Prophylactic antibiotics cannot prevent endoscopic retrograde cholangiopancreatography-induced cholangitis: a meta-analysis.

OBJECTIVES To determine the prophylactic effect of antibiotics on post-endoscopic retrograde cholangiopancreatography (ERCP) cholangitis or sepsis reduction in randomized controlled trials. METHODS Databases including MEDLINE, EMBASE, Cochrane Library, and Science Citation Index updated to June 2007 were searched. Main outcome measure was post-ERCP cholangitis or sepsis. RESULTS Seven trials were identified, and a total of 1389 patients were included; post-ERCP cholangitis occurred in 5.8% of controls (41/705) versus 3.4% of treated patients (23/684), without statistical significance (relative risk [RR], 0.58; 95% confidence interval [CI], 0.22-1.55). Subsequent sensitivity analysis on trials mainly targeted at patients with suspicious biliary obstruction showed that the incidences of post-ERCP cholangitis were 2.8% (12/425) and 5.4% (24/441) in the antibiotics and control groups, respectively, and this sensitivity analysis did not support antibiotics' preventive effect (RR, 0.33; 95% CI, 0.03-3.32). Another sensitivity analysis exclusively including trials with intravenous route of antibiotics administration also failed to confirm the prophylactic effect of antibiotics (RR, 0.53; 95% CI, 0.18-1.60). CONCLUSIONS Antibiotics cannot significantly prevent ERCP-induced cholangitis in unselected patients and should not be routinely recommended. Yet, more trials including patients with predicted incomplete biliary drainage are still required to prove the effectiveness of prophylactic antibiotics in this setting.