Abstract

Background: Interval colorectal cancers may be associated with a low serrated polyp detection rate (SDR) and advanced adenoma detection rate (AADR). We aimed to determine the SDR and AADR for endoscopists in a United States multicenter cohort.

Methods: We included average-risk screening colonoscopies from five medical centers in the United States. Endoscopists with data on at least 100 average-risk screening colonoscopies were included. We calculated median SDR and AADR for endoscopists with adequate adenoma detection rates (ADRs) > 25 %. We analyzed the relationship between ADR and SDR, and between ADR and AADR using nonparametric Spearman correlation coefficients, scatter plots, and linear regression.

Results: We included 3513 screening colonoscopies performed by 26 gastroenterologists. The mean age of patients was 56.8 years (SD 7.4) and 1585 (45 %) were male. All but one endoscopist had an ADR above 25 %. There was a significant positive but modest correlation between ADR and SDR (rho = 0.67,  < 0.01), and between ADR and AADR (rho = 0.56,  < 0.01). For endoscopists with an adequate ADR, median (interquartile range) ADR was 43 % (32.0 % - 48.6 %), median SDR was 8.4 % (7.3 % - 11.4 %), and median AADR was 9.3 % (6.4 % - 12.6 %).

Conclusion: A significant percentage of endoscopists have either a low SDR or low AADR despite an adequate ADR, justifying the need for separate SDR and AADR benchmarks. Based on our multicenter cohort, endoscopists with adequate ADRs had a median SDR and median AADR of about 8 % and 9 %, respectively.