Variation in Proximal Sessile Serrated Lesion Detection Rates During Non-screening Colonoscopies.

AIMS

Proximal sessile serrated lesions (PSSL) are increasingly recognized as significant precursors of interval colon cancer. We aimed to assess the PSSL detection rates during non-screening colonoscopies and whether there are associations between PSSL detection rate and the established colonoscopy key performance indicators (KPI).

METHODS

We retrospectively collected data of all non-screening colonoscopies performed by independent endoscopists at a large teaching hospital between June and December 2019. Data regarding endoscopists' KPIs, including polyp detection rate (PDR), cecal intubation rate (CIR), and colonoscopy withdrawal time (CWT), were retrieved from the national endoscopy database. SSL resected proximal to the splenic flexure were identified by an expert pathologist. Associations between PSSL detection rate and the different KPIs were assessed using Spearman's test.

RESULTS

A total of 2956 colonoscopies performed by 33 endoscopists were included. The mean PSSL detection rate was 0.7% (SD 1.5), the mean PDR was 37.1% (SD 17), the mean CIR was 91.3% (SD 6), and the mean CWT was nine minutes (SD 2). There was marked variability in PSSL detection rates between endoscopists (range 0 - 6.5%). PSSL detection rate positively correlated with CWT (r=0.34, p=0.04) but not with the other KPIs.

CONCLUSION

The wide variability in PSSL detection between endoscopists is concerning of high miss rates and despite achieving the national benchmarks for the established KPIs, many endoscopists still had low PSSL detection rates. Therefore, PSSL detection rate should be considered as an independent KPI.