Observer variance in the reporting of faecal occult blood results using the Okokit II method.

Abstract

Tests to detect blood in faeces in amounts or forms not observable on inspection (occult blood) are important in the diagnosis and treatment of lesions of the gastrointestinal tract. Faecal occult blood screening of patients has been shown to increase the detection of colorectal cancer! indeed the American Cancer Society recommends faecal occult blood tests every year for people over the age of fifty. 2 The Okokit IFM (Hughes and Hughes Ltd., Romford, Essex, UK) method used in this depart­ ment is based on the peroxidase-like activity of haemoglobin. Emergence ofa green colour within 2 min of reagent addition to the prepared faecal sample, in accordance with kit instructions, is regarded as a positive result, all other results being recorded as negative. A study was conducted to investigate the possibility of inter-observer variance in the reporting of results, without being concerned with the examination ofthe specificity and sensiti­ vity of the method. Four observers negative for red-green colour blindness were chosen from the laboratory staff. Over a period offive weeks 54samples received by the laboratory for routine faecal occult blood testing were analysed in batches of four to eight using the Okokit II kit. The test procedure was carried out on each sample by one person only, and all four observers were present. Observers simultaneously but independently recorded their results within the confines of the kit instructions. Results were classified as 'Agreement' when all four observers agreed on a result, 'Partial Dis­ agreement' when three agreed and 'Disagree­ ment' when only two agreed. It can be seen from Table I that 'Agreement' was recorded on only 57% of the samples assayed. Variation in the total results recorded as positive by individual observers was high, rang­ ing from 35% to 61%.