In a context of high detectors, is it necessary to monitor the detection rate of serrated lesions as a quality parameter in colonoscopy?

Aims Serrated lesion detection indicators have been proposed for inclusion in quality guidelines. However, there is no consensus on their usefulness or appropriate values as cut-off points. We aimed to assess whether, in the context of high-detecting endoscopists, the collection of quality indicators of serrated lesion detection allows differentiation between endoscopists with different performances.

Methods Retrospective analysis of a prospective database of CRC screening colonoscopies performed between January 2017 and January 2024 at a single center. Only endoscopists with≥100 examinations were included. Sessile serrated lesion detection rate (SSLDR) and combined serrated lesion detection rate (CSLDR=SSL+proximal hyperplastic polyp+hyperplastic polyp≥10mm+proximal serrated lesion+serrated lesion≥10mm+traditional serrated adenomas) were chosen as serrated lesion detection indicators. Endoscopists were divided into ‘Best’ (SSLDR and CSLDR>mean) and ‘Worst’ (SSLDR CSLDR<mean) groups. Both groups were compared for all other indicators [Lesion detection rate (LDR), Adenoma detection rate (ADR), ADR-plus, Advanced adenoma detection rate (AADR), cecal intubation rate (CIR), and withdrawal time≥6 min (WT6) or≥10 min (WT10)].

Results Nine endoscopists were evaluated (ADR≥55%). The mean SSLDR was 5% and the mean CSLDR was 15%. Five endoscopists were considered "Better" and 4 endoscopists "Worse". The comparison for each indicator between both groups of endoscopists was as follows: LDR 75.5% vs 66.7% (p=0.08); ADR 67.1% vs 59.2% (p=0.08); AADR 17.4% vs 18.1% (p=1); ADR-plus 4.8 vs 3.2 (p=0.01); CSLDR-plus 2.6 vs 2.1 (p=0.32); CIR 91.5% vs 90.6% (p=0.08); WT6 77.5% vs 64.8% (p=0.32); WT10 36.7% vs 11.7% (p=0.02)

Conclusions Endoscopists with above-average SSLDR and CSLDR values detect, on average, more adenomas (higher TDA-plus) and have longer withdrawal times (WT10). In a higher detector background, ADR is not enough to detect the best performers, and the use of indicators for sessile lesions detection may be necessary. An SSLDR≥5% and a CSLDR≥15% could be cut-off points.

Publication History

Article published online:
27 March 2025