Background: Job Exposure Matrix for exposures associated with occupational asthma
What does this tool to do?
This job-exposure matrix was developed, in order to allow researchers to use job title information obtained in respiratory disease questionnaires, to assign exposure risk groups, to survey participants, based on their job that places them at high risk for the development of occupational asthma.
What is included in the matrix?
This tool is two dimensional matrix with job codes on one axis (rows) and 23 exposure risk categories on the other axis (columns).
The ‘job’ axis of the matrix contains 506 job codes from the ILO International Standard Classification of Occupations: ISCO-88, published 1990. All submajor, minor, and unit group job titles are included; major group codes are excluded. These should be recoded to appropriate submajor or more precise codes prior to use of the matrix. The ISCO-88 codes and detailed descriptions (book or computer data file) can be obtained from: http://www.ilo.org/public/english/bureau/stat/isco/index.htm
The ‘exposure’ axis of the matrix contains:
- High risk: 18 columns for known risk factors for occupational asthma
- Low risk: 4 columns (1 for ‘low exposure to asthmogen’, 3 for other exposures linked to respiratory disease but NOT related to occupational asthma)
- Non exposed: 1 columns identifying jobs unlikely to be linked to exposures relevant for occupational asthma
- 2 columns indicating jobs that could benefit from a re-evaluation step (verification variables, see below)
- 1 column of detailed comments describing how to carry out the re-evaluation (comments)
- 1 column indicating jobs for which all exposure estimates are likely poor, even after re-evaluation (confidence)
A tablular version of the exposure axis of the matrix can be seen here.
How were the exposure categories chosen?
The risk factors for occupational asthma include both specific agents and groups of agents or settings, because most standardized job codes are too broad to allow valid exposure estimates for many of the specific agents linked to asthma The starting point for grouping was the categorization scheme of Chan-Yeung and Malo (Table of the major inducers of occupational asthma. In: Bernstein IL, Chan-Yeung M, Malo JL, et al, eds. Asthma in the workplace. New York: Marcel Dekker, 1993:595–623.)in which over 150 chemical and biological substances are stratified into high and low molecular weight (MW) agents. We included both specific agents and mixed environments because the exposure axis was based on both current knowledge about risk factors and the practical constraints imposed by the job coding system.
The main focus of this tool is identifying risks associated with occupational asthma. However, in response to specific requests from other researchers, additional columns are included that classify jobs on SOME risk factors for other non-asthma respiratory disease (ie. vehicle exhaust, environmental tobacco smoke, and possible exposure to other non-asthmagenic irritants). A final column identifies jobs unlikely to be associated with asthma or other respiratory disease risks.
What criteria were used to classify a job as exposed for each category?
Each cell in the matrix contains a yes/no classification for exposures risk (yes coded as 1, no coded as 0). For the asthmagens, because previous research has shown that job exposure matrices are optimized when specificity is favoured over sensitivity, a job was classified as exposed (ie.coded ‘1’) only it that job was assessed as having a strong likelihood of exposure relevant to occupational asthma. A strong likelihood means that the probability of exposure was expected to be high for a significant number of subjects having that job. For vehicle exhaust and ETS, exposure classification was also based on specificity (ie. jobs were only classed as exposured if the probability of exposure was high); for the « possible irritant exposure » column, sensitivity was favoured (ie. jobs tended to classified as exposed if the risk of exposure was considered even modest).
How confident are you in the exposure assignments? and how is this confidence indicated in the matrix?
Our experience indicate that the utility of the tool is enhanced when application of the matrix alone is followed by several steps to review, recode, and possibly re-assignment exposures for some jobs. Therfore, the matrix also contains three columns in the exposure axis that relate to the quality of the information. Two of these identify for job codes which we felt, a priori, could benefit from a second look at the textual job history. This includes job codes that are often imprecisely coded, those for which exposures could differ greatly by industry (e.g. production managers, painters) and others for which the job codes were not sufficiently precise (e.g. technicians, nurses). For each of these codes, detailed comments are included to describe the type of re-evaluation needed. Exposure estimates for subjects classified into this category required a second step (see below).
A final column in the matrix identifies job codes for which the exposure estimates generated by the matrix are likely to remain imprecise even after application of the re-evaluation step.
« Version 5.2, Dec 2004 – changes to exposure coding and arrangement from V 5.1 based on comments from Zock et al (European Community Respiratory Health Survey – Occupational Working Group, July 2004) see attached worksheet for list of changes »