Adenoma detection rate in fecal immunochemical test positive colonoscopies: a population-based observational study

Abstract

Background and aims: Neoplasia-related indicators, such as adenoma detection rate (ADR), are a priority in the quality improvement process for colonoscopy. Our aim was to assess and compare different detection and characterization indicators in fecal occult blood test (FOBT)-positive colonoscopies, to analyze the factors influencing their variance, and to propose benchmarks. Methods: Retrospective analysis of prospectively collected data from all colonoscopies performed between 2007 and 2019 after a positive guaiac-based FOBT or a fecal immunochemical test (FIT) in the population-based colorectal cancer (CRC) screening program conducted in Alsace, part of the French national program. Detection indicators included ADR, NewADR (including proximal serrated lesions), mean number of adenomas per colonoscopy, and proximal serrated lesion detection rate. Characterization indicators included non-neoplastic polyp detection rate. Results: Overall, 13.455 FIT-positive colonoscopies were performed by 116 endoscopists. The overall ADR was 57.6% (95%CI 56.8-58.5). For each 10 microg/g increase in fecal hemoglobin concentration, a 2% increase in ADR was observed. Endoscopists whose ADR was ≥55% were high detectors for all neoplasia, including proximal serrated lesions and number of adenomas. The non-neoplastic polyp detection rate was 39.5% in highest detectors (ADR >70%), significantly higher than in lower detectors (21.4%) (p<0.001). There was a strong correlation between detection and characterization indicators, e.g. between proximal serrated lesion and non-neoplastic polyp detection rates (Pearson = 0.73; p<0.01). Conclusions: A single indicator, NewADR, including proximal serrated lesions, is enough to assess the neoplasia yield of colonoscopy provided the target standard is raised between 55% and 70% in FIT-positive colonoscopies (65-80% in men, 45-60% in women).

Colonoscopy may be weak link in organised colorectal cancer screening programme with faecal immunochemical test

“…We adopted a cut-off of 20 annual FIT-positive colonoscopies as we found that the CIRs and ProxSLDRs of endoscopists performing <20 colonoscopies were significantly lower than of those performing ≥20. 24 Here, more than half of the low-volume, assessable endoscopists were low-performing. We thus considered that the patients were not being given the best possible chances when their endoscopists were non-assessable and performed <20 annual FIT-positive colonoscopies.…”

“…Like others, we observed quite a good correlation between ADR and ProxSLDR. 24 However, out of ten low-ADR endoscopists, only two had low ProxSLDRs, and of five low-ProxSLDR endoscopists, only two had low ADRs, giving an argument for measuring the two indicators separately.…”