Volume and accreditation, but not specialty, affect quality standards in colonoscopy

Abstract

Abstract Background The Global Rating Scale, defined by the Joint Advisory Group for Gastrointestinal Endoscopy, requires monitoring of endoscopic performance indicators. There are known variations in colonoscopic performance, and investigation of factors causing this is needed. This study aimed to analyse the impact of endoscopist specialty and procedural volume on the quality of colonoscopy. Methods Data collected prospectively from a UK hospital endoscopy service between June 2007 and January 2010 were analysed. The main endpoint was the adenoma detection rate (ADR). Secondary endpoints were polyp detection rate (PDR), reported caecal intubation rate (CIR) and reported complications. Multivariable binary regression models were built to adjust for confounding patient-level and endoscopist-level variation. Results A total of 10 026 colonoscopies were included, with an overall ADR of 19·2 per cent, a CIR of 90·2 per cent and a perforation rate of 0·06 per cent. In univariable analyses, surgeons had a higher ADR and higher PDR, but lower CIR, compared with physicians. Surgeons had a significantly different case mix in terms of age, sex and indication for colonoscopy. After adjusting for this case mix in multivariable analysis, specialty was no longer a significant predictor of ADR; however, surgeons retained their higher PDR and physicians their higher CIR. Endoscopists accredited for screening and those performing more than 100 colonoscopies per year had a higher ADR. Conclusion Adjusting for case mix, physicians and surgeons performed equally well in terms of ADR. Accreditation and a higher annual number of colonoscopies were more important factors in achieving quality standards.