No mortality benefit with longer antibiotic prophylaxis durations in cirrhosis, upper GI bleeds

In patients with cirrhosis and upper gastrointestinal (GI) bleeds, mortality outcome does not differ between shorter and longer antibiotic prophylaxis durations, according to a systematic review and Bayesian meta-analysis.

Researchers searched multiple online databases for randomized clinical trials (RCTs) in which longer durations of antibiotic prophylaxis were compared with shorter durations (or 0 days) in patients with cirrhosis and upper GI bleeding. The primary outcome was all-cause mortality with a prespecified 5-percent noninferiority margin on the risk difference (RD) scale. Secondary outcomes included early rebleed and bacterial infections.

A total of 14 RCTs involving 1,322 participants (mean age range 41.5–62.0 years, 74.2 percent male), of whom 1,202 (90.9 percent) had a variceal source of bleeding, were included in the meta-analyses. Study quality, as assessed using the Cochrane Risk of Bias tool, was low to moderate. Bacterial infections were heterogeneously defined, and none of the trials reported adverse events. Two RCTs compared longer (5–7 days) with shorter (2–3 days) prophylaxis durations, while 12 RCTs compared any prophylaxis duration (1–10 days) with no prophylaxis.

Pooled data showed that compared with longer durations, shorter antibiotic prophylaxis durations (including no prophylaxis) had a 97.3-percent probability of noninferiority for all-cause mortality (RD, 0.9 percent, 95 percent credible interval [CrI], −2.6 to 4.9).

For early rebleeding, shorter durations had a 73.8-percent probability of noninferiority (RD, 2.9 percent, 95 percent CrI, −4.2 to 10.0) but were associated with more study-defined bacterial infections (RD, 15.2 percent, 95 percent CrI, 5.0–25.9).

The probabilities of noninferiority of shorter durations for all three outcomes were higher in RCTs published after 2004.