A meta-analysis of randomized controlled trials comparing methylene blue-directed biopsies with random biopsies in the surveillance of Barrett's esophagus Abstract This study was designed to examine the shortfalls of methylene blue-directed biopsies (MBDB) and conventional random biopsies (RB) in the surveillance of Barrett's esophagus. A Cochrane and Medline search was performed, and fi ve randomized controlled trials (RCTs) comprising 213 patients were selected for the meta-analysis. Random biopsies failed to spot specialized intestinal metaplasia (SIM) in a signifi cantly higher number of patients than MBDB (risk ratio, 0.724; confi dence interval, 0.552-0.950). The rate of underdiagnosis of dysplasia was comparable between the two techniques (risk ratio, 0.766; confi dence interval, 0.273-2.147). More biopsies were obtained using the RB technique, and a signifi cantly higher number of biopsies was reported as normal in comparison to MBDB (P = 0.014). The use of endoscopic staining was associated with prolongation of the procedure time (P = 0.009). It was concluded that SIM is overlooked more often in conventional biopsy techniques, but the incidence of undetected dysplasia is comparable using both methods. The routine use of methylene blue in the surveillance of Barrett's esophagus seems to be unnecessary.