Risks Associated With Platelet Transfusion After Cardiac Surgery: Intuitive or Counterintuitive?

Abstract

We read with great interest the work by Fletcher et al.1Fletcher CM, Hinton JV, Xing Z, et al. Platelet transfusion after cardiac surgery [e-pub ahead of print]. J Cardiothorac Vasc Anesth. doi: 10.1053/j.jvca.2022.12.009. Accessed December 30, 2022.Google Scholar investigating the association between platelet transfusion and hospital mortality after cardiac surgery. Among the 1,621 patients who received postoperative platelet transfusion, 1,064 patients were 1:1 matched with propensity scoring. The authors reported that postoperative platelet transfusion had no association with in-hospital mortality, reduced odds of suspected infection, and shorter lengths of stay in the intensive care unit and the hospital. The current work supported a recent systematic review and meta-analysis on platelet transfusion in cardiac surgery, which concluded that platelet transfusion was not associated with operative mortality, reoperation for bleeding, stroke, and infection.2Yanagawa B Ribeiro R Lee J et al.Platelet transfusion in cardiac surgery: A systematic review and meta-analysis.Ann Thorac Surg. 2021; 111: 607-614Abstract Full Text Full Text PDF PubMed Scopus (14) Google Scholar The authors are commended for their efforts, but they were unable to match the propensity scores for 35% of the qualifying patients. This sample loss suggests important clinical differences between the 2 cohorts. The degree of overlap in propensity-score distribution between the 2 populations can affect the ability to match to reduce bias. The sample loss contributes to unintended bias, potentially skewing the model and its outcomes (eg, overfitting).3Hade EM Lu B. Bias associated with using the estimated propensity score as a regression covariate.Stat Med. 2014; 33: 74-87Crossref PubMed Scopus (53) Google Scholar These concerns are not necessarily related to the function of propensity-score matching analysis itself, but rather may be due to the suboptimal selection of variables and application of propensity-score matching. The lack of data for intraoperative administration of blood products, including platelets, prevented the authors from excluding a possible confounder that may affect the relationship between postoperative platelet transfusion and outcome. The authors included the amount of chest tube drainage with their propensity score matching, and compared the 2 cohorts with similar bleeding patterns in the intensive care unit, but they failed to consider the association of intraoperative bleeding and platelet transfusion with a possible change in outcome. The source data were collected over a 20-year period. The rate of postoperative platelet transfusion (13.5%) seems acceptable, but there have been major changes in antiplatelet regimens, perioperative transfusion practices, and hemostasis monitoring during this time interval. For example, glycoprotein IIb/IIIa inhibitors have largely been replaced by newer P2Y12 antagonists in recent years.4Gellatly RM Connell C Tan C et al.Trends of use and outcomes associated with glycoprotein-IIb/IIIa inhibitors in patients with acute coronary syndromes undergoing percutaneous coronary intervention.Ann Pharmacother. 2020; 54: 414-422Crossref Scopus (5) Google Scholar Available platelet products have also changed (eg, additives, strategies for pathogen reduction).5Fontaine MJ Lasola JJM Henderson RA. Perioperative platelet transfusion: Not all platelet products are created equal.Curr Anesthesiol Rep. 2022; 12: 320-328Crossref Scopus (0) Google Scholar Recent blood management guidelines have also substantially influenced transfusion practice patterns and reduced overall perioperative allogeneic transfusions in cardiac surgery.6Irving AH Harris A Petrie D et al.Impact of patient blood management guidelines on blood transfusions and patient outcomes during cardiac surgery.J Thorac Cardiovasc Surg. 2020; 160 (437-45.e20)Abstract Full Text Full Text PDF PubMed Scopus (12) Google Scholar Thus, it is important to consider when surgery was performed as a variable in the analysis. None. Platelet Transfusion After Cardiac SurgeryJournal of Cardiothoracic and Vascular AnesthesiaVol. 37Issue 4PreviewTo investigate the independent association of platelet transfusion with hospital mortality and key relevant clinical outcomes in cardiac surgery. Full-Text PDF