Short-term Complications for Proximal Humerus Fracture Surgery Have Decreased: An Analysis of the National Surgical Quality Improvement Program Database

Multiple surgical treatments exist for proximal humerus fractures (PHFs), but current practice patterns and short-term complication profiles remain poorly understood. This is in part due to changes in treatment paradigms over the past decade. A more thorough understanding of the evolution in management over this time as well as an appreciation of the preoperative factors associated with both the chosen surgical modality and short-term complications will help inform future surgical considerations.In this study, we sought to: (1) characterize trends in the surgical management of PHFs over time, including usage rates of various surgical modalities and changes in complication rates; (2) identify preoperative variables associated with the selection of surgical modality; and (3) assess the independent covariates of acute 30-day complications, including demographic variables, injury characteristics, and treatment type.The National Surgical Quality Improvement Program (NSQIP) database was queried using Current Procedural Terminology and ICD-9 and ICD-10 codes to identify individuals with PHF undergoing open reduction and internal fixation (ORIF), hemiarthroplasty (HA), or reverse total shoulder arthroplasty (RTSA) from 2007 to 2018. The NSQIP database was utilized because of its collection of detailed preoperative demographic information and large repository of clinically derived data, which is felt to be more accurate than claims or administrative data. In total, 5889 patients with PHFs met the inclusion criteria. Patients 17 years or older were included. Patients with isolated greater tuberosity, humeral shaft, and distal humerus fractures; nonunions; malunions; and those undergoing revision procedures were excluded. A Cochran-Armitage test was used to evaluate surgical trends over time. Multivariable logistic regression models were created to identify covariates associated with surgical modality and complications. Although complications were either classified as major or minor, specific complications were also individually analyzed to avoid potentially misleading conclusions associated with pooling.The proportion of patients with PHFs undergoing RTSA (4% in 2007 and 34% in 2018; p < 0.001) and ORIF (46% in 2007 and 57% in 2018; p < 0.001) increased over time, and the proportion of those undergoing HA (50% in 2007 and 9% in 2018; p < 0.001) decreased. Across each surgical modality, minor complication rates decreased over time (RTSA: 10% in 2018; p < 0.001; ORIF: 5% in 2018; p = 0.01; and HA: 6% in 2018; p = 0.01). After controlling for confounding variables like diabetes, chronic obstructive pulmonary disease, congestive heart failure, dialysis, and preoperative blood transfusion, the following factors were independently associated with an increased odds of a patient undergoing RTSA rather than HA: older age, higher BMI, independent functional status, and smoking. The following factors were independently associated with a decreased odds of a patient undergoing ORIF rather than RTSA and HA: older age, higher BMI, higher American Society of Anesthesiologists (ASA) classification, smoking, steroid/immunosuppressant use, as well as three- and four-part fractures. After controlling for age, ASA classification, functional status, and preoperative blood transfusion, we also found that in the latter part of the study period, arthroplasty (RTSA and HA) was no longer independently associated with 30-day major or minor complications compared with ORIF.The increasing utilization of RTSA and decreasing short-term complication rates for fixation and arthroplasty alike represent a substantial change compared even with recent historic norms in the management of proximal humerus fractures. Quantifying demographics, injury characteristics, and comorbidities associated with both the choice of surgical modality and complications serves as the groundwork for decision support tools, which can inform patients and surgeons of the probability of a particular surgical modality being chosen and the risk of complications, based on national benchmarks. Future studies should investigate longer term complication rates, as many differences between the approaches being compared might occur beyond the perioperative period and our study had no means to address questions about complications beyond that period, but obviously those must be considered when choosing a treatment for patients with these injuries. Future work might also investigate the mechanisms behind the decreasing rates of short-term complications.Level III, therapeutic study.