Use of exposure data in occupational reproductive studies.

Accurately characterizing worker exposure presents some unique problems in occupational reproductive studies. The latency period between exposure and the adverse event may be relatively brief - days, weeks, or months. Inaccurately identifying the specific exposure period which initiated the adverse event may then contribute to sizeable misclassification errors. This paper discusses different types of exposure models that can be used to help reduce misclassification and delineate the exposure-effect relationships more accurately. The advantages and disadvantages of exposure data sources are also discussed.