The primary purpose of this study is to assess the effectiveness of a colorectal screening program in a large defined population of HMO members using intermediate endpoints with known links to the subsequent development and progression of cancer. The screening program, which was implemented in 1985, consists of standardized laboratory hemoccult testing, a diagnostic algorithm for follow-up of positive hemoccults, physician education, and computerized monitoring through a Tumor Registry Information System. The effectiveness of the screening program in increasing proportion of asymptomatic diagnoses and early stage disease will be assessed through pre-program/post-program time series and case/control analyses. The study data is derived from the medical records of 1,100 cases diagnosed during 1980-1988 and from computerized Tumor Registry, inpatient and outpatient information systems. Population and subgroup analyses will be used to estimate average lead time associated with the screening and to assess the effects of length bias sampling. Direct and indirect costs of the screening program will be estimated and the cost-effectiveness of the program in producing various outcomes will be assessed.
