BackgroundProximal humerus nonunion is a challenging complication of fractures that can be treated surgically with either open reduction internal fixation (ORIF) or reverse total shoulder arthroplasty (RTSA). The few studies published on this subject have shown high rates of complications and revision surgery when RTSA has been performed for proximal humerus nonunion. The purpose of this study was to determine the rates of complications and revision of this procedure at our institution, as well as to identify any variables that may impact risks of complications and reoperations. MethodsA single-institution retrospective review of all patients who underwent RTSA for proximal humerus nonunion between 2005 and 2021 was performed. Nonunion was defined as imaging evidence of lack of union, at least 90 days after the index fracture. Patients with less than one year of clinical follow-up were excluded. Fifty patients were included, with the majority being female (78%). The mean age at time of RTSA was 71 (range: 54-86) years and most patients were initially treated nonoperatively (74%). Mean total follow-up was 49 (range: 11-130) months. Demographic and surgical variables were recorded. Primary outcomes were complications and reoperations. Complications were divided into surgical (those directly related to RTSA), or other (those unrelated to RTSA). Secondary outcomes included VAS pain scores and range of motion. ResultsA total of 17 shoulders (34%) sustained complications after RSA, with 10 (20%) requiring reoperation. Six patients (12%) sustained dislocations and 5 (10%) had radiographic evidence of humeral loosening. No variables examined, including nonoperative vs surgical management of the index fracture, prosthesis type, or management of tuberosities, influenced the risk of dislocation. Survivorship free from reoperation at 2 years was 73%. Younger age at time of RTSA and the presence of diabetes mellitus both increased the risk of reoperation significantly (p=0.013 and p=0.037, respectively). There was a trend towards increased risk of reoperation in patients who were treated with initial ORIF (HR=2.95); however, this did not reach statistical significance (p=0.088). Three patients (6%) sustained a periprosthetic fracture after a fall. ConclusionRTSA provides improved pain and function for properly selected patients with proximal humerus nonunion. Dislocation, humeral loosening, and reoperation rates remain high when RTSA is performed for nonunion compared to other diagnoses. In this study, younger age and diabetes mellitus increased the odds of reoperation. Every effort must be made to optimize implant stability and humeral component fixation when RTSA is performed for proximal humerus nonunion.