Cutoff value determines the performance of a semi-quantitative immunochemical faecal occult blood test in a colorectal cancer screening programme

BACKGROUND: The cutoff of semi-quantitative immunochemical faecal occult blood tests (iFOBTs) influences colonoscopy referrals and detection rates. We studied the performance of an iFOBT (OC-Sensor) in colorectal cancer (CRC) screening at different cutoffs. METHODS: Dutch screening participants, 50-75 years of age, with average CRC risk and an iFOBT value >= 50 ng ml(-1) were offered colonoscopy. The detection rate was the percentage of participants with CRC or advanced adenomas (>= 10 mm, >= 20% villous, high-grade dysplasia). The number needed to scope (NNTScope) was the number of colonoscopies to be carried out to find one person with CRC or advanced adenomas. RESULTS: iFOBT values >= 50 ng ml(-1) were detected in 526 of 6157 participants (8.5%) and 428 (81%) underwent colonoscopy. The detection rate for advanced lesions (28 CRC and 161 with advanced adenomas) was 3.1% (95% confidence interval: 2.6-3.5%) and the NNTScope was 2.3. At 75 ng ml(-1), the detection rate was 2.7%, the NNTScope was 2.0 and the CRC miss rate compared with 50 ng ml(-1) was <5% (N = 1). At 100 ng ml(-1), the detection rate was 2.4% and the NNTScope was <2. Compared with 50 ng ml(-1), up to 200 ng ml(-1) CRC miss rates remained at 16% (N = 4). CONCLUSIONS: Cutoffs below the standard 100 ng ml(-1) resulted in not only higher detection rates of advanced lesions but also more colonoscopies. With sufficient capacity, 75 ng ml(-1) might be advised; if not, up to 200 ng ml(-1) CRC miss rates are acceptable compared with the decrease in performed colonoscopies. British Journal of Cancer (2009) 101, 1274-1281. doi:10.1038/sj.bjc.6605326 www.bjcancer.com Published online 15 September 2009 (C) 2009 Cancer Research UK