Epidemiologic investigation of occupational carcinogenesis using a serially additive expected dose model.

The epidemiologic identification of occupational carcinogens is complicated by several problems including worker mobility between jobs, variation over time of chemicals and processes used, and the long latency period between exposure and discovery of a tumor. In the light of these problems, a method using the cumulative dose concept has been developed which involves calculating the expected yearly exposure for each case from work histories of all noncases close to the case in year of birth and year of hire. The data required for use of the method include information concerning exposure to the chemicals being studied for each job in each calendar year of the study. Use of the method is illustrated with a study of angiosarcoma of the liver and vinyl chloride exposure in a polymerization plant. The value of the method lies in the wealth of information generated concerning the association between chemical exposures and cancer, including exposure level relationships, latency information, and the possibility that two chemicals might be acting independently or jointly. The serially additive expected dose model is likely to prove particularly useful in the analysis of data collected by occupational health surveillance systems, as well as retrospective studies of the type illustrated.