questionnaire, ensured a private, confidential environment. Following an explanation about the purpose of the study the personal details from each subject (date of birth, years in current job). Individual cases were questioned about task in the three month period prior to absence. Controls were preceding three months prior to interview. 3.5

administrator obtained industry and years in components performed asked to consider the

Questionnaire Validation

A small validation exercise was performed at Colliery 1 with the aim of comparing the responses obtained by questionnaire with the observed frequency with which tasks were performed. An ergonomist observed a sample of 10 men. The men observed were known to have completed the questionnaire (either as a case or a control). The ergonomist was unaware of the status within the study (Case/control) of the men he observed. Each man was observed for a full shift. Three occupational groups were studied - heading crew, locomotive workers and fitters. The observation involved breaking tasks down into components and recording the frequency and durations of these task components. The ergonomist concentrated on lifting, pulling, pushing and use of transport underground. A proforma was developed for the transfer of information which could be summarised and subjected to analysis. A comparison of observational and questionnaire data was carried out to assess individuals' accuracy in recording details of task components carried out. Observational data were classified into the same categories as those used in the questionnaire, and cross tabulations were produced allowing description of the agreement achieved. Cases were observed performing

tasks that they had reported more commonly than controls. Cases and controls tended to overestimate the frequency and durations of a task with the controls overestimating to a larger degree than the cases. The questionnaire validation results are presented in detail in Appendix V. 3.6

Statistical Analysis

The analysis was carried out in two main stages; (1) descriptions of the study group in the context of the whole population, and of respondents' characteristics, (2) examination of relationships between task components and outcome using formal regression analysis. In addition power calculations were carried out to inform the interpretation of results. 3.6.1

Descriptions

(a) The interview response rate was described, for both cases and controls. The group of non-participants was examined for any systematic bias in age or sickness diagnosis. (b) A general description of the stratified case-control design was produced, including the distribution of cases and controls according to age and colliery. (c) The characteristics of cases and controls were described. This comprised frequency distributions of age, occupation, years in industry, sickness diagnosis and other variables of interest, with means and standard deviations for continuous variables, separately for each colliery. (d) Numbers of individuals were tabulated by type of job and by the different task components in which they were involved, in order to identify task components which were job specific and those which were common to all. (e) Cases and controls were matched by colliery and by 10 year age band, approximately in the ratio 1:2. The approach taken in analysis was that for category matched data, which treated the data as having a pool of cases and a pool of controls within each stratum. This gives a more efficient analysis than if the data had been considered to be individually matched, which in effect would over match by artificially linking a particular case and control (Kleinbaum, Kupper and Morgenstern, 1982). (f) Tables of numbers by case/control status, colliery and age, for each task component were produced, and empirical odds ratios within stratum calculated to inform interpretation of subsequent logistic regression analysis. (g) The responses to questions in section G of the questionnaire, on back pain history, were described for cases and controls. This included the incidence of "acute" and "chronic" back pain (as identified by responses to question 2 of section G) and the length of time since first attack. 3.6.2

Multiple logistic regression

General strategy. The outcome of interest in this study was a binary one, that is case-control status, and the purpose of analysis was to relate this to task components. Multiple logistic regression analysis was used as this is a suitable technique for this type of problem. The goodness of fit of a logistic regression

10

model to a data set can be summarised in a statistic known as the deviance, and different models can therefore be compared by comparing their deviances. In all models, terms were fitted for age group, colliery and the interaction between the two (which define the structure of the data), and the effect on the response of carrying out different task components was assessed. Relationships between the response and job type were investigated separately; job type was so highly correlated with individual tasks that simultaneous assessment would have given no information about tasks. Details of the application of logistic regression to case-control data are given in Hosmer and Lemeshow (1989). With such a large number of questionnaire responses to relate to the outcome (each task component having a frequency, duration without a break and total duration within a shift associated with it), several approaches to finding a suitable model were used within the regression analysis. (Questionnaire data from 11 surface workers were excluded at this stage since the study only proposed to look at tasks performed underground). Presentation of results. The change in deviance produced when a new variable is added to the model measures the improvement in goodness of fit brought about by addition of the variable. The change in deviance is distributed as chi-squared under the assumptions of the model and provides a significance test of the variable being added. Change in deviance and regression coefficients were presented, along with t statistics to show the direction and size of the effect of the variable added. To summarise findings, regression coefficients were converted to odds ratios and presented with 95% confidence intervals. For questions relating to tasks carried out, in general, respondents were given a choice of categories into which to place their response, as opposed to providing the exact quantity. For example, 20 to 50 times per shift, or 50 to 100 times per shift. As a result most of the explanatory variables to be assessed in relation to case-control status were of an interval nature. Regression modelling allowed assessment of differences in risk between the different intervals, and identification of trends of increasing risk across intervals for a given task. Modelling approaches (i) Firstly, all task components were considered for inclusion on the basis of whether the task was done or not. All components (excluding those associated with transport, which were examined separately) were entered for selection in a forward step wise analysis, in order to form an impression of those which would be important. This method gives all task variables an equal chance of being added to the baseline model (in which age and colliery had already been accounted for) and the one giving the biggest improvement to the model, in terms of explaining most variation, is selected. Then the remaining variables are reassessed and the selection process repeats. (ii) Variables were formed which combined the information on whether a task was carried out at all, and the subsequent information on frequency, longest and total duration in a shift. For example, one variable would indicate whether the frequency of a task was:-

11

not done at all carried out 100 times per shift Similar variables were formed for longest and total duration in a shift. There would be three of these ordered categorical variables associated with each task component. In some cases several categories within a variable were combined since the numbers were very small. This was sometimes the case with the higher frequency categories which were rarely used. Each variable was added in turn to a model in which age and colliery were already accounted for, and its association with case control status assessed. It was also of interest to examine the pattern of risk across the categories, within each variable. We might expect that the risk of being a case would increase on moving from the lowest to the highest categories. Thus, initially, a test for an increasing linear trend in risk was applied. In addition, odds ratios associated with each category were examined to assess the pattern of risk, paying particular attention to the number of individuals contributing information to each category. Also for each of the variables tables were constructed showing the proportion of cases and controls within each of the categories, allowing the investigation of any trends. (iii) Interactions between the individual tasks identified as being related to case/control status were investigated. Tasks may not be independent; they may be linked as part of a larger task or particular job. This was assessed by examining the effect, on one association, of adding another task into the model. (iv) Those task variables identified as being individually associated with the outcome were entered in a stepwise analysis with a view to producing a model containing the tasks which, in combination, were most important. All analyses were carried out using the GENSTAT statistical program (Genstat 5 Committee, 1987).

12

13

4. RESULTS 4.1 Response Rates At colliery 1 the selection procedure resulted in 56 cases being recruited. One hundred and twenty controls were available for the main study. Eighty-two percent (46) of cases and 88% (106) of controls subsequently attended for interview. At colliery 2 the selection procedures resulted in the recruitment of 93 cases to main study. Five of these 93 miners had initially been chosen as controls became cases prior to interview. This left 181 suitable controls available interview. Sixty-two percent (58) of cases and 60% (109) of controls attended interview.

the but for for

In total, three hundred and nineteen men were interviewed in this study; cases and 215 controls, between the two collieries.

104

4.2 Population Characteristics A description is given here of some of the subjects' characteristics of interest, relative to case/control status and colliery. Information was available on each man's age, years worked in current job, years worked in the industry, and job title. The diagnosis given on sickness absence certificates was available for cases, and for controls if they were off sick for a back pain reason. 4.2.1

Age distribution

The distribution of cases and controls to five year age bands is shown in Table 4.1. The mean age of the population was 37 years, and on average an individual had worked 11 years in his current job and 17 years in the coal industry (table 4.2). Cases had not spent more time in their current job or in the industry than controls. Participants at colliery 2 were younger than at colliery 1, with correspondingly fewer years experience. A likely explanation is the large number of recent redundancies at this colliery which would have resulted in the preferential loss of older workers, nearing retirement age. 4.2.2

Occupation

The largest group of workers interviewed was underground electricians and fitters, followed by face workers and then pitbottom workers (table 4.3). In colliery 1 the majority of cases occurred in face and heading workers, whereas at colliery 2 men working elsewhere underground, eg. electricians and fitters, were often cases, as were deputies and pit bottom workers (eg. beltmen, byworkers). Frequency of reporting of task components by job group. Most tasks associated with lifting were common to a large proportion of the workers surveyed across the different job types. One of the least common seemed to be lifting objects weighing more than 50kg. The proportion of workers performing this task ranged from 42% of general underground workers to 66% of underground transport workers (table 4.4). A relatively small proportion of underground transport workers were involved in lifting while kneeling.

14

Holding objects weighing more than 50 kg was generally less common among workers than other holding related tasks (table 4.5); underground transport workers less often reported holding away from the body. Carrying objects weighing 20 kg with one hand was common across job groups (table 4.6). Carrying objects weighing more than 50 kg was common to face and development workers but less so to other types of workers. Carrying while stooped and carrying away from the body were uncommon tasks except in face, development and underground transport workers. The proportion of individuals who recorded pushing while stooped was high across all job groups (table 4.7). Development work was noticeably associated with most pushing tasks, in particular twisting the body while pushing. Face workers and underground tradesmen also often reported pushing above shoulder height, and pushing in conjunction with overreaching or twisting. The pattern of pulling tasks across job groups was very similar to that for pushing (table 4.8). In the area of transport there were differences between methods used at the two collieries. At Colliery 1 the main method by which the workers travelled was man-riding belt, while at Colliery 2 it was by paddy. From table 4.9 it was noted for man-riding belt and paddy the proportions across job group were very similar. At Colliery 2 FSVs and Becorits were used for transport of materials, but at Colliery 1 these were not usually necessary. The use of FSVs and Becorits were restricted mainly to underground transport and general underground workers, while heading machines were operated by development and face workers. 4.2.3

Sickness absence

The most common reason for sickness absence, relating to the back, was back pain or strain, followed by back injury (table 4.10) A smaller number reported lumbago or sciatica. The incidence of back injury among cases was slightly higher at colliery 2 than at colliery 1 while the incidence of back pain/strain at colliery 2 was lower. No cases of arthritis of the spine were noted. Five individuals selected as controls had a few days self-certificated absence during the study period (due to backpain), but were, by definition, still eligible as controls. The remaining controls had no sickness absence associated with a back problem. The average length of sickness absence (related to the back) for cases was 16.7 days; the mean for cases at colliery 2 was 17.2 days and at colliery 1 16.2 days (table 4.10). At both collieries back pain or strain resulted, on average, in the longest absences. On average cases at colliery 2 with back pain/strain or lumbago/ sciatica had slightly longer spells of sickness absence than those at colliery 1. Examining the type of back pain in relation to the occupation of the individual was difficult because of the small numbers involved (table 4.11). Deputies appeared more susceptible to lumbago/sciatica than other workers. 4.2.4

Responses to questions on back pain

All cases said they had experienced back pain in the last 12 months (table 4.12). A large proportion (78%) of controls also reported having back pain. For the period more than 12 months before the study, 80% of cases and 72% of controls said that they had had attacks of back pain (table 4.13)

15

Back pain or ache during the last 12 months. Of the 104 cases, 46% (48) indicated that they had ache or pain on most days during the last year, compared with 24% (40) of controls (1 missing value) who said they had experienced backache or pain (Table 4.14). Seventy seven percent (43) of the remaining 56 cases had had more than one separate 'attack1. Seventy two percent (89) of a possible 123 controls had had more than one 'attack' during the last year (2 missing values). The median number of 'attacks' for cases was 4 (range 1-30) compared with 6 for controls (range 1-100). Cases were more likely than controls to respond as if they had chronic backache or pain. Controls were prone to more separate 'attacks' than cases; cases may however have experienced attacks of longer duration or severity. These two results seem to be consistent. Back pain or ache more than 12 months ago. Eighty percent (83) of cases had also experienced attacks of pain or ache in their back more than 12 months ago. Seventy two percent (154) of all controls had had an attack more than a year ago; 82% (136) of those controls who had had attacks within the 12 months, and 37% (18) of controls who had not, reported having experienced back pain more than 12 months ago. First attack of back pain. Respondents were asked to supply, if they could remember, the month and year of their first attack of back pain. One hundred and eighty six respondents out of a possible 237 remembered the year; only 24 could recall both month and year. Time since first attack of back pain was calculated using year only. The means and standard errors for the case and control groups, are presented in Table 4.15, and distributions for cases and controls are presented in Figure 4.1. There would appear to be no real difference between cases and controls with respect to length of time since the onset of back problems. 4.2.5

Non-participation

It was of interest to examine any differences between those individuals chosen for the study who participated, and those who did not. After being selected for the study there were a number of reasons why individuals may not have attended for interview. Since individuals had a choice as to whether to attend or not some self-selection may have occurred. Perhaps those with particular experience of back pain, or those in a certain job or age group were more likely to attend. These possibilities were investigated. Practical reasons, such as cessation of work at the colliery or the occurrence of sickness absence, may also have intervened. At colliery 2 there were many non-attenders due to redundancies taking place at the time. At the end of the study, 10 of the identified cases had not been interviewed at colliery 1, and 35 at colliery 2. All cases have, by definition, experienced some back problem. The distribution of cases who did and did not participate in the different diagnostic groups is presented in table 4.16. The proportions in each category are similar for the two groups. A Chi squared test (X2 = 0.19) suggested no association between the type of back problem and whether individuals participated or not.

