Continuous Quality Improvement of Screening Colonoscopy: Data from a Large Colorectal Cancer Screening Program

Abstract

Background and Aim: Suboptimal colonoscopy quality may decrease the effectiveness of colorectal cancer (CRC) screening programs. Although many professional societies recommend continuous quality improvement (CQI) programs, little is known about their impact on colonoscopy quality indicators such as adenoma detection rate (ADR) or cecal intubation rate (CIR). The aim of this study was to determine whether CQI program improves the ADR and CIR over time. Methods: CQI is an integral part of national colonoscopy-based CRC screening program in Poland (Regula J et al. N Engl J Med 2006). The central database of the program, containing all colonoscopy and pathology reports, was reviewed once a year in order to assess the completeness and consistency of data and to calculate individual endoscopists' quality indicators. Following the review feedback information on the performance was sent to the endoscopists. We analyzed the data on ADR and CIR for 41 endoscopists from 25 centers who performed at least 100 yearly screening colonoscopies between 2004 and 2007. The increase in mean ADR was compared between groups of endoscopists with acceptable (≥10%) and poor (<10%) baseline rates. Results: Forty-one participating endoscopists performed a mean 1,113 colonoscopies (range 465-3,642) over the four year study period (a total of 45,644 colonoscopies for the whole group). The mean baseline (year 2004) ADR was 14.1% (range 5.0-26.9%) whereas the mean baseline CIR was 94.5% (range 84.2-100.0%). The mean ADR adjusted to endoscopist specialization and colonoscopy experience increased in subsequent years (2005, 2006 and 2007) by 1.5, 3.6 and 4.6 percent points, respectively, as compared to the baseline value (p=0.038, p<0.001, p<0.001, respectively). The mean ADR of endoscopists with acceptable baseline ADR reached the recommended threshold values over the study period (20.5% ± 6.5%) but it remained suboptimal in the group with poor baseline ADR (11.1% ± 3.8%). The recommended ADR of 20% was achieved by 14 (46.7%) endoscopists with acceptable baseline ADR while none of colonoscopists with poor baseline ADR reached that goal. Mean CIR and the percentage of endoscopists with a CIR ≥95% was similar at the baseline (94.5% and 53.7%, respectively) and in the subsequent years (95.3%, 95.4%, 95.4%, p≥0.06 and 63.4%, 61.0%, 65.9%, p≥ 0.07, respectively). Conclusions: CQI integrated in a large CRC screening program improves the mean ADR achieved by participating endoscopists by about 5% over 4 years period. However, endoscopists with a poor baseline ADR do not improve their ADR sufficiently using only feedback information and should undergo additional intensive training.