Each epidemiologic study was abstracted by consultants to the committee, and the abstracted data for each study were entered into a computerized data base. The first use of this data base was to sort the studies by health outcome. Completed abstracting forms and studies were then sent to committee members with expertise in that field, who reviewed and verified the abstracted information. Throughout the study, this data base was used extensively to aid committee members in locating relevant studies on a variety of questions ranging from methods of exposure assessment to study design.