Both endoscopic band ligation (EBL) and non-selective beta-blockers are used for primary prophylaxis of variceal bleeding, but the performance of EBL may vary across trials. We aimed to evaluate how study design and patient characteristics influence EBL-related outcomes in this context.

We performed a comprehensive search of MEDLINE, Embase, Cochrane Library, and major conference proceedings from 1995 to April 2025 to identify randomized controlled trials comparing EBL with placebo or alternative interventions. Sample-weighted univariate and multivariate linear regression models were applied to examine factors associated with bleeding events, variceal eradication, and mortality. Weighting was based on the number of patients undergoing ligation in each study.

Forty RCTs were included, comprising 2055 patients treated with EBL. The pooled incidence of procedure-associated bleeding was 3% (95% CI 2–4%), while post-ligation bleeding occurred in 11% (95% CI 9–15%). In multivariate models, overtube use was independently associated with increased procedure-related bleeding (β = 0.710; 95% CI 5.389–26.124; p = 0.007). Alcohol-related liver disease showed a borderline association with higher post-ligation bleeding risk (β = 0.427; 95% CI –0.019–0.324; p = 0.078). Variceal eradication was achieved in 94% of cases (95% CI 88–97%), and a higher prevalence of red wale marks correlated with improved eradication rates (β = 0.537; 95% CI 0.002–0.335; p = 0.048). The overall mortality estimate was 18.3% (95% CI 14.0–22.2%), with longer follow-up duration significantly linked to higher mortality (β = 0.651; 95% CI 0.366–1.072; p < 0.001).

Outcomes of EBL for primary prophylaxis show substantial heterogeneity across trials. Procedural factors, patient profiles, and follow-up duration significantly influence bleeding, eradication, and mortality outcomes. Recognizing these determinants is essential for designing and interpreting future EBL studies.