Limited effect of red blood cell transfusion on long-term mortality among anaemic cardiac surgery patients

OBJECTIVES: Our goal was to investigate long-term mortality associated with red blood cell (RBC) transfusion among patients with anaemia undergoing cardiac surgery when adjusting for known risk factors. METHODS: Adults with preoperative anaemia as defined by World Health Organization criteria undergoing open-heart surgery from 2000 through 2017 were included. Cox regression was performed for long-term mortality (30days-5years), comparing patients who received >= 1 unit of RBC with those who did not. Unadjusted and multivariable analyses adjusted for risk factors were performed. RESULTS: The study included 1859 patients, 1525 (82%) of whom received RBC transfusion. A total of 370 (19.9%) deaths were registered between 30days and 5years; 88 patients (23.8%) died between 30days and 1year. The unadjusted hazard ratio (HR) associated with RBC transfusion was 2.09 (1.49-2.93, P<0.001) from 30days to 5years postoperatively. The HR for RBC transfusion were 4.70 (1.72-12.81, P=0.002) and 1.77 (1.23-2.55, P=0.002) for 30days-1year and 1-5years, respectively. Adjusting for perioperative risk factors, which included postoperative complications, the HR decreased to 1.16 (0.80-1.68, P=0.43), 1.79 (0.63-5.12, P=0.28) and 1.11 (0.75-1.65, P=0.61) for observation time from 30days to 5years, 30days to 1year and 1 to 5years, respectively. Results were similar when postoperative complications were excluded from the adjustment variables. CONCLUSIONS: No statistically significant association between RBC transfusion and long-term mortality was found when we adjusted for known risk factors. This study suggests that the observed difference in mortality in this patient group is largely due to patient-related risk factors.