Blood Transfusion in Primary Total Hip and Knee Arthroplasty. Incidence, Risk Factors, and Thirty-Day Complication Rates

Adam Hart,Jad Abou Khalil,Alberto Carli,Olga Huk,David Zukor,John Antoniou

Background: The aim of this study was to analyze NSQIP (National Surgical Quality Improvement Program) data to better understand the incidence, risk factors, and thirty-day complication rates associated with transfusions in primary total hip and knee arthroplasty.

Methods: We identified 9362 total hip and 13,662 total knee arthroplasty procedures from the database and separated those in which any red blood-cell transfusion was performed within seventy-two hours after surgery from those with no transfusion. Patient demographics, comorbidities, preoperative laboratory values, intraoperative variables, and postoperative complications were compared between patients who received a transfusion and those who did not. Multivariate logistic regression was used to identify independent risk factors for receiving a transfusion as well as for associated postoperative complications (thirty-day incidences of infection, venous thromboembolism, and mortality).

Results: The transfusion rate after total hip arthroplasty was 22.2%. Significant risk factors for receiving a transfusion were age (OR [odds ratio] per ten years = 10.1), preoperative anemia (OR = 3.6), female sex (OR = 2.0), BMI (body mass index) of <30 kg/m2 (OR = 1.4), and ASA (American Society of Anesthesiologists) class of >2 (OR = 1.3). Multivariate logistic regression analysis indicated that adjusted odds of infection, venous thromboembolism, and mortality did not differ significantly between patients who received a transfusion and those who did not. The transfusion rate after total knee arthroplasty was 18.3%. Risk factors for receiving a transfusion were age (OR per ten years = 10.2), preoperative anemia (OR = 3.8), BMI of <30 kg/m2 (OR = 1.4), female sex (OR = 1.3), and ASA class of >2 (OR = 1.3). Multivariate logistic regression indicated that a transfusion was significantly associated with mortality (OR = 2.7) but not with infection or venous thromboembolism.

Conclusions: We did not find a strong association between perioperative red blood-cell transfusion and thirty-day incidences of infection, venous thromboembolism, or mortality; however, the odds of mortality were higher in patients who received a transfusion during total knee arthroplasty.