16

In addition, the age distributions of participants and non-participants were compared, for cases and controls (tables 4.17 and 4.18). No difference in age was found between cases who were interviewed and cases who were not at either colliery (X2 = 0.77, X2 = 2.88 respectively). There was evidence that, at colliery 2, controls chosen for the study who did not attend were older than those who did. A likely explanation for this is the large number of older workers who were made redundant at colliery 2 at the time of the study. 4.3

Results of the Main Logistic Regression Analysis

4.3.1

Introduction

The number of cases and controls performing task components by colliery is presented in Table 4.19 for lifting tasks, Table 4.20 for holding tasks, Table 4.21 for carrying tasks, Table 4.22 for pushing tasks, Table 4.23 for pulling tasks and Table 4.24 for use of underground transport and heading machines. Some grouping of response categories for particular tasks was necessary in order to avoid problems with small numbers in cells. The resulting new categories are, however, apparent in the associated results tables. 4.3.2

Individual tasks

Certain tasks were with which these lifting tasks, Table for pushing tasks vehicles.

identified as associated with a risk of back pain. The frequency specific tasks were performed are presented in Table 4.25 for 4.26 for holding tasks, Table 4.27 for carrying tasks, Table 4.28 and Table 4.29 to 4.31 for aspects of travel on free steer

Lifting tasks. Two tasks associated with lifting were identified as contributing to the risk of back pain. The first was 'lifting an object weighing more than 50 kg', which was common to 56% of underground workers surveyed (Tables 4.19, 4.25). Certain frequencies of performing the task were identified as being more risky than others (Table 4.32). In particular, individuals lifting an object heavier than 50kg more than 20 times per shift were 1.6 times more likely to be cases than those who did not lift such objects. Although no significant linear trend was identified a pattern of increasing odds ratios across categories was apparent. This suggests a general increase in the risk of casehood with increasing frequency of performing this task. The second task identified was concerned with twisting the body while lifting objects (table 4.32). This activity was common to 95% of respondents (tables 4.19, 4.25) that is, only a small number of underground workers (15; 8 cases and 7 controls) said they did not twist when lifting. This suggests that there would be little information within this task on which to discriminate between cases and controls. The fact that this task produced a significant result should be interpreted with care. The regression coefficients, and consequently the odds ratios, suggest a large decrease in risk between the first two categories ie. 'not twisting and lifting 1 and "twisting and lifting < 10 times per shift'. This arises because there are equal (but small) numbers of cases and controls in the first category, but a much larger proportion of controls (82%) in the second. The apparently large difference in risk between these two is the basis for the significant finding; excluding the first category, the risk across the remaining categories shows only small random fluctuations. This finding is likely to represent a quirk of the data. In any case

17

this task could not be interpreted as a positive risk factor for back pain. Holding tasks. One aspect of holding was shown to have an effect on the risk of being a case; holding objects above shoulder height. Seventy seven percent of respondents reported performing this task component (Tables 4.20, 4.26). Risk did increase with the frequency of repetition, as indicated by the significant term for linear trend (table 4.32). It is useful to examine the odds ratios associated with the different categories. These suggest no real increase in risk between not holding above shoulder height and doing so up to 50 times per shift. The risk for those holding above shoulder height more often than 50 times per shift was estimated at about five times that for those never doing so. Carrying tasks. Carrying objects weighing more than 50 kg, which was common to 56% of those surveyed, (Tables 4.21, 4.27) was identified as being associated with back pain problems; it is probably directly related to lifting objects of more than 50 kg. An increasing linear trend in risk with increasing number of times this task was carried out per shift was identified (Table 4.32). This is borne out by the increasing odds ratios across categories. It is noticeable that the risk of casehood is raised slightly for frequencies of up to 50 times per shift, but that above this the risk is estimated at about five times greater than for those not performing the task at all. Pushing tasks. Twisting the body when pushing an object was found to have an effect on the risk of back pain; 64% of those surveyed reported doing this task (Tables 4.22, 4.28). In particular there was evidence that the risk increased linearly with the number of repetitions per shift (table 4.32). The estimated odds ratio increased with increasing frequency categories. For those performing 20 to 50 repetitions of this task per shift the risk was increased by a multiple of 1.6, and at a frequency of 50 or more a four-fold risk was estimated. Transport. An aspect of underground transport which was shown to affect the risk of back pain was driving a free-steered vehicle (Tables 4.29 to 4.31). The association was a negative one, which was exhibited with each of the quantities related to driving FSVs; the number of journeys, the duration without a break and the total duration within a shift. Chi squared values for these quantities suggested associations significant at the 2% level. Confidence intervals for the odds ratios, however, just included 1.00, seemingly contradicting the finding (table 4.32). In the linear regression setting the chi squared test of the change in the regression sum of squares and the t tests associated with regression coefficients are algebraically identical, but in logistic regression they involve different distributional approximations and the former tends to be more reliable. Consequently we would more readily draw conclusions based on a test of the change in deviance than on confidence intervals derived from the estimated regression coefficients. In any case, only 19 (6%) individuals (2 cases and 17 controls) were drivers of FSVs, results related to this task may be unstable and should be interpreted with caution. If there is an effect of driving an FSV it is not in the direction of increasing the risk of back problems. One explanation of an effect may be that those driving FSVs are spending less time than others on the more physical, high risk tasks. Sixty eight percent (13) of those driving FSVs were transport or general underground workers; 11% (2) were face workers and 16% (3) development workers.

18

4.3.3

Correlation between tasks, and a proposed model

Looking at tasks separately may tell us what effect each has on the risk of back pain, but what will not be clear is whether tasks are independent. Two tasks may be highly related, meaning that one will not provide any additional information after the other has been taken into account. A stepwise type technique allows us to investigate this by ordering the tasks according to how important they would be in the model. The task coming out at the top of the list would be entered into the model and the importance of the remaining tasks re-evaluated. Task components high up in the list may move down if a related task is entered into the model, indicating that it can no longer improve the model. Such a technique was applied to the data. The task components under consideration were those appearing in table 4.32, plus additional task components whose individual association with back pain did not quite achieve statistical significance. The task associated with 'twisting while lifting 1 (A3ii) was excluded from consideration, since its association had been questioned. The analysis was repeated excluding tasks associated with driving an FSV. This was done in order to concentrate on the tasks common to all underground workers; driving an FSV was a task specific to a small number of individuals at one colliery. The objective of the stepwise regression was to find the most important tasks associated with back pain, and also to identify which tasks were inter-related. We did not restrict inclusion of terms to those significant at the 5% level, but allowed the process to continue until no further improvement was possible. This gave information about the order of importance of other tasks left for selection, and how they were affected by tasks added beforehand. In the initial analysis the task identified by question F3iv (total duration of driving an FSV)) entered the model first (table 4.33, model I). There was little to separate F3ii, F3iii and F3iv in terms of the improvement they could make to the model. After F3iv was added, the other two tasks moved down the list of tasks free for selection, indicating that they could produce no further improvement. This showed the correlation between the three; this was not unexpected since they are associated with the same task. Added next was Alii (table 4.33, model II), the frequency of lifting more than 50 kg. Consequently Clii, the frequency of carrying more than 50 kg, became less important and moved down the list of available tasks. D4ii was then included (frequency of twisting when pushing) (table 15, model III), although results suggested that there was no difference between the improvement from this task and A4iv (total duration of lifting when overreaching). At this stage none of the remaining tasks could improve the model, therefore no change was made. It was noted that A4iv stayed top of the list of remaining tasks (ie. it was preferable to any of the others) indicating that it was not correlated with D4ii. In the second stage of analysis, when F3ii, F3iii and F3iv were excluded from consideration, the first task added was Alii (table 4.34, model I), which had been second choice in the initial analysis. The second term to be added was A4iv (table 4.34, model II) which was not surprising since it had also been a candidate for third place previously. No improvement could be made to the model by adding any other task, but D4ii remained preferable to the others. On the whole, results from the second analysis did not suggest anything different from that including FSV task components. If we include transport activities, total duration driving an FSV, is shown to be the most important task (out of those

19

identified), in relation to the outcome. In addition the frequency of lifting more than 50 kg was significant. This was, in fact, the only task component significant at the 5% level when transport tasks were removed from consideration. It would seem then that the frequency of lifting more than 50 kg is the greatest factor contributing to caseship in the wider population; in particular lifting more than 50 kg more than 20 times per shift. Both D4ii and A4iv were candidates for inclusion in the model but neither achieved significance at the 5% level. The two suggested models are therefore table 4.33, model III and table 4.34, model II.

4.4

Summary

The study has suggested that there are associations between increased risk of back related absence from work and increases in (1) lifting more than 50 kg with a frequency of more than 20 times in a shift, (2) holding weights above shoulder height more than 50 times in a shift, (3) carrying more than 50 kg more than 50 times per shift, and (4) twisting the body while pushing more than 50 times in a shift. In multiple regression analyses only lifting more than 50 kg was associated significantly with increased risk.

20

21

5. DISCUSSION S.I

Introduction

Individuals recruited in this study as cases or controls were not assessed clinically during this study. The cases had at least one episode of sickness absence of seven or more days duration due to one of a number of defined back complaints. An absence of seven or more days duration in the United Kingdom requires a certificate from a doctor indicating the medical reason for the absence from work. The definition of a case based on a minimum duration of sickness absence due to a back condition was used to improve the validity of the diagnosis of a back problem, in the absence of clinical assessment. Cases were not selected by the presence of self-reported back problems. The prevalence of back conditions in the responses of the selected controls indicates that such complaints occur commonly in the colliery workforce. The use of a specific duration of sickness absence as the criterion for case selection may mean that the cases had more severe back complaints at the time of the study than the controls who reported back problems, which had not resulted in sickness absence. The controls were matched by sex, by colliery and by age (in ten year age bands). Controls were not matched by occupation. It was considered this would lead to possible overmatching with regard to the tasks performed which would have led to false negative results. The questionnaire responses of cases and controls were compared. The validation exercise suggested that the responses of cases were more accurate than those of controls, when compared to the observations of an ergonomist. The results of the main study may be more conservative than is the real situation, and the chance of attributing an increased risk of back pain to actual differences in the tasks performed is weakened by this difference in accuracy in the responses of cases and controls. The main results from this study indicate a positive relationship between being a case (i.e. having at least one episode of sickness absence of seven or more days duration due to a back problem) and lifting weights of more than 50kg more than 20 times per shift. Twisting when pushing more than 50 times per shift and lifting when overreaching for more than one hour in total per shift were included in the statistical model which was used but did not achieve statistical significance. The driving of free steer vehicles (FSVs) which was only undertaken at colliery 2 was associated with a reduced risk of absence due to a back problem. The use of transport systems other than FSVs was not associated with an increased risk of being a case. The neutral relationship between the use of transport systems underground and back problems may be related to the study population comprising mainly underground colliery staff. Those who were selected as cases and controls would have had to use the systems available to them at each colliery (manriding belts and locomotives). A different study design would be required to examine this phenomenon further. These results are further support for the relationship between heavy lifting and back pain as previously noted by Lawrence (1955), Wickstrom et al. (1985) and Walsh, Cruddes and Coggon (1991).

