Abstract
BACKGROUND: Total shoulder arthroplasty (TSA) has seen a dramatic rise in utilization, particularly in the elderly population. While indications have expanded beyond glenohumeral osteoarthritis (OA) to include proximal humerus fractures (PHFs), robust evidence comparing outcomes between these groups remains limited. We sought to compare the rate of surgical and medical complications of TSA (reverse or anatomic) performed for PHFs versus OA in elderly patients at 1 and 3 years post-operatively. METHODS: The TriNetX Global Collaborative Network, comprising 183 million patients across 155 healthcare organizations, was retrospectively queried for patients aged ≥65 years undergoing TSA. Two mutually exclusive cohorts were identified: patients undergoing TSA within 3 weeks of a PHF diagnosis (PHF cohort) or those with a pre-existing diagnosis of OA (OA cohort). Exclusion criteria included age <65 years, concurrent diagnoses of PHF and OA, and polytrauma. Patients were propensity score matched 1:1 on demographics and comorbidities. Outcomes assessed included surgical complications (infection, periprosthetic fracture, prosthetic failure, reoperation, persistent pain, and opioid use) and medical complications (cardiac, thromboembolic, respiratory, infectious, and mortality) at 1 and 3 years post-operatively. RESULTS: A total of 57,203 TSA patients were identified, 52,597 (91.9%) for OA and 4,606 (8.1%) for PHFs. Following propensity matching, 9,190 patients (4,595 per cohort) were analyzed. At 1 year, PHF patients exhibited higher rates of nearly all complications, including periprosthetic fracture, prosthesis-related complications, readmission, emergency department visits, transfusion, and opioid use (P < .05), though persistent shoulder pain was lower compared with OA patients (risk ratio: 0.807; P < .0001). By 3 years, PHF patients remained at higher risk for periprosthetic complications, respiratory events, sepsis, and mortality (P < .05). CONCLUSION: Elderly patients undergoing TSA for PHFs face substantially greater risks of post-operative complications and healthcare utilization compared to those treated for OA, despite reporting lower rates of persistent pain at 1 and 3 years. These findings highlight the need for tailored perioperative management, risk counseling, and careful patient selection in this vulnerable population.