Multicenter, Prospective, Observational Trial of Nonoperative Versus Operative Treatment for High-Energy Midshaft Clavicle Fractures

OBJECTIVES: Evaluate if nonoperative or operative treatment of displaced clavicle fractures delivers reduced rates of nonunion and improved DASH scores METHODS: Design: Multicenter, prospective, observational Setting: Seven Level 1 Trauma Centers in the United States Patient Selection Criteria: Adults with closed, displaced (100% displacement/shortened >1.5cm) midshaft clavicle fractures (OTA 15.2) treated nonoperatively, with plates and screw fixation, or with intramedullary fixation from 2003-2018. Outcome Measures and Comparisons: DASH scores (2 weeks, 6 weeks, 3, 6, 12, and 24 months), reoperation, and nonunion were compared between the nonoperative, plate fixation, and intramedullary fixation groups. RESULTS: 412 patients were enrolled, with 203 undergoing plate fixation, 26 receiving intramedullary fixation, and 183 treated nonoperatively. The average age of the nonoperative group was 40.1 (range 18-79) years versus 35.8 (range 18-74) in the plate group and 39.3 (range 19-56) in the intramedullary fixation group (p=0.06). 140 (76.5%) patients in the nonoperative group were male compared to 154 (75.9%) in the plate group and 18 (69.2%) in the intramedullary fixation group (p=0.69). All groups showed similar DASH scores at 2 weeks, 12 and 24 months (p>0.05). Plate fixation demonstrated better DASH scores (median=20.8) than nonoperative (median=28.3) at 6 weeks (p=0.04). Intramedullary fixation had poorer DASH scores at 6 weeks, 3 and 6 months than plate fixation and worse DASH scores than nonoperative at 6 months (p<0.05). The nonunion rate for nonoperative treatment (14.6%) was significantly higher than the plate group (0%) (p<0.001). CONCLUSION: Operative treatment of displaced clavicle fractures provided lower rates of nonunion than nonoperative treatment. Except at 6 weeks, no difference was observed in DASH scores between plate fixation and nonoperative treatment. Intramedullary fixation resulted in worse DASH scores than plate fixation at 6 weeks, 3 and 6 months, and worse DASH scores than nonoperative at 6 months. Implant removal was the leading reason for reoperation in the plate and intramedullary fixation groups, while surgery for nonunion was the primary reason for surgery in the nonoperative group. LEVEL OF EVIDENCE: Therapeutic Level II. See Instructions for Authors for a complete description of levels of evidence.