Predicted Risk of Mortality, Transfusion, and Postoperative Outcomes in Isolated Primary Valve Operations

Background: Transfusion has been linked with increased postoperative morbidity and death after cardiac operations. The purpose of this study was to determine the associations among The Society of Thoracic Surgeons Predicted Risk of Mortality (PROM), transfusion, and postoperative outcomes in patients who underwent isolated primary valve operations.

Methods: A retrospective review of the local Society of Thoracic Surgeons database of 1,575 adults undergoing isolated primary valve operations from 2003 to 2013 at a United States academic center was performed. Patients were compared by their postoperative transfusion status (NONE vs TRANS) and by PROM. Taking into account procedure type and preoperative hemoglobin, three-step multiple linear or logistic regression analyses were performed to assess (1) the influence of PROM on postoperative outcomes, (2) influence of PROM on transfusion, and (3) influence of PROM and transfusion on postoperative outcomes.

Results: Of 1,575 patients studied, 1,245 (79%) received transfusions. The mean PROM was 1.2% (95% confidence interval [CI], 1.1 to 1.3) for patients in the NONE group, and was 2.7% (95% CI, 2.6 to 2.9) for the TRANS group. The correlation between PROM and total red blood cell units transfused was r = 0.31 (p < 0.0001). Patients with a PROM of 4% to 8% (odds ratio [OR], 2.10; 95% CI, 1.28 to 3.45) and exceeding 8% (OR 3.80, 95% CI, 1.35 to 10.68) were more likely to receive transfusions than the low-risk (<4%) PROM stratum. For each percentage increase in PROM, the odds of transfusion increased by 27% (95% CI, 16% to 39%), controlling for procedure type and preoperative hemoglobin. There were no 30-day deaths in the NONE group, and rates of stroke, renal failure, and mediastinitis were lower. Composite event rates increased with increasing PROM (OR, 1.39; 95% CI, 1.19 to 1.63), with TRANS patients consistently showing a higher risk of major adverse cardiac events than NONE patients (OR, 7.47; 95% CI, 2.08 to 26.80).

Conclusions: Increased PROM yielded higher risks of transfusion. Postoperative outcomes were worse in patients who received a transfusion. This study suggests that the association between transfusion and clinical outcomes may be partly explained by the higher PROM among patients who ultimately received transfusions.