It is essential to identify the factors in clinical practice that influence the technical performance of colonoscopy as a basis for quality improvement programs.
To assess the factors linked to two key indicators of colonoscopy performance, i.e., cecal intubation and polyp diagnosis.
Consecutives colonoscopies performed over a 2-wk period in 278 unselected practice sites throughout Italy were prospectively evaluated. A multivariate model was developed to identify determinants of the performance indicators of colonoscopy.
In total, 12,835 patients (mean age 60.5 yr, standard deviation [SD] 15.1, 53% men) were studied. Sedation and/or analgesia was administered in 55.3% of procedures: 28.8% of patients received intravenous (IV) benzodiazepines, 15.4% received benzodiazepines in combination with narcotics, 3.1% received propofol, and 7.5% received other sedation regimens. Overall, cecal intubation was achieved in 80.7% of procedures, and the polyp detection rate was 27.3%. Multivariate analysis showed that the strongest predictors of cecal intubation were the quality of bowel preparation (inadequate vs excellent: odds ratio [OR] 0.013, 95% confidence interval [CI] 0.009–0.018; fair vs excellent: OR 0.246, 95% CI 0.209–0.290; and good vs excellent: OR 0.586, 95% CI 0.514–0.667) and the use of sedation (IV benzodiazepines vs no sedation: OR 1.460, 95% CI 1.282–1.663; IV benzodiazepines and narcotics vs no sedation: OR 2.128, 95% CI 1.776–2.565; and propofol vs no sedation: OR 2.355, 95% CI 1.590–3.488). The colonoscopy setting (workload and organizational complexity of the center) and the endoscopist colonoscopy volume were other factors independently correlated with completion of the procedure. Detection of polyps partially depended on the quality of bowel cleansing (inadequate vs excellent: OR 0.511, 95% CI 0.404–0.647) and use of sedation (OR 1.172, 95% CI 1.074–1.286).
In usual clinical practice, the use of sedation/analgesia, the colon-cleansing quality, the endoscopist experience, and some features related to the colonscopy setting decisively influence the quality of colonoscopy. These factors indicate the targets of future corrective measures to boost the quality of this examination.
1Department of Gastroenterology, Valduce Hospital, Como, Italy; and 2Institute of Biostatistics, OPIS, Milan, Italy
Reprint requests and correspondence: Franco Radaelli, M.D., Department of Gastroenterology, Ospedale Valduce, Via Dante 11, 22100 Como, Italy.
This paper has been presented in part at the ASGE meeting, DDW, May 20–24, 2006, Los Angeles, CA.
Received August 3, 2007; accepted November 27, 2007.