Negative Haemoccult test in malignant and premalignant lesions of the colon. Validation of the Haemoccult test with total colonoscopy.

To date, no published report on a trial has provided sufficiently strong evidence of the accuracy of the Haemoccult Test assessed by false-negative error, and validated on the basis of a complete colonoscopy of all patients. Total colonoscopy (up to the cecum) was performed on 534 patients whose stools had been tested for occult blood. The results of the colonoscopy were classified as follows: --expected continual bleeding (target lesions: carcinoma, polyps greater than or equal to 10 mm) --intermittant bleeding (polyps less than 10 mm, colitis, etc.), and --no source of bleeding. Results were regarded as false-negative in patients who were found to have a target lesion (carcinoma, polyp greater than or equal to 10 mm) following a negative Haemoccult test. Apart from positive and negative prognostic values, sensitivity and specificity were taken into account as characteristics for the quality of the test. Following a positive Haemoccult test result 46 of the 534 patients were found to have polyps greater than or equal to 10 mm and 22 carcinomas; following negative Haemoccult test results 31 patients were found to have polyps measuring greater than or equal to 10 mm in diameter and 2 carcinomas. On the basis of the chosen classification a false-negative Haemoccult result was established in 8.6% and a false-positive result in 4.3%. The Haemoccult test showed a sensitivity of 67% and a specificity of 93% in patients with target lesions or patients without a source of bleeding.