Fecal occult blood screening for colorectal cancer in a Veterans Administration hospital

Abstract

Over a 3.5 year period beginning September 1978, we gave fecal occult blood test kits to 13,522 outpatient applicants for medical care. Two thousand nine hundred sixty-four (22 percent) kits were returned completed according to instructions. One or more specimens were positive in 135 (5 percent) of the tests returned. One hundred twenty of the 135 patients with positive results were investigated by history, physical examination, digital rectal examination, and colonoscopy. Seventy-six of the 120 also had double-contrast barium enemas. Fifty-nine (48 percent) of the 120 patients who completed the work-up had neoplastic lesions; 14 of them were malignant (11.4 percent) and 44 were benign (36.6 percent). Three of the malignant lesions were carcinoma in situ, three had stalk invasion but no residual tumor was found at surgery (Dukes' type A), four were Dukes' type B, and three were Dukes' type C. The classification of one was uncertain. There was stalk invasion, but no further surgery was performed. Thus, of the 14 malignant lesions detected during the first screening, 13 had known classifications and 10 of these 13 were therapeutically favorable (carcinoma in situ, Dukes' type A or B). Ten of the 14 malignant lesions were beyond the range of the rigid proctosigmoidoscope. Our results to date indicate the following: (1) There was excellent patient follow-up and compliance for diagnostic procedures and therapeutic intervention (88.8 percent) when the results of the fecal occult blood test were positive. (2) Neoplasia had a high predictive value (48 percent). (3) A favorable staging for malignant lesions was detected by this method (71 percent). (4) An unacceptably high percentage of malignant lesions were beyond the range of the rigid sigmoidoscope (71 percent). (5) Work-up for patients with a positive fecal occult blood test result should include full colonoscopic examination.