A multivariable model for perioperative transfusion risk in elective primary THA

BACKGROUND: Perioperative blood transfusion remains a relevant concern in elective total hip arthroplasty (THA), as it is associated with adverse outcomes and increased healthcare utilization. Reliable, procedure-specific risk stratification tools based on routinely available clinical data are needed to support individualized perioperative blood management. METHODS: This retrospective cohort study included consecutive patients undergoing elective primary THA between 2016 and 2023 at a single certified arthroplasty center. Demographic data, comorbidity burden (ASA classification), perioperative laboratory values, operative characteristics, blood loss parameters, and transfusion data were extracted from a prospectively maintained registry. The primary outcome was perioperative allogeneic red blood cell transfusion. Univariable and multivariable logistic regression analyses were performed to identify independent predictors of transfusion. Model performance was assessed using receiver operating characteristic analysis and calibration testing. RESULTS: A total of 648 patients were included, of whom 104 (16.0%) required perioperative transfusion. Transfused patients demonstrated lower preoperative hemoglobin levels, greater perioperative hemoglobin decline, and higher total and hidden blood loss. On multivariable analysis, lower preoperative hemoglobin, longer operative time, greater intraoperative blood loss, and lower body mass index were independently associated with transfusion, while male sex was inversely associated. Higher comorbidity burden (ASA ≥ III) showed a borderline association after adjustment. The final model demonstrated excellent discrimination with an area under the curve of 0.878 and good calibration. CONCLUSION: External validation of the present model is required before routine clinical application. Perioperative transfusion following elective primary THA can be predicted with good discriminative performance using routinely available clinical variables, supporting perioperative risk assessment and individualized blood management, although its applicability for preoperative patient counseling is limited. LEVEL OF EVIDENCE: Level III (retrospective cohort study).