Albumin Versus American Society of Anesthesiologists Score: Which Is More Predictive of Complications Following Total Joint Arthroplasty?

Abstract

Malnutrition is not uncommon in arthroplasty patients, and hypoalbuminemia has been shown to be predictive of postoperative complications. The authors sought to compare albumin concentration with a global assessment of physical health, the American Society of Anesthesiologists (ASA) score, to further discriminate the importance of albumin in predicting postoperative complications. A cohort of 128,412 patients undergoing total hip arthroplasty (THA) and total knee arthroplasty (TKA) were identified from 2005 to 2015 through use of the American College of Surgeons National Surgical Quality Improvement Program database. Patients were divided based on surgery performed, albumin concentration (with <3.5 g/dL defining hypoalbuminemia), and ASA score (≤2 vs >2). Postoperative complications were evaluated through multivariable regression analyses adjusted for age, sex, body mass index, and current smoking status. The study population included 48,751 THA and 79,661 TKA patients. On multivariable analyses, both hypoalbuminemia and ASA score were significant (P<.05) predictors of complications such as death, superficial infection, pneumonia, renal insufficiency, reintubation, transfusion, readmission, and reoperation. Furthermore, hypoalbuminemia more robustly predicted deep infection for THA patients, as well as superficial infection for TKA patients. American Society of Anesthesiologists score was otherwise predictive of more postoperative complications than hypoalbuminemia within the TKA group and provided similar predictability within the THA group. Albumin is associated with complications following THA and TKA, and if used in conjunction with ASA score, albumin contributes to enhanced risk stratification. The authors recommend that a preoperative albumin concentration be obtained and efforts made to modify this risk factor prior to elective arthroplasty. [Orthopedics. 2018; 41(6):354-362.].