Importance: Patients of physicians with higher adenoma detection rates (ADRs) during colonoscopy have lower colorectal cancer (CRC) risk after screening colonoscopy (ie, postcolonoscopy CRC). Among physicians with an ADR above the recommended threshold, it is unknown whether improving ADR is associated with a lower incidence of CRC in their patients.
Objective: To determine the association of improved ADR in physicians with a range of ADR values at baseline with CRC incidence among their patients.
Design, setting, and participants: A total of 789 physicians in the Polish Colonoscopy Screening Program were studied between 2000 and 2017, with final follow-up on December 31, 2022. Joinpoint regression analyses were used to identify trends between changes in ADR and postcolonoscopy CRC incidence. Rates of CRC after colonoscopy were compared between physicians whose ADR improved and those without improvement. ADR improvement was defined as either an improvement by at least 1 ADR sextile category or remaining in the highest category.
Exposure: Physician ADR.
Main outcomes and measures: Association of improved ADR with postcolonoscopy CRC incidence.
Results: Of 485 615 patients (mean [SD] age, 57 [5.41] years; 60% female), 1873 CRC diagnoses and 474 CRC-related deaths occurred during a median follow-up of 10.2 years. Among individual physicians at baseline, median (IQR) ADR was 21.8% (15.9%-28.2%) and maximum ADR was 63.0%. Joinpoint regression showed a change in CRC incidence trends at an ADR level of 26%, corresponding to a CRC incidence of 27.1 per 100 000 person-years. Patients of physicians whose ADR was less than 26% at baseline and improved during follow-up had a postcolonoscopy CRC incidence of 31.8 (95% CI, 29.5-34.3) per 100 000 person-years, compared with 40.7 (95% CI, 37.8-43.8) per 100 000 person-years for patients of physicians with an ADR of less than 26% at baseline who did not improve during follow-up (difference, 8.9/100 000 person-years [95% CI, 5.06-12.74]; P < .001). Patients of physicians whose ADR was above 26% at baseline and improved during follow-up had a postcolonoscopy CRC incidence of 23.4 (95% CI, 18.4-29.8) per 100 000 person-years, compared with 22.5 (95% CI, 18.3-27.6) for patients of physicians whose ADR was above 26% at baseline and did not improve during follow-up (difference, 0.9/100 000 person-years [95% CI, -6.46 to 8.26]; P = .80).
Conclusions and relevance: In this observational study, improved ADR over time was statistically significantly associated with lower CRC risk in patients who underwent colonoscopy compared with absence of ADR improvement, but only among patients whose physician had a baseline ADR of less than 26%.