Bayesian network meta‐analysis: Efficacy of air insufflation, CO 2 2 insufflation, water exchange, and water immersion in colonoscopy

摘要

Background and Aim Colonoscopy is an excellent screening tool for colorectal cancer. There are four colonoscopy techniques: air insufflation, CO 2 insufflation, water exchange, and water immersion. Some studies reported that the latter three methods are better than the criterion standard (air insufflation), whereas some studies did not. In order to evaluate the efficacy of the four colonoscopy techniques, a network meta‐analysis was carried out. Methods We searched randomized controlled trials (RCT) published up to September 2017 from PubMed, Embase, Cochrane library, and Web of Science. Studies referencing the comparison between at least two of air insufflation, CO 2 insufflation, water exchange, and water immersion were selected. Primary outcomes included pain score during insertion, polyp detection rate, and adenoma detection rate, and secondary outcomes included cecal intubation time and cecal intubation rate. Mean differences or odds ratios and their corresponding 95% credible intervals were pooled with Bayesian modeling. Results Forty RCT with 13?734 patients were included in this network meta‐analysis. Our analysis showed that air insufflation had the highest pain score (surface under the cumulative ranking curve [SUCRA]: 98.8%) and the lowest detection rate of adenoma (SUCRA: 21.3%) and polyp (SUCRA: 16.8%). Water exchange had the lowest pain score (SUCRA: 1.1%) and highest detection rate of adenoma (SUCRA: 96.0%) and polyp (SUCRA: 98.9%), although it led to the longest cecal intubation time (SUCRA: 86.9%). Conclusions Air insufflation might be the most unsatisfactory colonoscopy. Meanwhile, water exchange might be the most efficient colonoscopy.