Background Upper gastrointestinal bleeding (UGIB) is a significant cause of cirrhosis decompensation; however, there is still controversy on risk factors for poor outcomes. This study aims to explore the impact of additional variables, such as liver disease etiology and intensive care unit (ICU) accessibility, on mortality, rebleeding, and infection rates. Methods Cox regression analysis was employed to identify predictors associated with mortality and rebleeding, while Poisson regression analysis was utilized to assess associations with infections. Results A total of 228 patients were included, with the majority classified as Child‐Pugh B and C. Among them, 96 patients (42.1%) had alcohol‐associated liver disease (ALD), of which 45 (46.9%) were in alcohol abstinence. One hundred and ninety‐five patients (85.5%) survived, while 33 (14.5%) died. Intubation was required for 31 patients (13.6%), and 28 were transferred to the ICU. Antibiotic prophylaxis was administered to 219 patients (96%), and 55 (24.1%) developed infections. Both orotracheal intubation and transfusions were associated with high mortality, whereas ALD was linked to a lower risk of death ( p < 0.001, 0.029, and 0.005, respectively). Intubation was also correlated with an increased risk of rebleeding, while transfer to ICU reduced this complication ( p = 0.012 and 0.045, respectively). The Child‐Pugh score and length of hospitalization were associated with the occurrence of infections ( p = 0.037 and < 0.001, respectively). Conclusions Intubation, transfusions, liver disease etiology, and ICU accessibility are critical predictors following UGIB in cirrhosis. Additionally, the Child‐Pugh score and duration of hospitalization are directly proportional to the risk of infections in this context. Trial Registration: ClinicalTrials.gov identifier: NCT04662918
Marcondes et al. (Thu,) studied this question.