22

5.2

The Effects of Chance, Bias and Confounding

When considering these results the effects of chance, bias and confounding must be addressed. In this study a large number of task components have been considered. There is a real possibility that with this number of variables some individual significant results may have been due to chance. The number of significant associations which we obtained is greater than would have been expected due to chance alone. The sample size obtained had adequate power to detect real differences between the study groups (Schlessman, 1982). Correlation between individual task components (e.g. aspects of lifting, carrying and holding) is likely to occur. The use of stepwise linear regression modelling in this analysis accounted for this correlation. An observational bias could have occurred if the questionnaire administrator changed his style of questioning depending on the status within the study of the individuals questioned. The administrator was trained to ask the questions in a standard manner. He was observed at regular intervals throughout the study at each colliery administering the questionnaire to cases and controls. No changes in technique were observed over time. No differences were noted in the style of administration when questioning cases or controls. A recall bias has been suggested from the validation exercise. This may, however, be a function of the variability of the work performed by an individual underground. The ergonomist observed one day's work for each individual, whereas the questionnaire asked about a typical shift in the past three months. If there was a bias based on recall the results tend to be conservative and it should not be assumed that other tasks are free of risk. The use of questionnaires for the assessment of postural loads has been shown by Burdof and Laan (1991) to have low reliability. Our results may be seen as support of this view. The possibility that factors exist which were not assessed in this study but which are related to sickness absence due to back pain must be considered. Factors such as sporting activities, other leisure interests, caring for dependent relatives and non-work accidents were not assessed during this study. The past history of back problems was considered by the administration of the shortened IOM back pain questionnaire. There were no differences between cases and controls in length of time since the onset of back problems. The prevalence of backache or pain more than 12 months before interview was significantly different between cases and controls. The prevalence of chronic back pain or ache was significantly greater among cases than controls in the present study. The focus of the study was on tasks performed by underground miners and as such did not consider the history of back pain in the logistic regression analysis that was performed. 5.3

Risk Reduction

5.3.1

Introduction

The questionnaire incorporated the principal factors which have been established in the scientific literature as contributing to the incidence of back pain. Most of these related to manual handling activities as this has been widely implicated in both the acute and chronic development of back injuries. In a multicausal disease such as back pain it is not always easy to establish irrefutable associations

23

or causal links between potential risk factors and back problems, particularly where an individual case may have a history of insidious, gradual development prior to some possibly trivial 'trigger' event. This study sought to establish statistically robust associations between the risk factors identified from the literature and back pain and it was indeed possible to establish a number of such associations. It would be wrong, however, to conclude that a failure to determine any association between some of the other factors included in the questionnaire and back problems indicated that these factors did not have a role to play in the aetiology of such problems. It can, for example, be shown using biomechanical calculations that some such factors produce such an increase in mechanical force on the structures of the spinal column that it is at least highly likely that these factors contribute to back problems in some way. The complexities of manual handling tasks are such that variations, for example, in how an individual lifts 50 kg, or precisely how they stoop or overreach to lift, are likely to have a significant influence on spinal loading. To have developed a questionnaire to elucidate factors in such detail would have resulted in a considerable increase in questionnaire size and consequent time for completion, probably to the point of impracticability. It would appear to be likely that, at least in part, those factors which were identified as a result of the study are less subject to variations in associated risk. For example, a heavy load such as one weighing 50 kg or more bears an intrinsic and unavoidable risk because of that weight. In contrast, the risk associated with stooped lifting would be strongly influenced by major modifiers such as the horizontal distance of the load away from the feet (and lumbar spine). Direct measurements by a number of researchers have shown this latter factor to be a very powerful modifier of intradiscal pressure. This section will therefore not restrict itself to considering the contribution of those factors identified by the statistical analysis as bearing a significant association with back problems, adopting in part a wider view of risk reduction. Nevertheless, the fact that certain elements did demonstrate a. significant relationship must be taken to indicate a particularly robust involvement between these factors and back problems. This section will therefore address specific ways in which the impact of such factors can be reduced as well as a more general risk reduction strategy. 5.3.2

The reduction of injury risk from specific factors

The results of the analyses can be grouped activities, lifting, holding, carrying and pushing. in turn.

into four main manual handling Each of these will be addressed

Lifting tasks. Lifting objects weighing more than 50 kg on a frequent basis (more than 20 times a shift) was identified as a significant risk factor. Over half of all respondents reported lifting more than 50 kg and occasional lifting did not appear to differentiate between cases and controls. Lifting such weights was most common amongst underground transport workers although the difference between this group and three others (faceworkers, development workers and underground tradesmen) was not particularly large. The occurrence of three of these groups in such a list is not surprising. Faceworkers and development workers (particularly the latter) have traditionally handled considerable quantities of heavy items. And, of course, transport workers have kept them supplied with these heavy items. In recent years there have been a number of developments which, although not primarily introduced for the purpose, have had a significant impact on reducing the

24

loads handled by such workers. One such example is that of roofbolting. However, at present the benefit from this is reduced by the continuing use of props and bars as secondary supports. It is to be hoped that, in time, improved bolting standards and consequent increased confidence in their use will see roofbolting accepted as the sole means of support. However, observations of roofbolting during this project have indicated that, even with this task, a number of risk factors still remain. Some of these have been addressed in a recent British Coal ergonomics project on roofbolting and the implementation of the relevant recommendations of that report (Rushworth and Mason, 1990) would have significant further benefits. Previous IOM research (Agius et al. 1988) identified underground tradesmen as a particular risk group and studies of maintenance operations (Ferguson et al. 1986) have indicated that manual handling is a significant problem in their work. Recent experience in other industries such as power generation has shown that this is not unique to coalmining. Maintenance and repair work is often carried out in less than ideal conditions, without the facilities afforded by a modern workshop and often with the added pressures of the need to get expensive capital equipment operational again. Shortcomings of the equipment itself, from the point of view of maintainability, has recently been addressed by a British Coal ergonomics project (Coleman et al. 1990) which showed back strain to be a particular problem, especially with free-steered vehicles. Attention to the recommendations and procedures contained in that report should undoubtedly be of considerable benefit. But attention should not only be paid to the maintenance of underground machinery. Equipping a new district involves many contributions from underground tradesmen. Simple tasks such as installing a junction board or other switchgear and indicators/controllers are unnecessarily risky when positioned above transformers or other heavy plant. Stretching across to fit such devices adds additional risk factors to the weight of the units. The consideration of risks to health and safety should become a routine element in planning the design and installation of any new district. Twisting whilst lifting was a factor which was commonly reported by both cases and controls. Although no evidence for a clear positive relationship with case status was found, biomechanical and other studies have indicated the considerable risk, particularly of disc damage, which may be created by this activity. Under certain circumstances, such as twisting through congested areas (past roadheaders in developments; through face-ends; etc) careful consideration of layout, size of machinery installed etc. can reduce the need for this action. However, in many instances, a lack of awareness of the risk may mean that miners will adopt twisting postures unnecessarily. Risk awareness training may well be of benefit in such situations. Holding tasks. Holding items above shoulder height on a frequent basis (more than 50 times per shift) was identified as a strong risk factor. In lifting to this height and holding loads in such positions the lifter will commonly hyper-extend the spine, bending backwards to counterbalance the load. The group of workers most commonly reporting such activities were underground tradesmen. As was stated in the section on lifting, the manual handling activities associated with installing equipment and machinery in a new district can create considerable risk. Not traditionally regarded as a manual handling job, the role of the 'fitter 1 needs careful consideration, particularly when considered against the trend in recent years for larger, heavier equipment.

25

Carrying tasks. In mining, loads are seldom lifted without also being carried. As a result, as was suggested above, the close relationship between lifting and carrying as risk factors is not surprising. Nevertheless, it has been recognised that carrying also creates further risks beyond those shared with lifting (size and weight of load etc.). Asymmetric loading (carrying in one hand), sudden movement of loads, slipping or tripping etc. may all create an additional risk. This is recognised in the guidelines for manual handling in the coal industry produced by British Coal and the IOM (ECSC, 1990) and a number of suggestions for techniques or devices to reduce the risk of injury when carrying loads are included. The wider implementation of these and similar ideas should be of significant benefit. Pushing tasks. Pushing (and pulling) actions are frequently neglected by Health and Safety staff in considering manual handling and the risk of injury. Many mechanically assisted tasks still involve a great deal of manual force application in manoeuvring objects into place. This is particularly the case when space is restricted, making it difficult to use mechanical aids to their full potential. As a consequence, the limited space often results in combinations of twisting and pushing. Although the association is relatively weak it is not therefore surprising for twisting and pushing to be identified as a risk factor. It is also not surprising for development workers and faceworkers to be the two job groups who most frequently reported performing this activity because of the space constraints under which they frequently have to work. Whilst it would not be reasonably practicable to expect coal faces to be made higher (regardless of geological conditions) or to expect drivages to be routinely made of a larger section, some thought to the selection of appropriate equipment and the layout of equipment may well generate benefits in this (and other) aspects of the work. 5.3.3

General risk reduction

In addition to the recommendations discussed above in relation to specific risk factors identified by the study, it was apparent that the general nature of the manual handling risks warranted a broader approach. In a more controlled production environment, such as would be encountered in a factory, specific risk reduction solutions can be identified and implemented such as altering the height of shelving or worksurfaces; providing roller conveyors to convert a short lift and carry into a pushing task; etc. However, as a result of the complexity and variety of mining activities, such approaches are unlikely to be practicable. Working conditions and the handling environment vary widely between different parts of a single colliery and, on an almost continuous basis, vary over time at any one location. It would therefore appear that a different approach to the reduction of risk of manual handling injury is required. It also would seem to be appropriate, given the enactment across the European Community of the Council Directive on the Manual Handling of Loads, to devise an approach within the framework of compliance with this Directive. The precise enactment of the Directive within member states is of course a matter for national legislation. This section will therefore be framed within the specific structure of the UK Regulations enacting this Directive. However, the essence of the Directives is that their enforcement should produce a common policy within the Community. This is apparent from the paper on manual handling policy within the European Coal and Steel Community produced by Simpson (1992). Therefore, although there may be differences of detail, the approach recommended in this section should enable compliance with legislation from other National governments.

26

Under the UK Manual Handling Operations required to:

Regulations 1992

employers

are

(i) so far as is reasonably practicable, avoid the need for manual handling where there is a risk of injury; (ii) where this is not practicable, to assess the risk of injury and to take steps to reduce that risk to the lowest level reasonably practicable. The guidance accompanying the Regulations (HSE, 1992) sets out a hierarchy of steps involving mechanising the handling operation or providing mechanical assistance or examining the ergonomics of the workplace. Training is generally seen to have a subsidiary role - not to suggest that it is unimportant but to attempt to inhibit the tendency to regard manual handling training as the universal panacea. Selection of specific workers eg. on the basis of strength is explicitly discouraged because of the lack of clear scientific evidence to justify selection on any grounds other than a prior history of back problems. Training is however recognised as being of greater importance in those situations which are relatively unpredictable, or outwith the control of the employer. It also has a clear role to play in ensuring that carefully designed workplaces are used correctly. Observation of work activities at coalmines indicate that they could be subdivided into two general categories (although inevitably there will be tasks which fall into a grey area in between). The first are relatively straightforward activities occurring in a comparatively controlled, repeatable manner. Unlike in a conventional manufacturing industry, such tasks are fairly infrequent in coalmining although they do occur, particularly in relation to activities which are ancillary or secondary to the main task of winning coal. Many of these will take place on the surface, such as handling goods in and out of stores. Some however will be encountered underground. These tend to occur in established areas of the mine such as around the pit bottom and would include the manual pushing and/or pulling of minecars which is still carried out at some collieries, activities in locomotive/free-steered vehicle workshops etc. These can be assessed and risk reduction procedures identified, adopting the procedures and guidance laid down by the HSE (1992), supplemented where possible by the wealth of information available through previous ECSC projects carried out on the Ergonomics Programme. Information on these projects is widely promulgated by the Bureau of Information and Coordination in Luxembourg, including the regular publication of lists and summaries of projects in an information bulletin. In particular, the Guidelines for Manual Handling in the Coal Industry produced by British Coal with the assistance of the IOM referred to earlier provides simple, practical advice for reducing the risk of injury arising from manual handling. In many instances, risk factors can be designed out. For example, much of the manoeuvring of vehicles in the pit bottom area, especially into and out of lift cages, has been mechanised, with the widespread introduction of creeper systems and other devices. The second category of activities are those which, although the general nature of the activity is predictable, e.g. setting roof bolts, the specific characteristics of the task vary from time to time and site to site within a colliery and potentially even more so between different collieries. It is clearly impracticable to expect coalminers to carry out formal risk assessments of each handling action in the roofbolting cycle, or even for each cycle of advance. It is partly in recognition of this type of situation that the guidance associated with the Manual Handling

