Jan
2012

Transfusion of blood during cardiac surgery is associated with higher long-term mortality in low-risk patients.

Jakobsen CJ, Ryhammer PK, Tang M, Andreasen JJ, Mortensen PE
Eur J Cardiothorac Surg 2012; Jan 12 [Epub ahead of print].
NATA Rating :
Review by : O. Habler
NATA Review

In their retrospective analysis, Jakobsen et al. evaluated the effect of red blood cell (RBC) transfusion on long-term survival (>4500 days) after cardiac surgery in a historical, consecutively collected multicenter patient cohort (4 centers; 20001 patients ≥15 yrs; coronary artery bypass grafting alone or together with aortic or mitral valve replacement; data from the Western Denmark Heart Registry, Danish Civil Registration System and Danish Transfusion Registry; observation period 1999-2010). Patients with multiple entries, previous cardiac surgery or death within 30 days after surgery were not included. No specific transfusion algorithm was used until 2004.

The percentage of patients transfused significantly increased with increasing age and increasing residual EuroSore (EuroScore 0-2: 26.6%, EuroScore >11: 85.4%). Of the transfused patients, 55% received 1-2 units, 24.6% 3-4 units, 10.1% 5-6 units and 10.3% >6 units of RBC. The volume transfused increased significantly with an increasing EuroScore but not with increasing age. Long-term (>4500 days) survival rate was significantly lower for transfused patients if the EuroScore was 0-4 (0.637 vs. 0.745) and 5-9 (0.373 vs. 0.436). In patients with a EuroScore >9, survival was independent of transfusion. In low-risk patients (Euro Score 0-4), RBC transfusion was identified as an independent risk factor for long-term mortality leading to a difference in mortality of 10.8%.

Age of RBC and the consecutive storage lesion, as well as the immunomodulatory effects of allogeneic blood are proposed as potential explanations for the findings. Limitations of the analysis are (1) its retrospective character, (2) the lack of a consistent RBC transfusion protocol, (3) the exclusion of blood products other than RBC (e.g. platelet concentrates, fresh frozen plasma), (4) the lack of differentiation between leukodepleted and non-leukodepleted blood, (5) no information about the age of blood transfused and (5) no information about perioperative hemoglobin concentrations triggering transfusion. Moreover, the number of patients who were transfused during non-cardiac surgical procedures within the observation period and the impact of this “additional” transfusion on survival remains unknown.

– Oliver Habler