The proximal serrated polyp detection rate is an easy-to-measure proxy for the detection rate of clinically relevant serrated polyps

Background and aims: The adenoma detection rate (ADR) is the most important surrogate quality parameter for colorectal cancer (CRC) prevention. However, serrated polyps also are precursors of CRC. Large, prospective studies comparing the detection rate of serrated polyps among endoscopists in an era of awareness about the malignant potential of serrated polyps have not yet been performed. We aimed to compare the proximal serrated polyp (PSP) detection rate and the clinically relevant serrated polyp (RSP) detection rate among endoscopists and to analyze the association between these parameters and the ADR.

Methods: Colonoscopy data were retrieved in one expert center between January 2011 and July 2014 by using a structured reporting system, enabling prospective and automatic quality assessment. Endoscopists who performed at least 50 colonoscopies within the timeframe were included for analysis. Multivariate logistic regression was used to compare the ADR, PSP detection rate, and RSP detection rate among endoscopists. The association among these parameters was calculated by using the Pearson r correlation coefficient. All lesions were assessed by an expert pathologist.

Results: In total, 16 endoscopists and 2088 colonoscopies were included for analysis. The PSP detection rate ranged from 2.9% to 18.6% (mean 10.4%) among endoscopists. Corrected for confounders, the odds ratio to detect ≥1 PSP, compared with endoscopists with the highest detection rate, ranged from 0.79 (95% confidence interval [CI], 0.41-1.52) to 0.12 (95% CI, 0.03-0.55). The PSP detection rate was highly correlated with the RSP detection rate (ρ 0.94; P < .001), ranging from 4.3% to 20.9% (mean 13.9%). The PSP detection rate moderately correlated with the ADR (0.55; P = .03), which ranged from 23.2% to 49.2% (mean 35.2%).

Conclusions: The PSP detection rate is widely variable among endoscopists, strongly correlated with the RSP detection rate, and moderately correlated with the ADR. These results suggest a high miss rate of RSPs among endoscopists with low rates of PSP detection. Future research should determine the association between endoscopists' PSP detection rates and the risk of interval cancer.