Reproducibility and diagnostic value of Hemoccult-II test. A colonoscopic evaluation in asymptomatic patients.

Inter-observer agreement rate in double readings of 246 Hemoccult-II tests was 0.88 (Kappa analysis). The slides were not rehydrated. The Hemoccult-II test was performed immediately before 748 total colonoscopies in asymptomatic patients after previous polypectomy or radical surgery for cancer to estimate the diagnostic value of the test in mass screening, accepting a higher prevalence of neoplasia in the study group but realizing that colonoscopy can only be performed in high-risk groups. Another purpose was to investigate the possibility of replacing colonoscopy with Hemoccult-II in surveillance after previous adenoma or cancer. The 79 colonoscopies after a positive test showed cancer in 3, adenomas in 13, and other intestinal pathology in 34 cases. The 669 colonoscopies after negative tests showed cancer in 1, adenomas in 67, and other pathology in 141 cases. In conclusion, the test cannot be used as a guideline for colonoscopy in follow-up programs for patients who have had adenomas, since it missed 84% of the new cases in the present series. The test does not exclude the presence of colorectal carcinoma, but the risk of cancer is probably several times less in patients with negative tests (1 of 669) than in patients with positive tests (3 of 79), which may justify the use of the Hemoccult-II test for mass screening in asymptomatic populations.