Value of faecal alpha-1-antitrypsin, haemoglobin and a chemical occult blood test in the detection of gastrointestinal disease.

Specimens from hospital out-patients and in-patients sent for faecal occult blood tests were also analysed for faecal alpha-1-antitrypsin and faecal haemoglobin. 453 stool specimens from 222 patients were analysed. The clinicians were only aware of the faecal occult blood test results, and diagnoses were made using conventional clinical and investigative criteria. Gastrointestinal bleeding or putative sites of bleeding were diagnosed in 98 patients, whereas in 81 patients putative sites of bleeding were not found or other cause of anaemia diagnosed. In 41 patients there was insufficient information to reach a definitive diagnosis. Comparison of the 3 faecal tests using these grouping methods showed that faecal alpha-1-antitrypsin performed best, with an accuracy of 89%, specificity of 90% and sensitivity of 88%, all significantly better (P < 0.001) than the faecal occult blood test (68%, 60% and 73%, respectively). There was no significant difference between the performance of the faecal occult blood and faecal haemoglobin tests. Faecal alpha-1-antitrypsin measurement may be a useful investigation in situations where a faecal occult blood test would normally be requested.