Triaging jobs in a community-based case-control study to increase efficiency of the expert occupational assessment method.

OBJECTIVES: Expert assessment is useful to assess occupational exposures in cases where measured exposure data are not available. However, the process may be inefficient in a community-based study with low prevalence of exposure. This study aimed to determine if formally triaging the jobs as to likelihood of exposure before the experts review those jobs could improve study efficiency. METHODS: One thousand nine hundred and sixty-one jobs from a case-control study were triaged by study staff (non-occupational health professionals) into four groups depending on the likelihood of exposure to solvents. For jobs in one group, we had additional information available in the form of job-specific modules and automatic exposure assignments for solvents based on rules pre-programmed into the job-specific module. After the automatic assignment, two experts reviewed the jobs to assign exposure to solvents in order to evaluate the process. The prevalence of exposure and the agreement between the two raters and between the raters' and the automatic assignments were compared for the four triage groups. RESULTS: The majority of jobs (76%) were triaged as unexposed by study staff and very few of these jobs were assigned as exposed by the raters (1%). For jobs with automatic assignment (18% of total), the raters tended to agree with the automatic assignment if that assignment was unexposed or probably exposed. There was less agreement for jobs in which the automatic assignment was possible exposure. For jobs triaged as ones with potential exposure based only on job title but with no further information available, the level of disagreement between the raters tended to be higher. CONCLUSIONS: Formal triaging of jobs can improve the efficiency of the expert assessment process. Of the 75% of jobs initially triaged as unexposed, virtually no exposures were found, and omitting manual review of this group would save considerable time.