Exposure Assessment for a Study of Workers Exposed to Acrylonitrile. II. A Computerized Exposure Assessment Program

Publisher Website
Google Scholar
Abstract
The validity of exposure assessment in retrospective epidemiologic studies has often been questioned, in part because systematic methods have not been developed to estimate historical exposure levels when insufficient monitoring results are available. In addition, documentation as to how the exposure estimates are derived is often lacking. A previously published report described a data management system, called Job Exposure Profiles, that organized and maintained job-related information available on each job in a mortality study of workers exposed to acrylonitrile. This database, and a second one containing almost 19,000 acrylonitrile monitoring results with accompanying documentation (job, department, date, type of sample, etc.), were linked to a computerized, interactive exposure assessment program (EAP) to allow the user to develop historical exposure estimates. The EAP is the subject of this report. In this program, monitoring results were sorted into various subfiles based on their quantity and quality. In cases where monitoring results meeting specified criteria were available, the EAP directly calculated job/department means of the monitoring results for each time period over which there were no significant changes that affected exposure levels in the workplace environment. Where monitoring results were available but did not meet the criteria, the EAP allowed the user to select from various estimation methods. Some of these methods used the remaining monitoring results, while others modified estimates previously developed for other jobs or time periods. From these data and other information provided by the user, the computer program calculated the estimates for the job of interest. This software program is unique in that it provided a structured mechanism for the development of historical exposure estimates and simultaneously allowed the user access to a variety of estimation methods based on the available data. It allowed easy access to, and use of, thousands of monitoring results. Finally, it documented all decisions made by the user. This type of program can be used in other retrospective epidemiologic studies where limited monitoring data are available.