A Clinical Trial of Rapid Tests for Occult Blood in Faeces

TH E DE TEe T ION of occult blood in the faeces is an important diagnostic test in disease of the alimentary tract. Tn particular, the identification of traces of blood may suggest the presence of a malignant lesion and may hasten the performance of full radiological and endoscopic examinations or of a surgical operation. It is highly desirable, therefore, that a reliable test should be available for routine use. Needham and Simpson (1952) suggested that an ideal test for faecal occult blood should fulfil the following requirements: (1) It should be sufficiently sensitive to give a positive result with the small amounts of blood likely to be encountered in bleeding from a local ulcerative lesion in the gastrointestinal tract. (2) It should not give false positive results in patients free from gastrointestinal disease but who are eating meat or receiving oral iron therapy. (3) It should be sufficiently simple and hygienic for it to be carried out rapidly in an out-patient clinic on smears of faeces obtained by digital examination of the rectum. Most of the tests in clinical use are based on the benzidine reagent. This substance, when dissolved in acid solution and mixed with an oxidising agent such as hydrogen peroxide, reacts with even very small quantities of haemoglobin to form an intensely blue pigment. The standard benzidine test, as usually carried out on an emulsion of faeces, is extremely sensitive and is without significance unless the patient has been on a meatfree diet for three days before the specimen is obtained. Therefore it is unsuitable for use on out-patients. Adlers' (1904) and Gregersen's (1916) tests were developed to overcome these drawbacks. Tn the former, a solution of benzidine in glacial acetic acid is mixed with a diluted faecal emulsion. Tn the latter, a powder which contains benzidine and barium peroxide is dissolved in glacial acetic acid and brought in contact with a smear offaeces on a filter paper. Ham (1950) described a further test in which the reagent is benzidine dihydrochloride dissolved in glacial acetic acid and absolute alcohol. This reagent is mixed with a smear of faeces and hydrogen peroxide added. A careful investigation of Gregersen's test by Needham and Simpson indicated that in their hands it fulfilled the requirements that have been outlined, but a disturbing number of false positives in our own clinical practice and the need for a reliable test for use in surgical emergencies suggested that a reinvestigation of the rapid test might be valuable.