27

Operations Regulations allows for the concept of generic assessments. The principle underlying generic assessments is that a range of inherently similar tasks may have a common assessment - with of course an acknowledgement of the variations in risk which may be encountered. A logical extension of this principle is that the concept of generic risk assessment can equally be extended to provide for generic risk reduction measures. Therefore, in recognising that there is a basically similar fundamental task, which may or may not have certain risk factors present, it should be recognised that there will probably be a basic core of risk reduction measures with others that may be appropriate for the particular circumstances encountered in any one location at any one time. In its guidance, the HSE recognise that in variable circumstances, such as may be encountered in coalmining, a greater emphasis may be placed on education and training. In such circumstances however, a reliance on traditional training in manual handling techniques is unlikely to be sufficient. Miners in UK coalmines have, for many years, received training in kinetic handling techniques, and yet the levels of injury and sickness absence from back problems remain high. What would also seem to be appropriate, and is being applied in other industries with similar problems, is the concept of education in risk recognition and reduction. For example, for a given activity such as roofbolting, the generic assessment will probably identify certain risk factors which are always present, some which are often present and others which may be present in certain defined circumstances. Those factors which are always present will be addressed with appropriate risk reduction techniques which will hopefully involve solutions aimed at designing out the problem at source (wherever reasonably practicable), but may include instruction in the correct techniques for carrying out the work with the minimum risk of injury. Workers involved in the operation in question will clearly require this training but they will also need education in recognising when other risk factors are present and, equally importantly, recognising what control actions are therefore necessary. These controls may be handling aids or other devices which should be used to counter specific problems. Alternatively, there may be alternative techniques which are appropriate in particular circumstances. It should be recognised that these aids or techniques may make the job slower, but that this reduced rate of working may be appropriate in situations of increased risk. In this case, management or organisational procedures and schedules must accommodate this; workers should not be penalised for taking steps to protect their health and safety. The first element in any such education package should be to inform staff of the risks to their health, the likelihood of individuals experiencing such problems, and the potential impact on their general health and lifestyle. For, unless workers recognise that the implications of failure to apply recommended procedures can have serious consequences for their personal health, it will be difficult to motivate them to take the appropriate course of action. In summary, therefore, although risk reduction procedures generally applicable to all examples of a generic task may be identifiable, it is likely that a suite of procedures may be more appropriate. Miners will require to learn to recognise the presence of particular problems (in the same way as they are expected to recognise the presence of other problems which may require action). They will then need to select risk reduction techniques appropriate to the problems identified and will require to be sufficiently committed or motivated to adopt such techniques even where the ensuing action takes slightly longer.

28

5.4 Further Research The mean duration of absence due to a back problem among cases in this study was 16.7 days. In total 1737 days were lost due to back problems at both collieries among the cases during the six month period. The prolonged nature of these absences suggests a passive approach to the management of back problems. Recently, more active rehabilitation back to work has been advocated. This approach includes the limiting of bed rest to 48 hours with mobilisation thereafter and an early return to duties for those without neurological impairment (Deyo, 1988). The active management of back pain in coalminers merits consideration and could be usefully studied to determine if such a management programme has the benefits noted in other research results (Wiesel, Feffer and Rothman, 1984). The economic benefits from reduced sickness absence are obvious and there are also likely to be significant benefits for the individual miner with back pain from this approach, in terms of earlier return to full function.

29

6. CONCLUSIONS AND RECOMMENDATIONS Back pain or ache is a common complaint among underground miners. Sickness absence due to back pain was associated with tasks involving lifting weights of more than 50kg more than 20 times per shift. Efforts should continue to reduce the frequency with which miners are required to lift and carry heavy weights. Further reduction in manual handling by underground miners should be actively pursued. The wider availability underground of jacks, hoists and other handling aids which are sufficiently robust will assist in this reduction. In addition, the ease with which machinery can be maintained underground should be a consideration in the purchase of such equipment. This will assist in reducing the physical strain associated with manual handling by maintenance workers and fitters. A programme of risk awareness training of coalminers will be of benefit. Such training should cover how to recognise which risks are present in any planned task; understanding the relative significance of these risks and familiarisation with those risk reduction procedures likely to be applicable. Refresher training at regular intervals for risk reduction and manual handling will be beneficial. The colliery management must ensure that the training is appropriate to the industry and that the application of the techniques is facilitated underground and adhered to by all miners. Further research is recommended to complete the validation of the task component questionnaire. Some modification to the questions may improve the reliability of the responses and may result in a useful instrument for epidemiological research. The active medical management of back pain in coalminers merits further study.

30

31

ACKNOWLEDGEMENTS This study was jointly funded by British Coal and CEC (Contract 7280/04/022). We wish to thank the British Coal management and employees at both collieries for all the cooperation provided during this study. The work of Mr George Pugh, the questionnaire administrator, was essential to the study and our thanks to him are willingly noted. We also wish to thank Dr Geoffrey Carreck, British Coal Medical Services and the nurses and staff at the two colliery medical centres for assistance with this study.

32

33

REFERENCES Afacan AS. (1982). Sickness absence due to back lesions in coal Journal of the Society of Occupational Medicine; 32: 26-31.

miners.

Agius RM, Lloyd MH, Campbell SJ, Graveling RA, Hutchison PA, Muirhead M, Gurr D, Soutar CA. (1988). Epidemiological and clinical studies of back pain in the mining industry. Final report on CEC Contract 7248/15/019. Edinburgh: Institute of Occupational Medicine. (IOM Report TM/88/10). Biering-Sdrensen F, Thomsen G. (1986). Medical, social and occupational history as risk indicators for low-back trouble in a general population. Spine; 11: 720-725. Bongers PM, Boshuizen HC, Hulshof CTJ, Koemeester AP. (1988). Back disorders in crane operators exposed to whole-body vibration. International Archives Occupational Environmental Health; 60: 129-137. Brendstrup T, Biering-Sdrensen. (1987). Effect of fork-lift truck driving on low-back trouble. Scandinavian Journal of Work, Environment and Health; 13: 445-452. Burdorf A, Laan S. (1991). Comparison of methods for the assessment of postural load on the back. Scandinavian Journal of Work, Environment and Health; 17: 425-429. Coleman GJ. (1990). Ergonomic guidelines for the design of mining machinery to facilitate maintenance. (Final Report on ECSC Project 7249/12/011). Luxembourg: Commission of the European Communities. Daltroy LH, Larson MG, Wright EA, Malspeis S, Fossel AH, Ryan J, Zwerling C, Liang MH. (1991). A case-control study of risk factors for industrial low back injury: implications for primary and secondary prevention programs. American Journal of Industrial Medicine; 20: 505-515. Damkot DK, Pope MH, Lord J, Frymoyer JW. (1984). The relationship between work history, work environment and low-back pain in men. Spine; 9: 395-399. Deyo RA. (1988). The role of the primary care physician in reducing work absenteeism and costs due to back pain. In: Deyo RA (ed). Back pain in workers. Philadelphia: Hanley and Belfus: 17-30. (Occupational Medicine: State of the Art Reviews Vol 3(1)). Deyo RA, Bass JE. smoking and obesity.

(1989). Spine;

Lifestyle and low-back pain: 14: 501-506.

the influence of

Dupuis H, Zerlett G. (1987). Whole-body vibration and disorders of the spine. International Archives Occupational Environmental Health; 59: 323-336. English CJ, WA, Graves of relations work. Edinburgh:

Maclaren WM, Court-Brown C, Hughes SPF, Porter RW, Wallace RJ, Pethick AJ, Soutar CA. (1989). Clinical epidemiological study between upper limb soft tissue disorders and repetitive movements at Final Report on HSE Research Contract No. 1/MS/126/585/85. Institute of Occupational Medicine. (IOM Report TM/88/19).

34

European Coal and Steel Community. (1990). Guidelines for manual handling in the coal industry. Luxembourg: ECSC. (Report No. 14, Series 3). Ferguson CA, Mason S, Collier SG, Golding D, Graveling RA, Morris LA, Pethick AJ, Simpson GC. (1985). The ergonomics of the maintenance of mining equipment (including ergonomic principles in designing for maintainability). Final Report on CEC Contract No. 7247/12/008. Edinburgh: Institute of Occupational Medicine. (IOM Report TM/85/12). Frymoyer JW, Pope MH, Clements JH, Wilder DG, MacPherson B, Ashikaga T. (1983). Risk factors in low back pain: an epidemiological survey. Journal of Bone and Joint Surgery; 65-A: 213-218. Frymoyer JW, Pope MH, Costanza MG, Rosen JC, Goggin JE, Wilder DG. (1980). Epidemiologic studies of low-back pain. Spine; 5: 419-423. Garg A, Moore JS. (1992). Epidemiology of Low Back Pain in Industry. In: Moore JS, Garg A (eds). Ergonomics: low back pain, carpal tunnel syndrome and upper extremity disorders in the workplace. Philadelphia: Hanley and Belfus: 593-608. (Occupational Medicine: State of the Art Review Vol 7(4)). Genstat 5 Committee. (1987). Press.

Genstat 5 reference manual.

Oxford:

Clarendon

Health and Safety Executive. (1992). Manual handling. Manual Handling Operations Regulations 1992. Guidance on Regulations. London: Her Majesty's Stationery Office. Heliovaara M. (1987). Occupation and risk of herniated lumber intervertebral disc or sciatica leading to hospitalization. Journal of Chronic Disease; 40: 259-264. Hosmer DW, Lemeshow S. (1989). Wiley & Sons Inc.

Applied logistic regression.

New York:

John

Johanning E. (1991). Back disorders and health problems among subway train operators exposed to whole-body vibration. Scandinavian Journal of Work, Environment and Health: 17: 414-419. Kellgren JH, Lawrence JS. (1952). Rheumatism in miners. British Journal of Industrial Medicine; 9: 197-207.

Part 2:

X-ray study.

Kelsey JL, Githens PB, O'Conner T, Weil W, Calogero JA, Holford TR, White AA, Walter SK, Ostfield AM, Southwick WO. (1984). Acute prolapsed lumber intervertebral disc: epidemiologic study with special reference to driving automobiles and cigarette smoking. Spine; 9: 608-613. Kelsey JL, Hardy RJ. (1975). Driving of motor vehicles as a risk factor for acute herniated lumbar intervertebral disc. Americak Journal of Epidemiology; 102: 63-73. Lawrence JS. (1955). Rheumatism in coal miners. Part III: British Journal of Industrial Medicine; 12: 249-261.

Occupational factors.

Lawrence JS, Aitken-Swan J. (1952). Rheumatism in miners. Part Rheumatic complaints. British Journal of Industrial Medicine; 9: 1-18.

1:

35

Maeda K, Okazaki F, Svenaga T, Sakurai T, Tagamatsu M. (1980). Low back pain related to bowing posture of greenhouse farmers. Journal of Human Ergology; 9: 117-123. Magora A. (1973). Investigation of the relation between low back pain and occupation. Scandinavian Journal of Rehabilitation Medicine; 5: 186-190. Punnett L, Fine LJ, Keyserling WM, Herrin GD, Chaffin DB. (1991). Back disorders and nonneutral trunk postures of automobile assembly workers. Scandinavian Journal of Work, Environment and Health; 17: 337-346. Riihimaki H (1991). Low back pain, its origins and risk indicators. Journal of Work, Environment and Health; 17: 81-90. Reisbord LS, Greenland S. (1985). back-pain prevalence: A population-based 38: 691-702.

Factors study.

Scandinavian

associated with self-reported Journal of Chronic Disease;

Riihimaki H. (1985). Back pain and heavy physical work: a comparative study of concrete reinforcement workers and maintenance house painters. British Journal of Industrial Medicine; 42: 226-232. Riihimaki H, Mattson T, Zitting A, Wickstrflm G, HSnninen K, Waris P. (1990). Radiographically detectable degenerative changes of the lumbar spine among concrete reinforcement workers and house painters. Spine; 15: 114-119. Riihimaki H, Tola S, Videman T, Hanninen K. (1989). Low-back pain and occupation. A cross-sectional questionnaire study of men in machine operating, dynamic physical work and sedentary work. Spine; 14: 204-209. Riihimaki H, Wickstrom G, Hanninen K, Luopajarvi T. (1989). Predictors of sciatic pain among concrete reinforcement workers and house painters - a five-year follow-up. Scandinavian Journal of Work, Environment and Health; 15: 415-423. Rushworth AM, Mason S. (1990). Design study of manual roofbolting machines. Burton-on-Trent: British Coal Corporation. (Report No. SSL/90/164). Schlesselman JJ. Press.

