Perioperative complications after hemiarthroplasty and total shoulder arthroplasty are equivalent

Abstract

Total shoulder arthroplasty (TSA) results in superior clinical outcomes to hemiarthroplasty (HA); however, TSA is a more technical and invasive procedure. This study retrospectively compares perioperative complications after HA and TSA using the National Surgical Quality Improvement Program (NSQIP) database. The NSQIP user file was queried for HA and TSA cases from the years 2005 through 2010. Major complications were defined as life-threatening or debilitating. All complications occurred within 30 days of the initial procedure. We performed multivariate analysis to compare complication rates between the two procedures, controlling for patient comorbidities. The database returned 1,718 patients (HA in 30.4% [n = 523] and TSA in 69.6% [n = 1,195]). The major complication rates in the HA group (5.2%, n = 29) and TSA group (5.1%, n = 61) were similar (P = .706). Rates of blood transfusions for postoperative bleeding in patients undergoing HA (2.3%, n = 12) and TSA (2.9%, n = 35) were indistinguishable (P = .458). Venous thromboembolism was a rare complication, occurring in 0.4% of patients in each group (2 HA patients and 5 TSA patients, P > .999). On multivariate analysis, the operative procedure was not associated with major complications (P = .349); however, emergency case, pulmonary comorbidity, anemia with a hematocrit level lower than 36%, and wound class of III or IV increased the risk of a major complication (P < .05 for all). Multivariate analysis of patients undergoing TSA or HA in the NSQIP database suggests that patient factors-not the procedure being performed-are significant predictors of major complications. Controlling for patient comorbidities, we found no increased risk of perioperative major complications in patients undergoing TSA compared with HA.