Shoulder arthroplasty for proximal humerus fracture is associated with increased postoperative complications and hospital burden

Abstract

ObjectivesTo compare 30-day outcomes including complications, readmissions, discharge location, operative duration, and postoperative length of stay as a function of indication for shoulder arthroplasty (SA) – either osteoarthritis (OA), rotator cuff disease (RCD), or proximal humerus fracture (PHF). We hypothesize PHF patients will have increased morbidity after SA. MethodsThe American College of Surgeons National Quality Improvement Program was queried from 2005 - 2015. Patients undergoing primary SA, including anatomic and reverse, were identified and the indication for procedure was identified using International Classification of Outcome variables were compared with bivariate analysis or multivariate binary logistic or linear regressions analysis that adjusted for differences in baseline patient characteristics. ResultsIn total, 8,083 patients undergoing SA were identified, with 6,806(84.2%) performed for OA, 898(11.1%) for RCD, and 379(4.8%) for PHF. Patients with PHF were the oldest, had the most medical comorbidities, dependent functional status, and ASA class. On multivariate analyses adjusting for these differences, patients with PHF had an increased rate of any complication (odds ratio [OR] = 2.07, P = 0.048), blood transfusion (OR = 4.66, P<0.001), and non-home discharge (OR = 3.06, P<0.001) relative to OA patients, as well as increased risk of blood transfusion (OR = 5.01, P<0.001) and non-home discharge (OR = 2.71, P<0.001) relative to RCD patients. Operative length was increased by 23 and 34 minutes and postoperative length of stay was increased by 1.4 and 1.2 days in PHF patients relative to OA and RCD patients, respectively. ConclusionsShoulder arthroplasty performed for PHF is associated with significantly increased postoperative morbidity and utilization of hospital resources. As bundled payment systems become more prevalent in orthopedics, these results will be helpful in determining appropriate payment bundles for shoulder arthroplasty. Level of EvidenceLevel III; Cohort Study