(1982).

Case-control studies.

New York:

Oxford University

Simpson GC. (1992). The European Coal and Steel Community ergonomics programme's approach to evolving a corporate manual handling policy. Progress in Coal, Steel and Related Social Research; (13): 4-9. Svenson H-O, Andersson GBJ. (1983). Low-back pain in 40- to 47- year old men: Work history and work environment factors. Spine; 8: 272-276. Troup JDG (1984). Causes, prediction and prevention of back pain at work. Scandinavian Journal of Work, Environment and Health; 10: 419-428. Videman T, Nurminen T, Tola S, Kuorinka I, Vanharanta H, Troup JDG. (1984). Low-back pain in nurses and some loading factors of work. Spine; 9: 400-404.

36

Walsh K, Cruddas M, Coggon D. (1991). Interaction of height and mechanical loading of the spine in the development of low-back pain. Scandinavian Journal of Work, Environment and Health; 17: 420-424. Wickstrflm G, HSnninen K, Lehtinen M, Riihimaki H. (1978). Previous back syndromes and present back symptoms in concrete reinforcement workers. Scandinavian Journal of Work, Environment and Health; 4 (suppl 1): 20-28. Wickstrflm G, Niskanen T, Riihimaki H. (1985). Strain on the back in concrete reinforcement work. British Journal Industrial Medicine; 42: 233-239. Wickstrflm G, Nummi J, Nurminen M. (1978). Restriction and pain during forward bending in concrete reinforcement workers. Scandinavian Journal of Work, Environment and Health; 4 (suppl 1): 29-38. Wiesel SW, Feffer HL, Rothman RH (1984). Industrial low back pain. A prospective evaluation of a standardised diagnostic and treatment protocol. Spine; 9: 199-203.

37

Table 4.1 Distribution of cases and controls within c o l l i e r y and five year age group

COLLIERY COLLIERY 1 Case

25-29 30-34 35-39 40-44 45-49 50-54

1 4 15 10 4 8 4

Total

46

50 times Total

7 (15) 20 (43) 13 (28) 3 ( 7) 3 ( 7) 46

23 (23) 43 (44) 25 (26)

30 (21)

63 (44)

38 (26)

14 (26) 19 (36) 11 (21) 5

5

8

( 5) 2 ( 2)

( 6)

( 9)

5

4 ( 8)

98

( 3) 144

53

26 (26) 42 (42) 27 (27) 4 ( 4) 2 ( 2)

101

40 (26) 61 (40)

38 (25)

9 ( 6)

6 ( 4) 154

21 (21) 39 (39) 24 (24) 8 ( 8) 7 ( 7)

99

9 ( 5) 4 ( 2)

70 (23) 124 (42) 76 (26) 17 ( 6) 11 ( 4)

199

298

49 (25)

85 (43)

52 (26)

Table 4.27 Number (percentage) of cases and controls performing those CARRYING task components independently associated with case/control status, by frequency category and colliery. Carrying Task

| Colli ery 1 [Colliery 2 (Cases Controls Total [Cases Controls Total

TOTAL Cases Controls Totals

50 kg: No

< 20 t imes 20-50t imes > 50 times Total

12 (26) 20 (43) 11 (24) 3 ( 7) 46

42 (43) 40 (41) 14 (14) 2 ( 2) 98

54 (38) 60 (42) 25 (17) 5 ( 3)

144

29 (55) 19 (36) 2 ( 4) 3 ( 6) 53

48 (48) 49 (49) 3 ( 3) 1 ( 1)

77 (50) 68 (44) 5 ( 3) 4 ( 3)

101

154

41 (41) 39 (39) 13 (13)

6 ( 6) 99

90 (45) 89 (45) 17 ( 9) 3 ( 2)

131 (44) 128 (43) 30 (10) 9 ( 3)

199

298

53

Table 4.28 Number (percentage) of cases and controls performing those PUSHING task components independently associated with case/control status, by frequency category and colliery. Pushi ng Task

(Colliery 2 |Col 1 iery 1 | Cases ControlsTotal (Cases Controls Total

TOTAL Cases Cont rol s Tola 1 s

Twist : No

< 20 times 20-50times > 50 times

Total

16 (35) 16 (35) 9 (20) 5 (11)

46

38

(39) 51 (52) 5

( 5) 4 ( 4) 98

54 (38) 67 (47) 14 (10) 9 (6) 144

17 (32) 28 (53) 4 ( 8) 4 (8) 53

36 (36) 51 (50) 12 (12) 2 ( 2)

53 (34) 79 (51)

101

154

16 (10)

6 ( 4)

33 (33) 44 (44) 13 (13) 9 ( 9)

99

74

17 ( 9) 6 ( 3)

107 (36) 146 (49) 30 (10) 15 ( 5)

199

298

(37)

102 (51)

Table 4.29 Number (percentage) of cases and controls journeying by F.S.V., by frequency category and colliery. Journey by|Colliery 1 |Colli ery 2 TOTAL F.S.V. Cases Controls Total j Cases Controls Total (Cases Controls Totals Frequency: Ni 1

46 (100)

97 ( 99)

143

51

85

( 99) ( 96) ( 84)

1

1 ( 1)

2

1 ( 2) 1 ( 1)

3 4

1 ( 1)

1 3 ( 2) ( 3) 2 ( 2) 4 ( 4) 1 ( 1) 1 ( 1) 1 ( 1)

5 6 8 10 12

Total

3

( 3)

46

98

144

53

101

182 136 97 ( 88) ( 98) ( 91) 1 1 ( 1) ( 1) 1 1 ( 1) ( 1)

279 ( 94) 1

4 ( 2) 3 ( 2) 2 ( 1) 4 ( 2) 1 ( 1) 1 ( 1) 1 ( 1)

4 1) 4 1) 2 1) 4 1) 1

3 ( 2) 4 1 ( 3) ( 1) 2 ( 1) 4 ( 3) 1 ( 1) 1 ( 1) 1 ( 1) 154

99

199

1

( ( ( (

1 1

298

54 Table 4.30 Number (percentage) of cases and controls journeying by F.S.V., by category of longest duration and colliery. Journey by|Colliery 1 (Colliery 2 F.S.V. |Cases Controls Total (Cases Controls Total

TOTAL Cases Controls Totals

Longest durat ion

(hours) : Nil

46 (100)

97 ( 99)

1.10 1.27 1.43

1 (1)

1 .52 1.85 2.02 3.02 3.52 5.02

Total

46

98

97 182 143 51 85 136 279 ( 99) ( 96) ( 84) ( 88) ( 98) ( 91) ( 94) 1 1 1 1 ( 1) ( 1) ( 1) 2 1 2 1 3 3 ( 4) ( 1) ( 2) ( 2) ( 1) ( 1) 2 2 2 2 ( 2) ( 1) ( 1) ( 1) 1 4 4 5 5 ( 4) ( 3) ( 3) ( 2) (1) 1 1 1 1 ( 1) ( 1) ( 1) 4 4 4 4 ( 2) ( 1) ( 4) ( 3) 1 1 1 1 ( 1) ( 1) ( 1) 1 1 1 1 ( 1) ( 1) ( 1) 1 1 1 1 ( 1) ( 1) ( 1)

144

53

101

154

99

199

298

55

Table 4.31 Number (percentage) of cases and controls journeying by F.S.V., by category of total duration and colliery.

Journey by|Colliery 1 |Colliery 2 |TOTAL F.S.V. Cases Controls Total (Cases Controls Total Cases Controls Totals Total durat ion (hours) : 0.00

46 (100)

97 ( 99)

1.18

85 143 51 ( 99) ( 96) ( 84) 1 ( 1)

1

2.02

( 2) 2.52

1 (1)

1 (1)

98

144

3.02 4.02

. 4.52 5.02 6.02 7.02

Total

46

1 ( 1) 2 1 ( 2) ( 2) 3 ( 3) 1 ( 1) 4 ( 4) 3 ( 3) 1 ( 1) 53

101

97 182 136 279 ( 88) ( 98) ( 91) ( 94) 1 1 1 ( 1) ( 1) 1 1 1 ( 1) ( 1) 1 2 2 ( 1) ( 1) ( 1) 1 2 3 3 ( 2) ( 1) ( 1) ( 1) 3 3 3 ( 2) ( 2) ( 1) 1 1 1 ( 1) ( 1) 4 4 4 ( 3) ( 2) ( 1) 3 3 3 ( 2) ( 2) ( 1) 1 1 1 ( 1) ( 1)

154

99

199

298

56

Table 4.32 Logistic regression model for task components contributing to risk of back pain. Estimated regression coefficients have been expressed on the odds scale, as the ratio of the odds of being a case between different task categories. Approximate 95% confidence intervals for the odds ratio are presented; those which do not include 1.00 indicate signficance at 5%. (Parameters for age group, colliery and their interaction are not shown here, although they were included in the model.) Approximate 95% confidence interval

Task

Estimated odds ratio

Q.Alii Li ft ing more than 50 kg

Yes v. No:

1 .149

[0 .692, 1.909]

0 v . < 20 t imes : 0 v . > 20 t imes :

1 .314 1 .610

[0 .565, 1.647] [1 .239, 8.004]

Q.A3H Twist ing whi le lifting

Yes v. No:

0 .397

[0 .312,

1.198]

[0 .048, [0 .172, [0 .136, [0 .110,

0.642] 1.749] 1.446] 1.254]

Q.BSii Ho 1 d i ng above shoulder

Yes v. No:

0 0 0 0

0 0 0 0

v. v. v. v.

v. v. v. v.

< 10 t imes :

0.176 10-30 t i me s :0.548 30-50 t imes : 0.444 > 50 t imes : 0.372

1 .262

< 10 t imes : 1 .165 10-30 t imes : 1 .097 30-50 t imes : 2 .069 > 50 t imes : 5 .265 1 .169

[0 .690, 2.309] [0 .601, [0 .523, [0 .680, .317,

[1

2.260] 2.302] 6.298] 21.046]

[0 .701,

1.949]

Q.Clii Carrying more than 50 kg

Yes v. No :

0 v . < 20 t imes : 0 .944 0 v . 20-50 t imes : 1 .831 0 v . > 50 t imes : 5 .534

[0 .543, 1.640] [0 .763, 4.398] .258, 24.356]

Q.D4H Tw i s t i ng when pushing

Yes v. No :

[0 .702, 2.000]

1 .185

0 v . < 20 t imes : 0.949 0 v . 20-50 t imes : 1 .636 0 v . > 50 t imes : 3 .991

[1

[0 .542, 1.663] [0 .696, 3.844] .246, 12.785]

[1

0. 3 6. 586 (2. 540 1 inear trend) 2. 7

10. 695 (0.010 1 inear t rend)

0. 6 7. 208 (4. 185 1 inear trend)

0. 4 7. 464 (3. 966 1 inear trend) 0. 403 7. 475 (4. 433 1 i near t rend)

57

table 4.32

contdA

Task

Estimated odds ratio

Approximate 95% confidence interval

2 X

Q.F3 Travel 1 ing

Yes v. No :

0.239

[0.050, 1.143]

4.2

by

per increase of 1 journey:

0.678

[0.448, 1.028]

5.8

0.371

[0.132, 1.047]

5.7

0.627

[0.389, 1.011]

6.2

FSV

per increase in t ime, wi thout

a break, of 1 hour: per increase in total t ime of 1 hour:

58

Table 4.33 Results of fitting stepwise logistic regression model to case/control status of 298 men (21 surface workers excluded). Values are estimated regression coefficients; absolute value of ratio of estimate to standard error is in parentheses. Included in all models are terms for AGE GROUP, COLLIERY and the interaction AGEGROUP.COLLIERY; no parameters are presented for these. Residual deviances, degrees of freedom and p values are given at the foot of the table. The tasks included for selection were Alii, A4ii, A4iv, B5ii, Clii, Cliv, D2ii, F3ii, iii, F3iv. Task (per shift)

Baseline

Tot. duration driving FSV (F3iv)

I -0.468 (1.91)

No v. lift >50kg up to 20 times

lift >50kg more than 20 times (Alii) No v. twist & push up to 20 times

II -0.479 (1.96)

-0.486 (1.97)

0.030 (0.11) 1.201 (2.49)

0.027 (0.10) 1.005 (2.02) -0.078 (0.26) 0.469 (1.04) 1.102 (1.80)

twist & push 20 - 50 times twist & push more than 50 times

( D4 i i ) Residual deviance Residual D.F. Significance of added term (p value)

364.3 280

III

358.0 279

< 0.02

351.4

277 < 0.05

346.3

274 < 0.20)

(4 missing values)

59

Table 4.34 Results of fitting stepwise logistic regression model to case/control status of 298 men (21 surface workers excluded). Values are estimated regression coefficients; absolute value of ratio of estimate to standard error is in parentheses. Included in all models are terms for AGE GROUP, COLLIERY and the interaction AG EG RO UP. COLLIERY; no parameters are presented for these. Residual deviances, degrees of freedom and p values are given at the foot of the table. The tasks included for selection were Alii, A4ii, A4iv, B5ii, Clii, Cliv, D2ii. Task (per shift)

Baseline

No v. lift >50kg up to 20 times

I

II

-0.039 (0.14)

1 ift >50kg more than 20 times (Alii) No v. 1 ift when overr . 1 hour (A4iv)

Significance of added term (p value)

(2.29) 0.460 (1.05) -0.218 (0.46) -0.252 (0.50) 0.576 (1.23)

1 i ft when overr . 15-30 minutes

Residual deviance Residual D.F.

