ASSESSMENT OF ROUTINE TESTS FOR OCCULT BLOOD IN FAECES.

Abstract

Routine chemical methods for the detection of occult blood in faeces depend on the peroxidase-like activity of haemoglobin and its iron-containing degradation products. The clinical value of these tests was first demonstrated by Boas (1901), using a method based on the guaiacum test for blood described by van Deen (1861). Adler and Adler (1904) claimed that the colour reaction given by benzidine was better than that given by guaiacum. The benzidine test has been modified many times, and the method of Gregersen (1919) in which faeces are tested on a glass slide was widely accepted until Needham and Simpson (1952) recommended carrying out the test on filter paper. Benzidine is no longer obtainable in the U.K. because of its carcinogenicity and has been largely replaced by ortho tolidine. Hematest and occultest reagent tablets, which contain orthotolidine and a hydrogen-peroxide-generating sys tem, have been available commercially for several years. Hematest is the less sensitive and is recommended for use with faeces ; occultest is recommended for the detection of blood in urine but may be used with faeces if the patient has been suitably dieted for a few days before specimens are collected. The most recent addition to the range of commercially pro duced simplified tests for blood is the hemastix reagent strip, a cellulose strip, one end of which is impregnated with an organic peroxide and orthotolidine. A modified form of this strip is now marketed in the United States as a test for blood and for free haemoglobin in urine. The value of the various chemical tests for faecal occult blood has been widely debated, and the subject has been reviewed by Hughes (1952), Needham and Simpson (1952), Thornton and Illingworth (1955), and Steingold and Roberts (1961). Most authors think that the main defect is failure to differ entiate between the peroxidase activity of blood and peroxidases of dietary origin. Interpretation of results is also complicated by the possibility that faeces may contain inhibitors and perhaps accelerators of the peroxidase reaction. These difficulties and uncertainties are eliminated by the use of the 51Cr method for the determination of blood in faeces, first used in man by Owen, Cooper, Grindlay, and Bollmann (1954). This permits a quantitative assessment of faecal blood loss, but the technique is not at present suitable for routine use. Although there was a reasonable correlation between the results of chemical tests for occult blood and the amount of blood lost as measured by the isotopic method, gross discrepancies were found in indi vidual stools and were attributed to sampling errors by Banner man (1957), Ebaugh, Clemens, Rodnan, and Peterson (1958), and Cameron (1960). The present work was undertaken because of the paucity of information on this important aspect of testing for occult blood in faeces. The opportunity was also taken to reassess the value of some of the currently available chemical tests and to assess the value of hemastix as a test for blood in faeces.