Detection of advanced colorectal neoplasia and relative colonoscopy workloads using quantitative faecal immunochemical tests: an observational study exploring the effects of simultaneous adjustment of both sample number and test positivity threshold

Abstract

ObjectiveWhen screening for colorectal cancer (CRC) using quantitative faecal immunochemical tests (FIT), test parameters requiring consideration are the faecal haemoglobin concentration (f-Hb) positivity cut-off and the number of stools sampled. This observational study explored variation in f-Hb between samples and the relationship between sensitivity for advanced neoplasia (AN, cancer or advanced adenoma) and colonoscopy workload across a range of independently-adjusted parameter combinations.DesignQuantitative FIT data (OC-Sensor) were accessed from individuals undergoing personalised colonoscopic screening with an offer of 2-sample FIT in the intervening years. We estimated variation in f-Hb between samples in 12 710 completing 2-sample FIT, plus test positivity rates (colonoscopy workload) and sensitivity for AN according to parameter combinations in 4037 instances where FIT was done in the year preceding colonoscopy.ResultsThere was large within-subject variability between samples, with the ratio for the second to the first sample f-Hb ranging up to 18-fold for all cases, and up to 56-fold for AN cases. Sensitivity for AN was greatest at lower f-Hb cut-offs and/or using 2-sample FIT. Colonoscopy workload varied according to how parameters were combined. Using different cut-offs for 1-sample FIT and 2-sample FIT to return similar sensitivity, workload was less with 2-sample FIT when the sensitivity goal exceeded 35%.ConclusionVariation in f-Hb between samples is such that both parameters are crucial determinants of sensitivity and workload; independent adjustment of each should be considered. The 2-sample FIT approach is best for detecting advanced adenomas as well as CRC provided that the colonoscopy workload is feasible.