Predictors of Intraoperative Blood Transfusion in Free Tissue Transfer

Background Free tissue transfer has become a safe and reliable procedure and is routinely used in a variety of settings. However, it is associated with lengthy operating times and a high potential for blood loss and consecutive red blood cell transfusions (RBCTs). Methods To assess the risk for RBCTs, we retrospectively identified 398 patients undergoing free tissue transfer between 2005 and 2014. Based on a multivariate model of risk factors and their respective odds ratio, a risk score was developed to predict the likelihood of the need for intraoperative RBCT. Results The median age at the time of operation was 51.3 ± 15 years, and 278 (70%) patients were male. The average body mass index was 25.9 ± 4 and the median ASA score was 2 (range: 1-4). Mean duration of surgery was 319.8 ± 108 minutes and mean duration of hospital stay was 45.8 ± 40 days. A total of 231 patients (58%) required perioperative RBCTs, all of which were allogenic. RBCTs were performed 0 to 48 hours preoperatively in 36 patients (11.3%), intraoperatively in 166 patients (41.7%), and 0 to 48 hours postoperatively in 125 patients (31.4%). The mean amount of overall RBCTs given was 2.5 ± 3.7 units and 1.1 ± 1.9 units for intraoperative transfusions. The following risk factors were statistically significant in the multivariate regression analysis and included in the risk score: age >60 years; a preoperative hemoglobin concentration of <11 g/dL; a preoperative platelet count of >400/nL; history of renal (RI) and cardial insufficiency (CI); defect localization on the proximal extremities, head and neck, or trunk; and the use of myocutaneous flaps. This score assessed the risk for RBCTs with a sensitivity of 77%, a specificity of 81%, and an AUC of the ROC curve of 0.86. Conclusion We were able to develop a risk score that allows for the assessment of RBCT likelihood. While most of the identified risk factors cannot be prevented or corrected, it still allows for improved patient counseling and can potentially reduce the number of ordered but not transfused RBCTs.