-0.052 (0.18)

364.3 280

357.8 278

350.9 274

< 0.05

< 0.20

(4 missing values)

60

61

Figure 4.1 Histograms of number of whole years from first attack of back pain to interview, for cases and controls Controls

Cases Years _

2 4 6 -

8 10 12 14 16 18

-

2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34

8 16 6 7

9 8

3

******** **************** ****** ******* ********* ******** *** ***** * * *

5 1 1 1 20 22 0 1 * 24 1 * 26 0 28 3 *** 30 32 0 1 * 34 Mi ss ing val ues : 33

Years 2 4 6 8 10 12 14 -

2 12 **• ************ 4 21 **: ********************* *: 6 12 * ************ 8 6 •k-k: ****** 10 20 **^ ******************** 12 11 **: *********** 4 **• 14 16 7 •k-k; 16 - 18 7 **• 1 * 18 - 20 6 **• 20 - 22 4 **• 22 - 24 24 - 26 0 1 * 26 - 28 28 - 30 0 2 ** 30 - 32 32 - 34 0 34 1 * Missing values : 100

62

63

APPENDIX 1 Ergonomic Surveys of Study Collieries 1. Procedure The purpose of the ergonomic surveys were to assess tasks underground for task elements which have a potential risk of inducing back pain. In order to do this an ergonomics checklist was prepared based upon elements identified in the literature associated with back pain. 2. Ergonomic Surveys Work carried out underground can broadly be grouped into four categories:( i) Transport and handling ( ii) Development work ( i i i ) Maintenance work ( iv) Facework

Colliery 2 Transport and handling At colliery 2 the men travel around the pit bottom by locomotive and there are no man-riding belts. Due to the uneven surfaces underground man-riding by locomotive over long distances places the men under vibratory forces for long periods. This is made worse by the cramped conditions of the cars with men having to assume awkward postures particularly those men carrying a lot of equipment. Due to the pit's size some of the miners can spend over 30 minutes travelling by locomotive to their place of work. However, it is the locomotive drivers who are possibly at greatest risk as they spend the majority of their shift on a locomotive, albeit in somewhat less cramped conditions. As for materials transport in most cases the locomotive transport system does not extend as far as the materials point of use and therefore additional transportation is required. Free-steered vehicles (FSV) and Becorits (monorail transporter) serve this purpose. An FSV is operated by one man, the driver. He is responsible for loading and unloading materials from his vehicle and carries out a large amount of manual handling. The terrain over which these vehicles travel is usually very uneven and the drivers may experience large vibratory forces. The other transporter known as the Becorit (manufacturers name) runs on an overhead monorail and has two man-riding cars, one at each end. At any one time there may be three to four men working with a Becorit as it is a large capacity transporter and therefore there is a lot of manual handling when loading and unloading. The man-riding cars on the Becorit are small and so shaped that the men have to adopt an awkward travelling posture which they may have to maintain for long periods. There is a large amount of manual handling of materials at transfer points where the materials are unloaded for direct use or loaded onto other vehicles. Additionally, there is a lot of materials handling at the point of use. When handling materials men were observed lifting items often weighing more than 20Kg from deep sided vehicles on uneven surfaces. Additionally when handling materials throughout the pit bottom men were observed overreaching, lifting and

64

carrying objects away from the body and pushing/pulling at below knee height. Development work At the front of a heading a cutting machine advances the drivage and the newly exposed driveway has then to be supported. The cutting machine is operated by one man. During the periods when the machine is cutting, the operator is potentially subjected to large vibratory forces, depending upon the geological conditions. Once the driveway is exposed it is supported by arch girders or large bolts fixed to the roof and to the sides of the roadway. A considerable amount of manual handling is required by members of the heading team in order to transport materials to the front of the heading for the purpose of supporting the roadway. Items such as fishplates, dowel rods for roof bolting, sections of wire meshing, arch girders (crown and legs) are lifted and carried to the front of the cutting machine. Occasionally very heavy and awkward loads are transported up to the front of the heading by a smaller monorail system. Men were observed extending the length of such a monorail up the heading. One end of the new monorail section was chained to the adjacent arch girder crown by a man working with his arms above his head whilst standing on a platform. Two men below lifted the other end of the monorail upwards above their heads and it was also chained to the roof. The monorail was then hammered into position from above using a sledgehammer. Due to the weight of the monorail and the uneven surfaces underfoot this process involved men applying large forces whilst having to adopt awkward postures. In some situations makeshift platforms might be devised, potentially resulting in an unstable footing with a consequent increased risk of injury, not just from the risk of falling but also due to the additional risks associated with handling loads overhead in such circumstances. When the arch supports are being set the arch crown and legs are placed onto a beam on the front of the cutting machine. The crown is bolted to the two legs and the arch is raised up by the cutting machine and put into place. Timber and wire-meshing are then brought forward in order to pack out the sides of the arches. When packing out the arches men were observed overreaching and pushing/pulling above head height. The other method used for supporting roadway is by roof and side bolting. A borer usually operated by two men is used to drill a hole for the bolt. Once the hole has been drilled a fixing resin is placed into the hole. The bolt is then drilled into the hole using the borer. In order to keep the borer in position men were observed leaning backwards when roof bolting and applying a large force. Maintenance work Due to large amounts of mechanical and electrical hardware underground most of the working teams have individuals responsible for the maintenance of equipment. Where in the pit there is any heavy machinery there also is a mechanical attendant known as a fitter, and an electrician. On the whole the majority of fitters and electricians do not do very much heavy lifting. However, particularly mechanical fitters may from time to time be required to carry out repair or maintenance work underground which may involve extensive manual handling. A previous IOM project (Ferguson et al. 1985) identified the manual handling of

65

machine components in restricted headroom conditions as a major problem area in underground mines. Checker fitters, for example, working at the coalface are responsible for replacing chock legs (powered roof supports). These chock legs represent a very heavy, bulky load and handling this load in conditions of restricted headroom underneath the chock line is difficult. In this study, chock fitters were observed overreaching, pushing and pulling with hands below knee height, not keeping the load close to the body, stoop-walking and applying force in a prolonged awkward posture. Belt-men are responsible for the maintenance of conveyors. The function of a belt maintenance man can vary from cleaning spillages to replacing rollers and sections of belt. Cleaning spillages involves shovelling away loose coal so that it doesn't obstruct the belt. Bad floor conditions compounded by restricted headroom was found to lead to poor posture. Replacing rollers and sections of conveyor is another manual handling function which is made difficult due to the structure of the conveyor and men were observed handling loads away from the body, overreaching, pushing and pulling below knee height and above head height. The greatest load is provided by the conveyor belts which are rolled up and then replaced. Backripping is the process of maintaining and repairing roadway which is in a bad condition. Damaged or worn arch crowns and legs are replaced by new arch settings. The old arch settings are pulled down using a tirfor (mechanical aid used for pulling) and new arch sections are put into place manually. Firstly the legs are placed into position. In some cases the sides of the roadway may be in a bad condition and need to be cleared in order to allow the new leg sections to be erected. In these cases the sides are trimmed using air or hand picks and the debris shovelled away. In order to get high enough to place the arch crown in position the miner may be tempted to improvise a platform. He lifts the crown manually onto what is known as a horsehead which is a mobile girder supported by rings on the underside of the already set crowns. This girder slides forward through the rings and juts out in front of the previously set crown so that it can support the new crown section. The new crown section is lifted on top of the horsehead, which holds it in place and the crown is then bolted to its two legs on either side of the roadway. The backripping man, during removal and setting of the new arches, often has to lift, push and pull heavy arch sections with his hands above head height. The nature of the work is such that the backripper has to adopt many awkward postures often when applying force. Additionally when he is using the air pick to trim the sides of the roadway his body experiences vibratory forces from the tool. Lastly it is a significant factor that a lot of his work is carried out in old roadways which are often very cold and windy. Facework Men working at the coal face are responsible for ensuring that the faceline is being properly maintained and that the coal is being efficiently fed away from the face. Most of the time these men are working directly underneath the chocks and standing room is therefore dependent upon the height of the coal seam. The faces are reasonably high and some of the men observed were standing whilst others were having to stoop when walking up the faceline. The majority of the handling is carried out by the fitters but occasional handling can be observed by the rest of the team e.g. when using timber to pack out the front of the chocks. Observation of this handling provided examples of overreaching with a load, twisting and pushing above shoulder height with the

66

hands positioned away from the body. Colliery 1 Transport and handling Men can travel around the pit by locomotive or man-riding belt. The type of locomotives used at Colliery 1 are the same as at Colliery 2. The difference between the pits is the presence of man-riding belts at colliery 1, where the men either sit up or lie down on their fronts to travel on the conveyor belt. The belt, which lies on rollers provides an up and down vibratory force on the body during riding. The major problem with man-riding belts however is caused by getting off them. The belt does not stop and therefore the men have to step off at designated landing stages. Unfamiliarity can often result in falls. As for materials colliery 2 to the manual handling materials handled

transport the locomotive transport system extends further than at materials point of use. This therefore reduces the amount of to the unloading of the locomotive at the point of use. The and procedures used are the same as at colliery 2.

Development work The methods and practices used in development work at colliery 1 are the same as those documented in the corresponding section for colliery 2. There does tend to be more roof bolting at colliery 1 than was apparent at colliery 2. Maintenance work The different types of maintenance work described for practices at colliery 2 are the same as those carried out at colliery 1. However, in addition to these jobs track laying was also observed. When a new length of track is being laid, steel sleepers are laid down on the ground at equidistant intervals. Rails carried by two men are placed along each side of the line of sleepers and fastened onto the sleepers using a mallet to hammer them into place. Once the rails are in place they are bolted to the existing track using fish-plates which are tightened into place using a ratchet socket. Having laid the new track these men level the ground using picks and mallets. The working practices observed in track laying involved a large amount of manual handling, lifting, carrying, lowering etc. and various tools such as picks and mallets were used with great force by the men, who operated in a bent (stooped) position. Facework The heights of the coalfaces at colliery 1 are approximately the same as those at colliery 2. Working practices observed at the face were also much the same.

67

APPENDIX 2 Study Questionnaire

68

8 Roxburgh Place • Edinburgh EH8 9SU

IOM

Task Components and Back Pain Study Confidential Questionnaire

Colliery Study identity number Name of Interviewee (Surname, Initials) Date of Birth (dd mm yy) Works Identity number Job Title Code at IOM Years in current job at this colliery Years in coal industry Date of Interview (enter as day, month, year eg. 9th July 1991, as 090791) Time of Interview (enter as hour and minute according to 24 hour clock eg. 2.05pm as 1405)

J

69

I am going to ask you • some questions about working underground. I will ask a question and show you a card with a selection of answers. I would like you to pick out one answer for me. All of the questions are about work during a typical shift.

Section A The questions in this first section are about lifting objects when working underground during a typical shift. Interviewer to show miner card listing objects weighing more than 50kg. This card lists some objects used underground which weigh more than 50kg (that is about 8 stones). 1.

i)

During a typical shift do you, on your own, lift any object weighing more than 50kg ? If NO, go to question 2 ii)

Y/N

How many times in a shift do you lift objects like these weighing more than 50kg ? Show miner Card A

iii)

What is the longest time without a break that you spend repeatedly lifting objects weighing more than 50kg ? Show miner Card B

iv)

What is the total amount of time in a shift that you spend lifting objects weighing more than 50kg ? Show miner Card C

Interviewer to show miner Card listing objects weighing more than 20kg This card lists some objects used underground which weigh more than 20kg (that is about 3 stones). 2.

