Predicting Postoperative Anemia and Blood Transfusion Following Total Knee Arthroplasty

Background While transfusion and clinically relevant anemia after elective primary total knee arthroplasty (TKA) are uncommon, there remains a question of who needs postoperative hemoglobin monitoring, especially in the setting of increasing incidence of outpatient TKA. The purpose of this study was to create predictive models for postoperative anemia and blood transfusion to guide clinical decision-making. Methods The records of consecutive TKA patients were reviewed from February 2016 to December 2020 at a single institution. Two multivariable logistic regression models, for postoperative anemia (hemoglobin < 10 g/dL) and allogeneic blood transfusion included 8 variables: age, sex, body mass index, preoperative hemoglobin level, tranexamic acid total dose, American Society of Anesthesiologists level, operative time, and drain use. Model performance was assessed using accuracy, area under the curve (AUC), sensitivity, and specificity. Results The records of 14,901 patients were included in this study. Patients had a mean (± standard deviation) age of 67.9 ± 9.2 years and mean body mass index of 31.3 ± 6.5 kg/m2. The postoperative anemia model had an accuracy of 88% (95% confidence interval [CI], 87%-89%) and AUC of 0.88 (95% CI, 0.87-0.89). The blood transfusion model had an accuracy of 97% (95% CI, 96%-97%) and AUC of 0.90 (95% CI, 0.87-0.93). Conclusion The postoperative anemia and blood transfusion model accurately predicted each outcome. Patients with less than a 5% probability of postoperative anemia may not benefit from a complete blood count at postoperative day 1. Application of these criteria may save the healthcare system hundreds of millions of dollars. Level of Evidence Prognostic Level IV. See Instructions for Authors for a complete description of levels of evidence.