Early Re-Exploration versus Conservative Management for Postoperative Bleeding in Stable Patients after Coronary Artery Bypass Grafting: A Propensity Matched Study

Background: Postoperative bleeding requiring re-exploration in cardiac surgery has been associated with complications impacting short-term outcomes and perioperative survival. Many aspects of decision-making for re-exploration still remain controversial, especially in hemodynamically stable patients with significant but not acutely cumulating chest drain output. We investigated the impact of re-exploratory surgery on short-term outcomes in a "borderline population" of CABG patients who experienced significant non-acute bleeding, but that were not in critically hemodynamic unstable conditions. Methods: A prospectively collected database of 8287 patients undergoing primary isolated elective CABG was retrospectively interrogated. A population of hemodynamically stable patients experiencing significant non-acute or rapidly cumulating bleeding (>1000 mL of blood loss in 12 h, <200 mL per hour in the first 5 h) with normal platelet and coagulation tests was identified (N = 1642). Patients belonging to this group were re-explored (N = 252) or treated conservatively (N = 1390) based on the decision of the consultant surgeon. Clinical outcomes according to the decision-making strategy were compared using a propensity score matching (PSM) approach. Results: After PSM, reoperated patients exhibited significantly higher overall blood product consumption (88.4% vs. 52.6% for red packed cells, p = 0.001). The reoperated group experienced higher rates of respiratory complications (odds ratio 5.8 [4.29-7.86] with p = 0.001 for prolonged ventilation), prolonged stay in intensive care unit (coefficient 1.66 [0.64-2.67] with p = 0.001) and overall length of stay in hospital (coefficient 2.16 [0.42-3.91] with p = 0.015) when compared to conservative management. Reoperated patients had significantly increased risk of multiorgan failure (odds ratio 4.59 [1.37-15.42] with p = 0.014) and a trend towards increased perioperative mortality (odds ratio 3.12 [1.08-8.99] with p = 0.035). Conclusions: Conservative management in hemodynamically stable patients experiencing significant but non-critical or emergency bleeding might be a safe and viable option and might be advantageous in terms of reduction of postoperative morbidities and hospital stay.