I)

During a typical shift do you lift any objects with one hand which weigh over 20kg ?

If NO, go to question 3 ii)

How many times in a shift do you lift objects weighing more than 20kg with one hand ?

Show miner Card D iii)

What is the longest time without a break that you spend repeatedly lifting objects like these with one hand ?

Show miner Card E iv)

What is the total amount of time in a shift that you spend lifting objects weighing over 20kg with one hand ? Show miner Card B

Y/N

70

3.

0

During a typical shift do you twist your body when lifting any object ? Show miner Photo 2

Y/N

If NO, go to question 4 ii)

How many times in a shift do you twist your body when lifting objects ? Show miner Card D

iii)

What is the longest time without a break that you spend repeatedly twisting your body when lifting objects ? Show miner Card E

iv)

What is the total amount of time in a shift that you spend twisting your body when lifting objects ? Show miner Card B

4.

i)

During a typical shift do you lift any objects when overreaching ? Show miner Photo 2

Y/N

If NO, go to question 5 ii)

How many times in a shift do you lift objects when overreaching ? Show miner Card D

iii)

What is the longest time without a break that you spend repeatedly lifting objects when overreaching ? Show miner Card E

iv)

What is the total amount of time in a shift that you spend lifting objects when overreaching ? Show miner Card B

5.

i)

During a typical shift do you lift any objects above shoulder height ? Show miner Photo 3

If NO, go to question 6 ii)

How many times in a shift do you lift objects above shoulder height ? Show miner Card A

iii)

What is the longest time without a break that you spend repeatedly lifting objects above shoulder height ? Show miner Card B

Y/N

71

iv)

What is the total amount of time in a shift that you spend lifting objects above shoulder height ? Show miner Card C

6.

i)

During a typical shift do you lift any objects over an obstruction ?

Y/N

If NO, go to question 7 ii)

How many times in a shift do you lift objects over an obstruction ? Show miner Card A

iii)

What is the longest time without a break that you spend repeatedly lifting objects over an obstruction ? Show miner Card B

iv)

What is the total amount of time in a shift that you spend lifting objects over an obstruction ? Show miner Card C

7.

i)

During a typical shift do you lift any objects whilst kneeling ?

If NO, go to Section B ii)

How many times in a shift do you lift objects whilst kneeling ? Show miner Card D

iii)

What is the longest time without a break that you spend repeatedly lifting objects whilst kneeling ? Show miner Card E

iv)

What is the total amount of time in a shift that you spend lifting objects whilst kneeling ? Show miner Card B

y/N

72

Section B Interviewer to show miner photo 4 The questions in this section are about holding objects whilst standing still during a typical shift. Interviewer to show miner Card listing objects weighing more than 50kg. 1.

i)

During a typical shift do you, on your own, hold objects weighing more than 50kg ?

y/N

If NO, go to question 2 ii)

How many times in a shift do you hold objects weighing more than 50kg ?

Show miner Card D iii)

What is the longest time without a break that you spend holding any objects weighing more than 50kg ?

Show miner Card F

iv)

What is the total amount of time in a shift that you spend holding objects weighing more than 50kg ? Show miner Card B

Interviewer to show miner Card listing objects weighing more than 20kg. 2.

i)

During a typical shift, do you hold any object weighing more than 20kg with one hand ?

Y/N

If NO, go to question 3 ii)

How many times in a shift do you hold objects weighing more than 20kg with one hand ? Show miner Card A

iii)

What is the longest time without a break that you spend holding objects like this with one hand ? Show miner Card F

iv)

What is the total amount of time in a shift that you spend holding objects weighing more than 20kg with one hand ? Show miner Card B

n

73

Interviewer to show miner Photo 5 3.

i)

During a typical shift do you hold any object whilst stooped over ?

If NO, go to question 4 ii)

How many times in a shift do you hold objects whilst stooped over ? Show miner Card D

iii)

What is the longest time without a break that you spend holding an object like this ? Show miner Card F

iv)

What is the total amount of time in a shift that you spend holding objects whilst stooped over ? Show miner Card B

4.

i)

During a typical shift do you hold any objects away from your body ?

Y/N

If NO, go to question 5 ii)

How many times in a shift do you hold objects away from your body ? Show miner Card D

iii)

What is the longest time without a break that you spend holding an object away from you body ? Show miner Card F

iv)

What is the total amount of time in a shift that you spend holding objects away from your body ? Show miner Card B

5.

i)

During a typical shift do you hold any objects above shoulder height ?

If NO, go to Section C

Y/N

1 1

1 1

74

ii)

How many times in a shift do you hold objects above shoulder height ? Show miner Card D

Hi)

What is the longest time without a break that you spend holding an object above shoulder height ? Show miner Card F

iv)

What is the total amount of time in a shift that you spend holding objects above shoulder height ? Show miner Card B

Section C Interviewer to show miner Photo 6 The questions in this section are about carrying ie. holding a load when walking during a typical shift.

Interviewer to show miner card listing objects weighing more than 50 kg.

1.

i)

During a typical shift do you carry any objects weighing more than 50 kg ?

If NO, go to question 2 ii)

How many times in a shift do you carry objects weighing more than 50kg ? Show miner Card A

iii)

What is the longest time without a break that you spend carrying an object weighing more than 50kg ? Show miner Card F

iv)

What is the total amount of time in a shift that you spend carrying objects weighing more than 50kg ? Show miner Card B

75

Interviewer to show miner card listing objects weighing more than 20kg. 2.

i)

During a typical shift do you carry with one hand any object weighing more than 20kg ?

If NO, go to question 3 ii)

How many times in a shift do you carry objects weighing more than 20kg with one hand ?

Show miner Card D iii)

. What is the longest time without a break that you spend carrying an object weighing more than 20kg with one hand ?

Show miner Card F iv)

What is the total amount of time in a shift that you spend carrying objects weighing more than 20kg with one hand ?

Show miner Card B 3.

i)

During a typical shift do you carry any object whilst stooped over ? Show miner Photo 7

Y/N

If NO, go to question 4 ii)

How many times in a shift do you carry objects whilst stooped over ?

Show miner Card A iii)

What is the longest time without a K eak that you spend carrying an object whilst stooped over ?

Show miner Card F iv)

What is the total amount of time in a shift that you spend carrying objects whilst stooped over ?

Show miner Card B 4.

i)

During a typical shift do you carry any objects held away from your body ?

If NO, go to Section D

Y/N

76

ii)

How many times in a shift do you carry objects held away from your body ? Show miner Card D

iii)

What is the longest time without a break that you spend carrying an object held away from your body ? Show miner Card F

iv)

What is the total amount of time in a shift that you spend carrying objects held away from your body ? Show miner Card B

Section D The questions in this section are about pushing during a typical shift.

1.

i)

During a typical shift do you push any object whilst stooped over ?

••

.

If NO, go to question 2 ii)

How many times in a shift do you push objects whilst stooped over ? Show miner Card D

iii)

What is the longest time without a break that you c pend repeatedly pushing objects whilst stooped over ? Show miner Card E

iv)

What is the total amount of time in a shift that you spend pushing objects whilst stooped over ? Show miner Card B

2.

i)

During a typical shift do you push any object when overreaching ?

If NO, go to question 3 ii)

How many times in a shift do you push objects when overreaching ? Show miner card A

I

'

77

HO

What is the longest time without a break that you spend repeatedly pushing objects when overreaching ? Show miner Card B

iv)

What is the total amount of time in a shift that you spend pushing objects when overreaching ? Show miner Card C

3.

i)

During a typical shift do you push any objects which are above your shoulder height ?

If NO. go to question 4 ii)

How many times in a shift do you push objects above shoulder height ? Show miner Card A

iii)

What is the longest time without a break that you spend repeatedly pushing objects above shoulder height ? Show miner Card B

iv)

What is the total amount of time in a shift that you spend pushing objects above your shoulder height ? Show miner Card C

4.

i)

Do you twist your body when pushing any object ?

If NO, go to Section E ii)

How many times in a shift do you twist your body when pushing objects ? Show miner Card A

iii)

What is the longest time without a break that you spend repreatedly twisting your body when pushing objects ? Show miner Card B

iv)

What is the total amount of time in a shift that you spend twisting your body when pushing objects ? Show miner Card C

Y/N

78

Section E The questions in this section are about pulling during a typical shift. 1.

i)

During a typical shift do you pull any object whilst stooped over ?

Y/N

If NO, go to question 2 ii)

How many times in a shift do you pull objects whilst stooped over ? Show miner Card D

iii)

What is the longest time without a break that you spend repeatedly pulling objects whilst stooped over ? Show miner Card E

iv)

What is the total amount of time in a shift that you spend pulling objects whilst stooped over ? Show miner Card B

2.

i)

During a typical shift do you pull any object when overreaching ?

If NO, go to question 3 ii)

How many times in a shift do you pull objects when overreaching ? Show miner Card A

iii)

Wh^at is the longest time without a break that you spend repeatedly pulling objects when overreaching ? Show miner Card B

iv)

What is the total amount of time in a shift that you spend pulling objects when overreaching ? Show miner Card C

3.

i)

During a typical shift do you pull any objects which are above your shoulder height ?

If NO, go to question 4

Y/N

79

ii)

How many times in a shift do you pull objects above shoulder height ? Show miner Card A

iii)

What is the longest time without a break that you spend repeatedly pulling objects above shoulder height ? Show miner Card B

iv)

What is the total amount of time in a shift that you spend pulling objects above shoulder height ? Show miner Card C

4.

i)

During a typical shift do you twist your body when pulling any object ?

Y/N

If NO, go to section F ii)

How many times in a shift do you twist your body when pulling objects ? Show miner Card A

iii)

What is the longest time without a break that you • spend repeatedly twisting your body when pulling objects ? Show miner Card B

iv)

What is the total amount of time in a shift that you spend twisting your body when pulling objects ? Show miner Card C

Section F The questions in this section refer to the use of transport systems and heading machines underground during a typical shift. 1.

i)

Do you travel by man riding belt ?

y/N

If NO, got to question 2 ii)

How many times in a shift do you travel by man-riding belt ? Show miner Card G

iii)

What is the longest time without a break that you spend travelling by man-riding belt ? Show miner Card E

\

1

80

iv)

What is the total amount of time in a shift that you spend travelling by man-riding belt ? Show miner Card B

2.

i)

Do you travel by paddy ?

Y/N

If NO, go to question 3 ii)

How many journeys in a shift do you make by paddy ? Show miner Card G

III)

What is the longest time without a break that you spend travelling by paddy ? Show miner Card E

iv)

What is the total amount of time in a shift that you spend travelling by paddy ? Show miner Card B

3.

i) ii)

How many journeys in a shift do you make by F.S.V. ?

iii)

What is the longest time without a break that you spend travelling by F.S.V. ?

iv)

4.

Do you drive a free-steered vehicle ?

i)

••

Y/N

What is the total amount of time in a whole shift that you spend travelling by F.S.V. ?

Do you travel on a Becorit ?

Y/N

If NO, go to question 5 ii) iii) iv)

How many journeys in a shift do you make on a Becorit ? Hrs Mins

What is the longest time without a break that you spend travelling on board a Becorit ? What is the total amount of time in a shift that you spend travelling on a Becorit ?

'

Hrs Mins ' ' '

81

5.

i)

Do you operate a heading machine ?

Y/N

If NO, go to question 5 ii) iii)

iv)

How many times in a shift do you cut? Hrs Mins

What is the longest time without a break that you spend cutting or loading? What is the total amount of time in a shift that you spend cutting and loading ?

Section G Back Pain Questionnaire 1.

Have you had a pain or ache in your back during the past 12 months ? If YES, go to question 2

2.

-

If NO, go to question 5

Did you have the pain or ache in your back on most days of every month during the last year ?

Y/N

If YES, go to question 5 3.

Have you had more than one separate attack of back pain or ache in the last 12 months ?

Y/N

If NO, go to question 5 4.

In the last 12 months how many separate attacks of back pain or ache have yea had ?

5.

More than 12 months ago did you have any attacks of pain or ache in yQur back ?

Y/N

If NO, go to end 6.

Can you remember when you had your first attack of back pain ?

