Population colorectal cancer screening with fecal occult blood test.

Screening for early detection of colorectal cancer using fecal occult blood testing has been shown to be effective in reducing mortality from this disease. The largest Health Maintenance Organization in Israel initiated the use of Hemoccult Sensa in 1992 to evaluate the field performance of this test. All primary care physicians were invited to order home-based tests for their asymptomatic patients 50-74 years of age. This report summarizes the results of 45,166 tests performed, 22,193 in the prevalence round. Seventy-eight cancers, 60 patients with adenomas, and 163 patients with polyps were detected, yielding a cancer detection rate of 2.61/1,000 screened in the prevalence round. Of these, 21.6% were in the right colon. Of screen-detected cancers, 44.5% and 58.9% were detected in Dukes' A and in situ stages in the prevalence and incidence rounds, correspondingly. The overall estimated sensitivity of the test (median follow-up, 35 months) was 85.3% for the prevalence round with a specificity of 95.5%. The sensitivity for left-side tumors (87.9%) was higher than for right-side or rectal tumors (78.6%). The positive predictive value for cancer increased with increasing number of positive fields. Four or more positive fields had a positive predictive value for cancer of 16-26% and a positive predictive value of 46-71% for all tumors combined. Population screening with a sensitive fecal occult blood test performs well outside a trial setting, detecting a high proportion of expected tumors with favorable stage distribution. Given its proven power to significantly reduce mortality, use of this test is strongly advised to both medical organizations and the healthy population at average risk.