Predicting red blood cell transfusion for primary hip and knee arthroplasty: Designing a hub and spoke high-volume arthroplasty model.

Background: Waiting lists for elective surgery are at record levels. Consolidating elective care into superhubs where high-volume surgery can be undertaken efficiently is an attractive solution. These pathways should be safe but also efficient and convenient for patients.We undertook this study to develop a model that would reliably predict healthy (American Society of Anaesthesiologists [ASA] class 1 and 2) patients who could be treated on a streamlined pathway and who would reliably not require red blood cell transfusion during their inpatient admission. Methods: We retrospectively identified all patients undergoing primary total hip arthroplasty (THA) or total knee arthroplasty (TKA) at our centre over a five-year period. We used binary logistic regression to develop a predictive model based on these variables. Results: We identified 13-preoperative candidate variables from our literature search and used these to construct our predictive model. The final validated model was highly effective in predicting those patients who would not need a red blood cell transfusion (area under curve = 0.945). Conclusion: Our model reliably predicts those healthy (ASA 1 or 2) patients undergoing primary THA or TKA surgery who will not require a red blood cell transfusion during their admission. These patients are suitable for this streamlined pathway without the need for unnecessary preoperative patient tests and travel. Level of evidence: III.