Basic variables at different positivity thresholds of a quantitative immunochemical test for faecal occult blood

Abstract

OBJECTIVES: Screening by faecal occult blood testing (FOBT) is effective in decreasing mortality and incidence of colorectal cancer (CRC). Immunochemical tests have proved to be more cost effective than guaiac FOBTs. The latex agglutination test (LAT) has the advantage of being a fully automated, quantitative test. The aim of this study is to interpret the overall experience with LAT according to different positivity thresholds. SETTING: A population based screening programme is currently running involving subjects aged 50–70, invited every 2 years to have an FOBT. LAT is the standard screening test and has a positivity threshold for further diagnostic tests of 100 ng haemoglobin/ml of sample solution. METHODS: Positivity rates, detection rates for CRC high risk adenomas, and positive predictive values for CRC, high risk adenomas, and low risk adenomas were calculated for several positivity thresholds. RESULTS: 19 132 attendances at screening were recorded (11 774 at first screening, 7358 at subsequent screenings). Progressively increasing the positivity threshold from 100 to 200 ng/ml showed (a) a decrease in positivity rate; (b) a decrease in detection rates for CRC or high risk adenomas; (c) an increase in positive predictive values for cancer; (d) an increase in positive predictive value for high risk adenomas. CONCLUSIONS: Increasing the positivity threshold of the LAT reduces recall rate and improves positive predictive value for cancer or high risk adenomas but substantially decreases the detection rate of CRC and high risk adenomas. For this reason increasing the positivity cut off for LATs is not advisable. On the other hand decreasing the positivity threshold would increase recall rate and sensitivity of screening. Careful evaluation of sensitivity of the quantitative results of the LAT for interval cancers is needed to definitively assess the optimal positivity threshold for LATs in population based screening programmes.