LAUSR

BACKGROUND Postoperative re-exploration for bleeding (RB) is a frequent complication following cardiac surgery. We aim to assess incidence, risk factors, and prognostic significance of RB in a large cohort of cardiac patients. MATERIALS AND METHODS We reviewed prospectively collected data for all patients who underwent cardiac surgery at our institution from 2007 to 2015. Logistic regression analysis was used to identify independent predictors of RB and specific outcomes. Propensity matching using a 1:1-ratio compared outcomes of patients who had RB with patients who did not. RESULTS During the study period, 7381 patients underwent cardiac operations. Of them, 189 (2.6%) underwent RB. RB was an independent predictor of in-hospital mortality (Odds Ratio (OR):2.62 Confidence Interval (CI):1.38-4.96; p = 0.003), major adverse events (OR:3.94, CI:2.79-5.62; p < 0.001), gastrointestinal events (OR:3.54 CI:1.73-7.24), renal failure (OR:2.44, CI:1.23-4.82), prolonged ventilation (OR:3.83, CI:2.60-5.62, p < 0.001), and sepsis (OR:2.50, CI:1.03-6.04, p = 0.043). Preoperative shock (OR:3.68, CI:1.66-8.13; p = 0.001), congestive heart failure (OR:1.70 CI:1.24-2.32; p = 0.001), and urgent and emergent status (OR:2.27, CI:1.65-3.12 and OR:3.57, CI:1.89-6.75; p < 0.001 for both) were predictors of RB operative mortality. Operative mortality, incidence of major adverse events, gastrointestinal events, and respiratory failure were all significantly higher in the propensity matched RB group (p = 0.050, p < 0.001, p = 0.046, and p < 0.001 respectively). CONCLUSIONS RB significantly increases in-hospital mortality and morbidity after cardiac surgery.