Narrow-Band Imaging for Detection of Neoplasia at Colonoscopy: A Meta-analysis of Data From Individual Patients in Randomized Controlled Trials

Abstract

Background & Aims: Adenoma detection rate (ADR) is an important quality assurance measure for colonoscopy. Some studies suggest that narrow-band imaging (NBI) may be more effective at detecting adenomas than white-light endoscopy (WLE) when bowel preparation is optimal. We conducted a meta-analysis of data from individual patients in randomized controlled trials that compared the efficacy of NBI to WLE in detection of adenomas.

Methods: We searched MEDLINE, EMBASE, and Cochrane Library databases through April 2017 for randomized controlled trials that assessed detection of colon polyps by high-definition WLE vs NBI and from which data on individual patients were available. The primary outcome measure was ADR adjusted for bowel preparation quality. Multilevel regression models were used with patients nested within trials, and trial included as a random effect.

Results: We collected data from 11 trials, comprising 4491 patients and 6636 polyps detected. Adenomas were detected in 952 of 2251 (42.3%) participants examined by WLE vs 1011 of 2239 (45.2%) participants examined by NBI (unadjusted odds ratio [OR] for detection of adenoma by WLE vs NBI, 1.14; 95% CI, 1.01–1.29; P = .04). NBI outperformed WLE only when bowel preparation was best: adequate preparation OR, 1.07 (95% CI, 0.92–1.24; P = .38) vs best preparation OR, 1.30 (95% CI, 1.04–1.62; P = .02). Second-generation bright NBI had a better ADR than WLE (second-generation NBI OR, 1.28; 95% CI, 1.05–1.56; P = .02), whereas first-generation NBI did not. NBI detected more non-adenomatous polyps than WLE (OR, 1.24; 95% CI, 1.06–1.44; P = .008) and flat polyps than WLE (OR, 1.24; 95% CI, 1.02–1.51; P = .03).

Conclusions: In a meta-analysis of data from individual patients in randomized controlled trials, we found NBI to have a higher ADR than WLE, and that this effect is greater when bowel preparation is optimal.