Estimating Historical Exposures to Formaldehyde in a Retrospective Mortality Study

Abstract

Dose-response is an important consideration in any epidemiologic study. This paper describes the development of exposure classifications for 30,000 persons exposed to formaldehyde since the early 1930's. The general approach was to make initial estimates of historical exposures from existing data, to monitor current exposures, to have plant industrial hygienists review the ranking of the jobs, and to review all three sources of information to develop final estimates. Job histories were abstracted from personnel files. The job titles from the abstracts were standardized and aggregated, reducing the number from 300,000 to 6700 job titles. Walk-throughs of the plants were made to gather historical and current production, control, and monitoring data. Historical exposure information for each job was coded and computerized. A list of the jobs, ranked by level of formaldehyde exposure, was generated and sent to each workplace for review. While the exposures were being coded, approximately 2000 air samples were taken. The samples were used to standardize the plants' monitoring data, to determine exposures of jobs which had never been monitored, and to supplement the historical estimates. The current sampling results, the employer comments, and the historical estimates were reviewed by the study industrial hygienists, and final estimates were derived. The review by the plants provided less information than expected, probably due to the volume of material. The monitoring, on the other hand, provided valuable data which in some cases confirmed the exposure estimates and in other cases caused the estimates to be changed. Stewart, P. A.; Blair, A.; Cubit, D. A.; Bales, R. E.; Kaplan, S. A.; Ward, J.; Gaffey, W.; O'Berg, M. T.; Walrath, J. Estimating historical exposures to formaldehyde in a retrospective mortality study.