B25-20 Predictors of Short-term Mortality in ICU Patients With Gastrointestinal Bleeding: A Systematic Review and Meta-analysis

Abstract Rationale Gastrointestinal bleeds (GIB) are a significant source of fatal outcomes in the intensive care unit (ICU) with estimated mortality rates up to 25%. Identifying prognostic factors including presentation, laboratory values, risk scores, and need for therapeutic interventions refine risk stratification and assist triages. Despite advances in in-hospital mortality risk prediction (e.g., Glasgow-Blatchford score), to date, no meta-analyses have collectively assessed and summarized such outcomes in critically ill patients admitted to the ICU. This meta-analysis aims to synthesise prognostic factors for short-term mortality among patients with GIBs, requiring ICU admission. Methods We performed a PRISMA-compliant systematic review, and searched MEDLINE, EMBASE, and PubMed. Studies were included if they: (1) were observational or randomized; (2) involved patients admitted to the ICU with GIB; (3) reported short-term mortality (in-hospital, 30-day). Risk of bias was assessed using the Quality in Prognostic Studies tool, and certainty of evidence was rated with the Grading of Recommendations, Assessment, Development, and Evaluation (GRADE) framework. Results We included 33 full texts encompassing 16 251 patients (37.1% female). The pooled mortality rate was 13.3% (n = 2153). Pre-existing malignancy (OR 2.14, 95%CI 1.40-3.28, p 0.001, GRADE: low) was associated with increased mortality, while pre-existing hypertension (OR 0.71, 95%CI 0.51-1.00, p = 0.047, GRADE: low) was associated with reduced mortality. Presentation factors associated with increased mortality included: hepatic encephalopathy (OR 3.38, 95%CI 2.03-5.64, p 0.001, GRADE: low), sepsis (OR 4.44, 95%CI 1.98-9.96, p 0.001, GRADE: low), and heart failure (OR 1.43, 95%CI 1.10-1.85, p 0.001, GRADE: low). The APACHE II (OR 1.17, 95%CI 1.12-1.23, p 0.001, GRADE: low), SOFA (OR 1.75, 95%CI 1.43-2.14, p 0.001 GRADE: low) and MELD (OR 1.19, 95%CI 1.13-1.24, p 0.001, GRADE: low) risk scores were associated with short-term mortality. Conclusion Short-term mortality in GIB in the ICU appears driven by multi-organ dysfunction over bleeding parameters, as evidenced by indices consistently incorporating signs of end-organ damage. Hypertension’s protective effect is likely confounded by age, chronic medications, and greater hemodynamic resilience. Prognosis in ICU-admitted GIB reflects systemic vulnerability, underscoring the need for ICU-specific risk models incorporating organ dysfunction and host reserve. Our review’s limitations include retrospective analyses, limited adjustment for confounders, and heterogeneity across studies. This abstract is funded by: None