Y/N

If No, go to end

7.

When was it ? (enter as month and year eg. May 1986 as 0586)

That was the last question. Thank you for your help in answering these questions, it is greatly appreciated.

82

83

APPENDIX 3 Examples of items weighing more than 20kg and more than 50kg. Photographs of tasks

EXAMPLES OF ITEMS WEIGHING 50KG OR MORE (About 7.5 stones) 1 Arch-girder leg. 1 Metre of conveyor belt. 1 Crown. 1 Steel bar.

EXAMPLES OF ITEMS WEIGHING 20KG OR MORE (About 3 stones) 1 Stonedust bag. 1 Drum of oil. 1 Agalite block. 1 4' Piece of chockwood. 1 Pair of fishplates 4 Tie bars.

84

PHOTO 1.

TWISTING

PHOTO 2.

OVERREACHING

86

PHOTO 3.

HOLDING/LIFTING ABOVE SHOULDER HEIGHT

PHOTO 4.

HOLDING

88

PHOTO 5.

HOLDING WHILST STOOPED OVER

PHOTO 6.

CARRYING

90

PHOTO 7.

CARRYING WHILST STOOPED OVER

91

APPENDIX 4 Cards with responses ranges for tasks CARD A 1. Less than 20 times per shift 2. 20 to 50 times per shift 3. 50 to 100 times per shift 4. More than 100 times per shift CARD B 1. Less than 15 minutes 2. 15 to 20 minutes 3. 30 minutes to 1 hour 4. More than 1 hour CARD C 1. Less than 30 minutes 2. 30 minutes to 1 hour 3. 1 to 3 hours 4. More than 3 hours CARD D 1. Less than 10 times per shift 2. 10 to 30 times per shift 3. 30 to 50 times per shift 4. More than 50 times per shift CARD E 1. Less than 5 minutes 2. 5 to 15 minutes 3. 15 to 30 minutes 4. More than 30 minutes CARD F 1. Less than 1 minute 2. 1 to 3 minutes 3. 3 to 5 minutes 4. More than 5 minutes CARD G 1. Once per shift 2. Twice per shift 3. 2 to 5 times per shift 4. More than 5 times per shift.

92

93

APPENDIX 5 COMPARISON OF THE DATA FROM THE OBSERVATION OF 10 MEN UNDERGROUND AT COLLIERY 1 WITH INFORMATION FROM THEIR BACK PAIN QUESTIONNAIRES Methods Data on the observations made at colliery 1 by an ergonomist, were examined in relation to the information given in the questionnaire by the 10 men involved. The frequencies and durations which make up the observational data were grouped into the same categories as those used in the completion of the questionnaire, allowing a direct comparison of the two data sets. Comparisons were made at the following levels: 1. 2. 3. 4.

Whether the task was reported as being carried out and whether it was observed Whether the frequency category from questionnaire and observation coincided. Whether the longest duration without a break category from questionnaire and observation coincided Whether the total duration category from questionnaire and observation coincided.

In the cases of 2, 3 and 4 the direction of the deviation from the observed was of particular interest. A two-way table is presented at the end of this section for each task concentrated on in the observation, showing whether individuals had recorded on their questionnaire that they performed the task, and whether they were observed carrying out the task. These tables give an impression of how often an individual's reporting coincides with what was observed. These should be interpreted with care since any deviation may be due to the fact that observation was not carried out during a 'typical 1 shift. Another possibility is that an individual's impression of a typical shift when answering the questionnaire may be inaccurate. Results Non-observation of tasks previously reported in the questionnaire was fairly common (eg.7 individuals reported twisting the body while lifting but were not observed doing so). However observation of tasks NOT reported in the questionnaire was relatively uncommon; for 9 of the 16 tasks there was no occurrence of this. There was good agreement between questionnaire and observation on aspects of travel (ie.using man riding belt or paddy), probably because such activities are well defined. In cases where reporting and observation coincided (ie. when men said they did a certain task and were observed doing it), the accuracy of estimating frequency, longest duration and total duration was examined. In almost all tasks there was some degree of over-estimation in each of frequency, longest duration and total duration. Only in using man riding belts did estimates fall below those observed. This is indicated for each task after the corresponding table.

94

For most tasks the frequency was over-estimated fewer times than longest and total duration, the exceptions being travelling by man riding belt and paddy where the reverse was true. Frequency estimations were usually out by only one, or occasionally two, categories eg. '20 to 50 times per shift' recorded in the questionnaire as opposed to Mess than 20 times per shift' during observation. Longest duration and total duration were more often estimated two or three categories higher than observed. If we assume that the observation was carried out during fairly typical shifts, then we might conclude that there is a tendency for individuals to report tasks that they do not actually carry out. It may be the case that individuals are considering whether they EVER carry out such a task rather than whether they do it TYPICALLY. It is encouraging that on very few occasions were individuals observed doing tasks they reported they did not do during a typical shift. When a task is correctly reported the related quantities, durations in particular, seem to be consistently over-estimated. An assessment was made as to whether there were differences in accuracy between cases and controls. Since each of the men observed had completed a questionnaire their case/control status had already been determined. Cases would have recently experienced a back problem which had resulted in a period of absence, and this may have heightened their awareness of the tasks they perform. Three of the 10 men observed were cases and the remainder were controls. A comparison was made, for each task component, between the proportion of cases and controls whose reporting of tasks carried out agreed with the observation. For all 10 task components where data existed for comparison, a larger proportion of cases than controls gave coinciding information. Controls were more likely to have reported tasks but not been observed carrying them out. With regards to the recording of actual frequency and durations, cases seemed slightly more accurate in this respect also. Across all the task components agreement between questionnaire and observation in these respects was as follows: Frequency Duration Longest duration

CASES 50% 37% 20%

CONTROLS 38% 23% 15%

Discussion As mentioned before disagreement was almost always in the direction of over-estimation. This is perhaps an indication of more awareness about tasks carried out among the cases. Cases seem to be more able to identify tasks and to estimate associated quantities accurately. The effect of such a phenomena in an analysis relating the performance of certain tasks with the probability of being a case would be to give conservative results. Although this comparison is based on only a very small sample of the study population it is useful to consider the findings when interpreting the main study results. In a previous study of tasks in coalminers Agius et al. (1988) identified significant variability between and within workers. The differences between questionnaire responses and observations noted in this current validation exercise could also be explained by within worker variability of tasks performed.

95

Tables comparing reporting and observation of tasks underground: LIFT 50KG TASK OBSERVED? yes no total TASK REPORTED?

yes no

3 0

5 2

total

8 2 10

Reported and observed agree 2 0 0

FREQUENCY LONGEST DURATION TOTAL DURATION

Reported higher than observed 1 3 3

Reported lower than observed 0 0 0

Reported higher than observed 3 3 3

Reported lower than observed 0 0 0

LIFT AND TWIST TASK OBSERVED? yes no total TASK REPORTED?

yes no

3 0

7 0

10 0

total

3

7

10

FREQUENCY LONGEST DURATION TOTAL DURATION

Reported and observed agree 0 0 0

96

LIFT AND OVERREACH TASK OBSERVED? yes no total TASK REPORTED?

yes no

6 0

2 2

8 2

total

6

4

10

Reported and observed agree

2 2 1

FREQUENCY LONGEST DURATION TOTAL DURATION

Reported higher than observed 4 4 5

Reported lower than observed 0 0 0

Reported higher than observed 1 4 4

Reported lower than observed 0 0 0

LIFT ABOVE SHOULDER TASK OBSERVED? yes no total TASK REPORTED?

yes no

5 0

4 1

9 1

total

5

5

10

FREQUENCY LONGEST DURATION TOTAL DURATION

Reported and observed agree 4 1 1

97

LIFT OVER OBSTRUCTION TASK OBSERVED? yes no total TASK REPORTED?

yes no

3 0

6 1

9 1 10

total

Reported and observed agree 1 0 0

FREQUENCY LONGEST DURATION TOTAL DURATION

Reported higher than observed 2 3 3

Reported lower than observed 0 0 0

Reported higher than observed

Reported

LIFT WHILE KNEELING TASK OBSERVED? yes no total TASK REPORTED?

yes no

1 0

3 7

2 7

10

total

Reported and observed agree FREQUENCY LONGEST DURATION TOTAL DURATION

0 0 0

1 1 1

1 owe r t han observed

0 0 0

98

PUSH AND STOOP TASK OBSERVED? yes no total TASK REPORTED?

yes no

4 0

5 1

9 1

total

4

6

10

Reported and observed agree 1 0 0

FREQUENCY LONGEST DURATION TOTAL DURATION

Reported higher than observed 3 4 4

Reported lower than observed 0 0 0

Reported higher than observed 1 3 3

Reported 1 owe r t han observed 0 0 0

PUSH AND OVERREACH TASK OBSERVED? yes no total TASK REPORTED?

yes no

5 1

1 3

6 4

total

6

4

10

FREQUENCY LONGEST DURATION TOTAL DURATION

Reported and observed agree 4 2 2

99

PUSH ABOVE SHOULDER TASK OBSERVED? yes no total TASK REPORTED?

yes no

3 1

1 5

4 6

total

10

Reported and observed agree

observed

1 2 3

2 1 0

FREQUENCY LONGEST DURATION TOTAL DURATION

Reported

hi ghe r t han

Reported lower than observed

0 0 0

PUSH AND TWIST TASK OBSERVED? yes no total TASK REPORTED?

yes no

total

FREQUENCY LONGEST DURATION TOTAL DURATION

0 0

6 4

6 4

10

10

Reported and observed agree 0 0 0

Reported higher than observed 0 0 0

Reported lower than observed

0 0 0

100

PULL AND STOOP

TASK OBSERVED? yes no total TASK REPORTED?

yes no

4 1

3 2

total

7 3 10

Reported and observed agree 1 0 0

FREQUENCY LONGEST DURATION TOTAL DURATION

Reported higher than observed 3 4 4

Reported 1 owe r t nan observed 0 0 0

Reported higher than observed 0 1 1

Reported lower than observed, 0 0 0

PULL AND OVERREACH TASK OBSERVED?

yes TASK REPORTED?

yes no

total

FREQUENCY LONGEST DURATION TOTAL DURATION

1 1

no

7 1

total

8 2 10

Reported and observed agree 1 0 0

101

PULL ABOVE SHOULDER TASK OBSERVED?

yes TASK REPORTED?

yes

no

1 1

no

5 3

total

total

6 4 10

Reported and observed agree 1 1 1

FREQUENCY LONGEST DURATION TOTAL DURATION

Reported higher than observed 0 0 0

Reported lower than observed 0 0 0

Reported higher than observed 1 1 1

Reported lower than observed 0 0 0

PULL AND TWIST TASK OBSERVED?

yes TASK REPORTED?

no

total

yes no

1 2

7 0

8 2

total

3

7

10

FREQUENCY LONGEST DURATION TOTAL DURATION

Reported and observed agree 0 0 0

102

MAN RIDING BELT TASK OBSERVED? yes no total TASK REPORTED?

yes no

8 0

1 1

9 1

total

8

2

10

Reported and observed agree

3 7 3

FREQUENCY LONGEST DURATION TOTAL DURATION

Reported higher than observed 4 1 1

Reported lower than observed 1 0 4

Reported

Reported lower than observed

PADDY TASK OBSERVED? yes no total TASK REPORTED?

yes no

1 1

0 8

1 9

total

2

8

10

Reported and observed agree FREQUENCY LONGEST DURATION TOTAL DURATION

0 1 1

h i ghe r t nan observed

1 0 0

0 0 0

(A20115) IOM (R) ReportCov art

3/15/06

12:32 PM

Page 2

HEAD OFFICE:

Research Avenue North, Riccarton, Edinburgh, EH14 4AP, United Kingdom Telephone: +44 (0)870 850 5131 Facsimile: +44 (0)870 850 5132 Email: [email protected]

Tapton Park Innovation Centre, Brimington Road, Tapton, Chesterfield, Derbyshire, S41 0TZ, United Kingdom Telephone: +44 (0)1246 557866 Facsimile: +44 (0)1246 551212

Research House Business Centre, Fraser Road, Perivale, Middlesex, UB6 7AQ, United Kingdom Telephone: +44 (0)208 537 3491/2 Facsimile: +44 (0)208 537 3493

Brookside Business Park, Cold Meece, Stone, Staffs, ST15 0RZ, United Kingdom Telephone: +44 (0)1785 764810 Facsimile: +44 (0)1785 764811