Screening for colorectal neoplasms with a new immunological human faecal haemoglobin and albumin test

In Germany, screening for colorectal cancer shows low efficiency, which is partly due to demographic changes with a rising mean age of the population, a low participation rate and an unsatisfactory sensitivity of guaiac tests for detecting faecal occult-blood. Therefore, a pilot screening study with a new immunological faecal haemoglobin and albumin test was performed in Ostringen, Germany to assess its compliance, performance characteristics and cost-effectiveness. Two thousand, seven hundred and eighty-five persons (1,498 women and 1,287 men) collected 1 ml samples from two different sites of one stool. The upper limit of normal was 10 micrograms/g stool for haemoglobin and 100 micrograms/g stool for albumin. The compliance was 82%; 224 persons (8%) had a positive test result. Of these, 184 underwent full colonoscopy. We detected 14 colorectal cancers, 10 of which were Dukes' stage A carcinomas removed by endoscopic polypectomy, 34 large adenomas and 43 small adenomas. The detection rate for colorectal neoplasms was above the rate described for other immunological haemoglobin tests and for Haemoccult tests. The specificity of the test--defined with false-positive results if a normal colon mucosa and no other reasons for upper or lower gastrointestinal bleeding were found--was 99.5%. The cost-effectiveness was assessed by comparing the diagnostic costs with the savings resulting from prevention of colorectal carcinomas by endoscopic polypectomy of malignant polyps (Dukes' stage A). The savings in our screening study exceeded the diagnostic costs by approximately 2.3 times. The combined immunological faecal haemoglobin and albumin test should substitute the Haemoccult test in colorectal cancer screening because of its higher sensitivity and specificity combined with cost-effectiveness and good patient compliance.