RESEARCH REPORT 164 - Health and Safety Executive

1 HSE Health & Safety Executive Questionnaire predictors of asthma and occupational asthma Prepared by Institute of Occupational Medicine for the Health and Safety Executive 2003 RESEARCH REPORT 164 HSE Health & Safety Executive Questionnaire predictors of asthma and occupational asthma BG Miller, MK Graham, KS Creely, HA Cowie, CA Soutar Institute of Occupational Medicine 8 Roxburgh Place Edinburgh EH8 9SU This work was designed to investigate the identification by questionnaire of asthma in general, and work-related asthma in particular, to assist in producing a questionnaire to identify work-related asthma. A previous postal survey of a community in central Scotland provided a complete data set for almost 12,000 subjects of working age, with responses to a range of questions on respiratory symptoms and their relationship to occupation.

2 For those subjects who had recorded the occupation held when symptoms began, the existing data were augmented by an exposure assessment to characterise exposures typical of those occupations. Certain symptoms were associated, tending to occur together, and a number of syndromes were defined in terms of these. Some limited associations between these syndromes and occupational exposures were identified, particularly with irritant and highly reactive chemicals, and exposure to proteins and other organic substances. Occupational histories from the postal survey have not yet been processed and future work on could provide further insight into occupational associations of syndromes.

3 The syndromes suggested a much higher prevalence of work-related asthmatic symptoms than reported by the Labour Force Survey questions, but the reasons for this excess were not clear. For a few subjects, clinical assessments had been made, but from an initial look the agreement with the questionnaire-based syndromes was fairly poor. More intensive reworking of the clinical data revealed relationships with childhood asthma and exposure to known asthmagens that may help to guide future questionnaire design. This REPORT and the work it describes were funded by the Health and Safety Executive (HSE). Its contents, including any opinions and/or conclusions expressed, are those of the authors alone and do not necessarily reflect HSE policy.

4 HSE BOOKS Crown copyright 2003 First published 2003 ISBN 0 7176 2759 4 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means (electronic, mechanical, photocopying, recording or otherwise) without the prior written permission of the copyright owner. Applications for reproduction should be made in writing to: Licensing Division, Her Majesty's Stationery Office, St Clements House, 2-16 Colegate, Norwich NR3 1BQ or by e-mail to ii 1234567 CONTENTS SUMMARY v 1. INTRODUCTION 1 Background 1 Previous studies 1 Legacy of data 2 Unanswered questions & needs 2 2.

5 OBJECTIVES 3 Detailed objectives 3 Plan of work 3 3. METHODS 5 Literature review 5 Data extraction & checking 5 Multivariate analyses: clusters and factors 13 Exposure Assessment 14 Statistical analysis of associations 20 4. RESULTS 21 Literature review 21 Analysis of symptoms data 25 Definition of syndromes 31 Exposures 37 Clinical records 45 5. DISCUSSION 52 5. Quality of available data 52 5. Exposure assessment 52 5. Symptom combinations 52 5. Defining syndromes 53 5. Associations between syndromes and occupational exposures 53 5. Comparisons with clinical examinations 54 5.

6 Design of questionnaires 54 iii Possibilities for future work 54 6. ACKNOWLEDGEMENTS 56 7. REFERENCES 58 APPENDIX 1 60 APPENDIX 2 70 APPENDIX 3 74 APPENDIX 4 76 APPENDIX 5 78 iv SUMMARY Introduction Reported incidence of occupational asthma is greater than for other common occupational respiratory disorders. Epidemiological studies of occupational asthma rely on questionnaires on respiratory symptoms, but there is little agreement or standardisation on what questions best reveal occupational asthma. A community study of respiratory Health was carried out in central Scotland in the mid 1990s, in which almost 17,000 adult subjects completed postal questionnaires on self-reported work related illness, respiratory symptoms, smoking habit and occupation.

7 A small number of these (101) attended clinical examinations. The data from this study represent an important resource for understanding the relationships between symptom questions and their relationship with occupational disease. We REPORT on a project involving detailed reanalysis of this data set, followed by a comparison with the clinical assessments. The study was intended to provide information towards the development of a questionnaire to identify work related asthma. Objectives The detailed objectives were: i. To investigate which groups of respiratory symptoms tend to be reported together. ii. To identify clinically relevant syndromes based on these symptom clusters.

8 Iii. To investigate whether and how these syndromes differentiate between individuals who were defined to have work-related and non-work related asthma in the previous questionnaire study. iv. To investigate whether and how these syndromes differ by occupation and exposure, with particular reference to differences between probable irritant and allergic asthma. v. To identify symptom combinations associated with clinical diagnosis of work-related asthma, and compare these to the syndromes identified by clustering. vi. To identify any aspects of the clinical examination and diagnosis process that could usefully be incorporated into a self-administered questionnaire.

9 Methods Recent literature on development and validation of questionnaires to identify asthma and other respiratory diseases was collected and reviewed. Questionnaire data from the previous community study were extracted from computer files. Only the first two mailings used the complete questionnaire, giving data for 14,090 working age adult subjects. Of these, 2,149 were omitted from statistical analysis because of missing data on individual questions. The data from the community study included the job that was current when respiratory symptoms had first been experienced, and these had previously been coded to SIC and SOC classifications.

10 Occupations were grouped and characterised in terms of likely exposure to a v range of important asthmagens and respiratory irritants. Indices of exposure included presence or absence of each hazard, and likely frequency of exposure. An additional assessment was made of whether there was at least a moderate probability of accidental or periodic exposure to very high levels of irritant gases or fumes. Statistical analyses of patterns amongst the symptoms used principal components and factor rotation to identify combinations of symptoms to form syndromes. Separate analyses used clustering techniques on the symptom variables, to identify combinations of responses that tended to occur together.