Abstract
INTRODUCTION: Gastrointestinal (GI) complications after cardiac surgery, although uncommon (0.4-3%), are associated with high mortality rates (13-63%). This study aimed to describes outcomes of GI complications and propose a risk score predicting their occurrence. METHODS: We conducted a retrospective case-control study including 8,544 patients undergoing cardiac surgery between 2005 and 2019. GI complications-defined as bleeding, ischemia/perforation, obstruction, or pancreatitis-were used to select cases for 1:2 propensity score matching with controls. Multivariable logistic regression was performed to determine independent predictors of a composite endpoint (postoperative pulmonary, renal, and cardiac complications). A GI complication risk score was then developed. RESULTS: After matching, 162 patients were analyzed (54 GI group, 108 controls). Smokers and patients with significant coronary artery disease were more frequent in the GI group (p = 0.0049 and p = 0.0013). GI complications included ischemia (50.6%), hemorrhage (31.5%), pancreatitis (3.7%), and obstruction (14.8%), with 24.1% occurring as part of multiorgan failure. Overall mortality in the GI group was 38.8% compared with 0.9% in controls (p < 0.0001). Independent predictors of the composite endpoint were GI complications, NYHA class >2, and preoperative acute myocardial ischemia. The risk score, incorporating smoking, age, chronic kidney disease, and mitral valve replacement, showed good discrimination (area under the curve 0.735, 95% CI 0.653-0.816) and calibration (Hosmer-Lemeshow p = 0.934). CONCLUSIONS: Mortality remains high among patients who develop GI complications after cardiac surgery, regardless of treatment modality. This risk score represents a potentially valuable tool for identifying patients at increased risk and underscores the importance of close monitoring and timely intervention in this population.