JAMA Study Questions Routine Prophylactic Antibiotics use for Upper-GI Bleeding in Cirrhosis
New research in JAMA Internal Medicine has challenged the current guideline recommending 5–7 days of preventive antibiotics for cirrhotic patients with upper-GI bleeding, citing limited supporting evidence and growing calls for reassessment. The study was conducted by Connor P. and colleagues.
International guidelines have recommended this length of treatment to lower mortality, rebleeding, and infections, but the current systematic review and bayesian meta-analysis indicates that the mortality benefit for making the recommendation is not supported by high-quality, up-to-date evidence. The authors warn that infection reporting strategies used in earlier trials varied and were frequently high-risk for bias, which further undermines the strength of existing guidelines.
The review examined randomized clinical trials (RCTs) of longer antibiotic regimens (5–7 days) versus shorter regimens (2–3 days) or no antibiotics in cirrhotic patients with upper GI bleeding. Embase, MEDLINE, and CENTRAL databases were searched through September 25, 2024. RCTs with mortality or early rebleeding results only were included; pediatric literature, observational studies, gray publications, and nonsystemic antibiotic trials were excluded.
A total of 14 RCTs with 1322 participants were considered. The range of mean age was 41.5 to 62 years, and 74.2% (981 patients) were male. Most—90.9% (1202 patients)—had variceal bleeding as the source. Two trials compared directly 5–7 days versus 2–3 days of prophylaxis, and the remaining 12 compared any prophylaxis (1–10 days) with no prophylaxis.
Results
Mortality: Shorter lengths (including none) had a 97.3% likelihood of noninferiority for all-cause death versus longer regimens (Risk Difference [RD] 0.9%, 95% credible interval [CrI] −2.6 to 4.9).
Early Rebleeding: The likelihood of noninferiority in shorter lengths was 73.8% (RD 2.9%, 95% CrI −4.2 to 10.0).
Infections: Shorter lengths were associated with 15.2% more study-defined bacterial infections (95% CrI 5.0 to 25.9). Infection definitions, however, differed substantially between studies.
Subgroup Analysis: In post-2004 published studies, when endoscopic and supportive care was better, probabilities for noninferiority regarding mortality, rebleeding, and infection outcomes were even more favorable.
This meta-analysis finds that the mortality benefit underlying the recommendation for 5–7 days of antibiotic prophylaxis against cirrhosis complicated by upper GI bleeding is not supported by high-quality evidence. The authors emphasize that high-quality, contemporary RCTs are imperative to define the optimal duration of antibiotics. Until then, clinicians should recognize that current guidelines are based on outdated and low-to-moderate quality evidence.
Reference:
Prosty C, Noutsios D, Dubé L, et al. Prophylactic Antibiotics for Upper Gastrointestinal Bleeding in Patients With Cirrhosis: A Systematic Review and Bayesian Meta-Analysis. JAMA Intern Med. Published online August 11, 2025. doi:10.1001/jamainternmed.2025.3832
