Cancer-Attributable Costs of Diagnosis and Care for Persons ith Screen-Detected Versus Symptom-Detected Colorectal ancer

m o w ackground & Aims: Colorectal cancer screening is efective and cost-effective, but little data from health lan settings are available inform decision-makers rearding direct economic implications of colorectal caner screening programs. The purpose of this study was o compare the prediagnosis evaluation and first-year reatment costs of persons diagnosed with colorectal ancer, stratified by whether the cancer was detected by creening using fecal occult blood testing or evaluation f symptoms. Methods: This retrospective study anayzed persons diagnosed with colorectal cancer from 993 to 1999 in Group Health Cooperative, a large ealth maintenance organization in Washington state. otal health care costs during 3 months before and 12 onths following diagnosis were compared for screenetected versus symptom-detected individuals. Results: uring this time, 206 cancers were detected by screenng and 717 by symptoms. In the 3 months before iagnosis, total costs were $7346 for persons with creen-detected versus $10,042 for those with sympom-detected cancer (P < 0.01). Stratified by stage, iagnosis costs were significantly lower for persons with tage B cancer ($7282 vs. $11,682; P < 0.01) and onsignificantly lower for other stages. A total of 53% of creen-detected cases were Dukes’ stage A or in situ at iagnosis versus 30% of symptom-detected cases (P < .01). Overall costs were lower for the screen-detected roup in the 12 months following diagnosis ($22,369 vs. 29,471; P < 0.01). Conclusions: Colorectal cancer creening can substantially reduce prediagnosis evaluaion costs. These savings are of interest to health plans nd should be factored into cost-effectiveness evaluaions of screening programs.