Displaced Proximal Humerus Fractures in Older Patients: Shoulder Surgeons Versus Traumatologists

Proximal humerus fractures are relatively common, with treatment dependent on fracture-specific, patient-specific, and surgeon-specific factors. This study sought to identify preferences among shoulder specialists and orthopedic traumatologists in the treatment of proximal humerus fractures. An anonymous Internet survey of fellowship-trained shoulder surgeons and traumatologists was conducted with radiographs and select computed tomography images of 15 unique displaced proximal humerus fractures. Participants were asked to classify each case according to Neer criteria and choose management from a list of options. Groups were analyzed using chi-square test for independence, paired t test, and Fleiss' kappa within and between each group. Among shoulder surgeons, there were a total of 19 cases selected for nonoperative management, 204 cases selected for open reduction and internal fixation (ORIF), and 122 cases selected for arthroplasty. Among traumatologists, there were 44 cases selected for nonoperative management, 234 for ORIF, and 67 for arthroplasty. Fleiss' kappa for intraobserver agreement on treatment choice was 0.26 for shoulder surgeons and 0.18 for traumatologists, and chi-square test for independence was significant between the 2 groups (P<.001). Paired t test of the average treatment proportions was significant for nonoperative management and arthroplasty (P=.003) but not significant for differences in rates of ORIF. These results confirm poor consistency in Neer classification among surgeons and suggest that shoulder surgeons were more likely to consider arthroplasty for treatment and that traumatologists were more likely to use ORIF or to manage patients nonoperatively. These variations in care may translate to differences in outcome and cost.