Differences in endoscopy practices between providers with highest and lowest post endoscopy upper gastrointestinal cancer rates using national endoscopy database

Introduction Post endoscopy upper gastrointestinal cancer (PEUGIC) is an important key performance indicator of endoscopy quality. PEUGIC rates vary more than threefold between endoscopy providers in England. We examined differences in endoscopy practices between providers with the highest and lowest PEUGIC rates to identify factors which can potentially be addressed to improve endoscopy quality in England. Methods PEUGIC rates were calculated for all endoscopy providers in England using the National Cancer Registration and Analysis Service and Hospital Episode Statistics databases. Endoscopy providers were categorised into quartiles based on their PEUGIC rates and those in the highest and lowest quartiles were included in this study. Data for all diagnostic UGI endoscopies performed in these providers between January 2019 and February 2020 were extracted from the National Endoscopy Database (NED). Multivariable regression analysis was performed to identify endoscopy practices associated with low PEUGIC rate providers, after adjusting for patient characteristics and indications. This project was funded by the National Institute for Health and Care Research (Grant Reference Number NIHR 201571). Results 330,851 endoscopies performed at 60 endoscopy providers were included. UGI cancers were diagnosed on 0.9% (n=3,071) of endoscopies. PEUGIC rates ranged from 5.1% to 7.6% in the low PEUGIC rate providers and 9.3% to 13.0% among high PEUGIC rate providers. Low PEUGIC rate providers were associated with: training lists (Odds Ratio 2.31 (95% CI 2.26–2.36)); more high risk endoscopy indications (alarm features or patients ≥55 years of age) (1.08 (1.06–1.1)); intravenous sedation use (1.24 (1.22–1.26)); endoscopists with an average UGI endoscopy annual volume of 101–200 (1.10 (1.07–1.12)), compared with <100; and endoscopy lists with 7 to 9 points per list (1.23 (1.21–1.25)). Compared with Midazolam alone, use of combined sedation (Midazolam and Fentanyl) was also more common among low PEUGIC rate providers (1.17 (1.15–1.19)). Endoscopies at low PEUGIC rate providers were less likely to be performed by endoscopists not on specialist, nursing or trainee registers (0.88 (0.85–0.90) and biopsies (other than Helicobacter testing) were less common (0.83 (0.81–0.84). Conclusions In the National Endoscopy Database, formal training, more high-risk endoscopy indications, endoscopist annual endoscopy volume, intravenous sedation, less biopsies and lower intensity endoscopy lists were associated with low PEUGIC rate providers. These findings can guide endoscopy quality improvement efforts and hopefully reduce PEUGIC rate in future.