Screening for colorectal cancer: A comparison of 3 fecal occult blood tests

Background: Colorectal cancer is the second leading cause of cancer deaths in the United States. Fecal occult blood testing has become a standard screening test for large-bowel cancers in the average asymptomatic population. Performance characteristics of the test and physician and participant compliance are the 2 major elements that impact the success of screening and early detection. Objectives: To evaluate the nonhydrated Hemoccult, rehydrated Hemoccult, and Hemoccult SENSA tests (SmithKline Diagnostics Inc, Palo Alto, Calif) and to assess participant and physician compliance. Methods: A mass community-based screening study in an urban setting. Kits were distributed by a local pharmacy and at community sites. Diagnostic tests were completed through physicians' offices and clinics. Participants were asymptomatic and aged 50 years or older. Those who tested positive were advised to follow up with a physician. Results: An overall positivity rate of 16% was reported for the 8293 kits that were processed. Rehydrated Hemoccult had a positivity rate of 15%; Hemoccult SENSA, 7%; and nonhydrated Hemoccult, 5%. The positive predictive value of nonhydrated Hemoccult was 14%; rehydrated Hemoccult, 7%; and Hemoccult SENSA, 11%. Of those who tested positive, 59% had a colonoscopy or flexible sigmoidoscopy and double-contrast barium enema examination on follow-up. Recommended follow-up was more frequent for those who consulted a gastroenterologist. Conclusions: Rehydrated Hemoccult yielded a higher positivity rate and lower positive predictive value than either Hemoccult SENSA or nonhydrated Hemoccult. Hemoccult SENSA approached the positive predictive value of nonhydrated Hemoccult. Adequacy of follow-up of patients testing positive for fecal occult blood needs improvement. Arch Intern Med. 1997;157:970-976