Abstract
Background and aims: Adenoma detection rate (ADR) is a key quality indicator for colonoscopy; however, it is cumbersome to obtain. We investigated if detection rates (DRs) for adenomas, serrated polyps (SPs) and clinically relevant SP (crSPDR) can be accurately estimated by individualized DR ratios (DRRs) in a multicenter primary colonoscopy screening cohort of average-risk individuals.
Methods: DRRs were calculated by dividing DRs for a certain polyp entity by polyp detection rate (PDR) for each endoscopist individually on the basis of his/her first 50 (DRR50) and 100 (DRR100) consecutive colonoscopies. DRs were estimated for each endoscopist by multiplying his/her DRR for a certain polyp entity with his/her PDR of subsequent colonoscopies in groups of 50 (DRR50) and 100 (DRR100) consecutive colonoscopies. Estimated and actual DRs were compared.
Results: Estimated DRs showed a strong correlation with actual DRs for adenomas (r = 0.86 and 0.87; each p < .001), SPs (r = 0.85 and 0.91; each p < .001) and crSPs (r = 0.82 and 0.86; each p < .001) using DRRs derived from first 50 and 100 consecutive colonoscopies. Corresponding root mean square error (RMSE) between individual estimated and actual DRs using DRR50 and DRR100 was 5.3(±4.6)% and 4.5(±4.8)% for adenomas, 5.2(±4.1)% and 3.9(±2.8)% for SP, 3.1(±3.1)% and 2.8(±2.5)% for crSP, respectively. RMSE was not significantly different between DRR50 and DRR100 for ADR (p = .445), SPDR (p = .178) and crSP (p = .544).
Conclusions: DR for all relevant polyp entities can be accurately estimated by using individual DRRs. This approach may enable endoscopists to easily track their performance measures in daily routine.
Disclosure statement
No potential conflict of interest was reported by the author(s).