Outcomes of Clavicle Nonunion Repair: A Retrospective Cohort Study of 125 Patients

OBJECTIVES: To describe outcomes following operative repair of clavicle nonunions and identify risk factors for recalcitrant nonunion. METHODS: Design: Retrospective cohort study. Setting: Two academic Level-1 trauma centers. Patient Selection Criteria: Included were adults who underwent repair of a clavicle fracture (AO/OTA 15) nonunion from January 2004 to 2024. Outcome Measures and Comparisons: The primary outcome was recalcitrant nonunion, defined as nonunion requiring additional revision surgery after nonunion repair or absence of healing at final follow-up. Univariate and multivariable logistic regression was performed to determine associations between patient, nonunion, and treatment characteristics ( fixation construct and position, use of bonegraft, substitutes or compression) with recalcitrant nonunion. The secondary outcome was reoperation for complications other than recalcitrant nonunion. RESULTS: 125 patients were included (mean age 44 years [range 18–82], 55% male). The median follow-up was 16 months and 82% of acute fractures were treated nonoperatively. No patients presented with confirmatory criteria of fracture-related-infection. Twenty-four patients (19%) developed recalcitrant nonunion, with 17 undergoing revision. BMI (5-point increase, OR 3.38, p < 0.001), smoking (OR 4.49, p = 0.020), nonunion duration (3-month increase, OR 1.04, p = 0.013), age (10-year increase, OR 1.62, p = 0.042), and non-diaphyseal nonunion location (OR 4.79, p = 0.013) were identified as independent risk factors for recalcitrant nonunion in multivariable analysis. No treatment characteristics were associated with recalcitrant nonunion in univariate analysis (p > 0.05). Twenty-five patients (20%) underwent reoperations for complications other than recalcitrant nonunion. CONCLUSIONS: Operative repair for clavicle nonunion failed in 1 in 5 patients. Higher BMI, smoking, longer nonunion duration, older age, and non-diaphyseal nonunion locations were associated with increased risk of failure. No treatment characteristics were associated with failure. Surgeons may target modifiable risk factors, such as smoking and BMI, to achieve more reliable healing rates. LEVEL OF EVIDENCE: Prognostic Level III.