Transfusion Association with Adverse Clinical Outcomes in Cardiac Surgery: An Analysis of a Large Patient Cohort.

Abstract

Abstract 3158Poster Board III-95 BackgroundPrevious studies have demonstrated significant variation in red cell transfusion in cardiac surgery. Allogeneic red cell transfusion has been reported as associated with adverse post-operative outcomes in this setting, whilst there is conflicting evidence on potential adverse effects of platelets. We investigated the variation in transfusion practice (including red cell, platelets and plasma transfusion) across six large cardiac surgery centres and whether peri-operative transfusion was independently associated with clinical outcomes. MethodsData were prospectively collected on 9363 cardiac surgery patients at six major Australian hospitals from January 2005 to December 2008 through the Australasian Society of Cardiac and Thoracic Surgeons (ASCTS) Cardiac Surgery Database. The independent association of transfusion (including red cells, platelets and plasma) with clinical outcomes was determined by stepwise logistic regression analysis. Patient factors (including co-morbidities and medication use such as anticoagulants and anti-platelet agents) and surgical factors (including type of procedure, previous cardiothoracic intervention, urgency of surgery, cardiopulmonary bypass time, cross-clamp time, return to theatre and other peri-operative complications) were included in the analysis. ResultsProcedure types were: coronary artery bypass graft (CABG) surgery in 60%, valve surgery in 14%, CABG and valve in 10% and other procedure in 16%. There was significant variation in transfusion of all blood components between the six hospitals, which was not accounted for by patient or surgery related factors. The adjusted odds ratio (OR) for the risk of receiving a transfusion of red cells varied between the hospitals from 0.22 to 1.98, for platelets from 0.39 to 3.3 and for plasma from 0.27 to 2.0. Transfusion of red cells, platelets and plasma were each associated with all outcome measures. In multiple logistic regression analysis, red cell transfusion was independently associated with multi-system failure (OR, 2.97; 95% confidence interval [CI], 1.41 - 6.28; p = 0.004), peri-operative myocardial infarction (OR, 2.63; 95% CI, 1.45 - 4.77; p = 0.001), stroke (OR, 1.76; 95% CI, 1.13 - 2.72; p = 0.012), prolonged ventilation (OR, 2.6; 95% CI, 2.19 - 3.06; p < 0.001), pulmonary embolism (OR, 4.29; 95% CI, 1.4 - 13.16; p = 0.011), pneumonia (OR, 2.12; 95% CI, 1.8 - 2.72, p < 0.001) and septicaemia (OR, 2.29; 95% CI, 1.31 - 4.0; p = 0.004). Platelet transfusion was independently associated with multi-system failure (OR, 2.17; 95% CI, 1.3 - 3.63; p = 0.003) and pneumonia (OR, 1.25; 95% CI, 1.02 - 1.52; p = 0.03). Plasma transfusion was independently associated with in-hospital (OR, 2.46; 95% CI, 1.82 - 3.32; p < 0.001) and 30 day mortality (OR, 2.2; 95% CI, 1.61 - 2.99; p < 0.001), multi-system failure (OR, 1.97; 95% CI, 1.18-3.31; p = 0.010), prolonged ventilation (OR, 1.81; 95% CI, 1.55 - 2.13; p < 0.001) and septicaemia (OR, 1.83; 95% CI, 1.17 -2.85; p = 0.008). ConclusionThere was significant variation in transfusion practice for red cells, platelets and plasma across these six major hospitals that was not accounted for by patient or surgery related factors. Peri-operative transfusions of red cells, platelets and plasma, after adjusting for established clinical risk factors, were each independently associated with an increased risk of adverse events. These findings from a large cohort of patients are in keeping with previous reports of adverse events associated with red cell transfusion in cardiac surgery and support the need for further research into the effects of transfusion on patient outcomes. An improved understanding of factors contributing to the significant variation in transfusion practice is also required as this may have important implications for patient outcomes. DisclosuresNo relevant conflicts of interest to declare.