Retrospective evaluation of occupational exposures in cancer epidemiology: A European Concerted Action of Research

Assessing exposures in the investigation of occupational cancer is difficult because of the historical nature of exposures. Yet, misclassification of exposures can have a profound effect on the interpretation of risk. It is therefore important to reduce misclassification as much as possible. In an effort to improve our ability to assess exposures, the European Concerted Action of Research has identified three areas which will be given the attention of researchers: elaboration of job-exposure matrices (JEMs), methods to build ad hoc and general questionnaires, and statistical issues. JEMs should be developed for specific industries after evaluation of individual worksites; improved by using standardized coding systems and by identifying level, probability, calendar time, etc.; and evaluated by comparing their performance with other methods of assessment. Questionnaires for specific exposures, methods to code information obtained from questionnaires to allow better evaluation of exposures, and quality control procedures should be developed. In addition, evaluations of exposures using questionnaires should be compared to evaluations using JEMs. Statistical measures should be developed that evaluate exposure assessment methods, including evaluations of accuracy, i.e., how well a method compares to the real exposure; evaluations with other methods; and determinations of whether a method is able to reproduce well-known relationships. Ways of analyzing epidemiologic data should also be developed including indices of cumulative exposure and the use of probability of exposure.