Control group selection in studies involving compensation for occupational diseases.

Many occupational diseases which are compensable are of interest in epidemiological studies. In addition, diseases of interest which are not compensable may be discovered during the evaluation of eligibility for compensation. Certain peculiarities of the group of workers who seek compensation (those who receive compensation and those who do not) may make the selection of an appropriate control group difficult. In two recent studies, problems associated with finding controls for cases with a compensable disease and a disease discovered during the evaluation of eligibility for compensation pointed out some of these potential biases. This paper describes the results obtained using different control groups for these studies, proposes reasons for differences in the results and suggests steps which can be taken to ensure the validity of a control group in such studies.