Background and objective: Optimization of the preoperative hemoglobin (Hb) level is an effective way to reduce allogeneic transfusion in total knee arthroplasty (TKA) though the procedure is expensive, requires close monitoring and is often inconvenient for patients with reduced mobility. Our aim was to investigate the value of preoperative Hb levels to predict transfusion and thereby tailoring Hb optimization to patient characteristics.
Materials and methods: All consecutive patients who undergone primary TKA in our center over 2 years, and received tranexamic acid intraoperatively, were reviewed. The adjusted association between preoperative Hb levels and transfusion was assessed by multivariate logistic regression, and the estimated probability of transfusion for individual patients was derived from the logistic model.
Results: Out of the 784 patients who meet the inclusion criteria, risk of transfusion was associated with poorer performance status, as measured by the America Association of Anestesiology (ASA) score III/IV (OR: 3·3, P < 0·001) and lower preoperative Hb level (OR 3·8 for each g/dl below 13 g/dl; P < 0·001). According to the Hb level, the estimated probability of transfusion was 0·03 (range: 0·03-0·64) for ASA I/II patients and 0·10 (range: 0·10-0·84) for ASA III/IV.
Conclusion: Not all the patients undergoing TKA who receive tranexamic acid need the same preoperative Hb optimization target. Two easily available factors, such as the ASA score and the Hb level, can help individualize the Hb optimization target.
Keywords: arthroplasty; erythrocyte transfusion; hemoglobin; knee replacement; tranexamic acid.
© 2014 International Society of Blood Transfusion.