S1322 “Early Is Not Always Better”: Impact of Endoscopy Timing in Cirrhotic Patients With Acute Variceal Bleeding—An Updated Systematic Review and Meta-Analysis

Introduction: The optimal timing of endoscopy in cirrhotic patients with acute variceal bleeding (AVB) remains controversial, with contradictory results from recent studies. This updated meta-analysis incorporates newer, higher-quality evidence since the last similar analysis, to clarify the impact of endoscopic timing on patient outcomes. Methods: Following PRISMA guidelines (PROSPERO: CRD420251063584), PubMed, Scopus, Cochrane CENTRAL, and Embase were searched by 2 reviewers. We included observational studies and clinical trials comparing early versus delayed endoscopy based on timing from admission to endoscopy (≤6 h, ≤12 h, ≤48 h). We defined very early endoscopy as ≤6 h and early endoscopy as ≤12 h from admission. Outcomes included all-cause mortality, rebleeding, and adverse events (AEs, like perforation, infection, aspiration). Results were pooled using RevMan 5.4.1, random-effects model (P < 0.05 for significance). A priori subgroup analyses (mortality: 6 week/in-hospital and rebleeding: early/delayed/undefined), sensitivity analyses, and Egger's tests were conducted to address bias. Results: Fifteen studies (11 retrospective, 3 prospective, 1 randomized controlled trial; N = 9,046 patients, >3× prior meta-analyses) were included. Endoscopy within 48 h significantly reduced all-cause mortality vs >48 h (RR 0.53, 95% CI 0.30-0.94, I2 = 26%). In contrast, very early endoscopy (≤6 h) was associated with increased mortality (RR 1.57, 95% CI 1.05-2.34, I2 = 0%). Early endoscopy (≤12 h) showed no mortality benefit. In-hospital mortality was also lower with endoscopy ≤48 h (RR 0.56, 95% CI 0.34-0.92, I2 = 15%). Rebleeding rates did not differ overall or by subgroup. However, early endoscopy (≤12 h) significantly increased overall AEs (RR 2.98, 95% CI 1.83-4.84, P < 0.0001, I2 = 5%). Conclusion: Endoscopy within 48 h reduces overall and in-hospital mortality in cirrhotic AVB. However, very early endoscopy (≤6 h) may increase mortality—possibly due to inadequate stabilization or selection bias. Early timing also increases AEs, highlighting the need for individualized decision-making. Unlike prior analyses that pooled heterogeneous thresholds, our study clearly defines very-early and early timing and presents low-heterogeneity evidence. These findings challenge the “early is always better” paradigm and support a more nuanced approach to endoscopy timing in AVB.