Neither operative nor nonoperative approach is superior for treating displaced midshaft clavicle fractures: a partially blinded randomized controlled clinical trial

Aims To compare the functionality of adults with displaced mid-shaft clavicular fractures treated either operatively or nonoperatively and to compare the relative risk of nonunion and reoperation between the two groups. Methods Based on specific eligibility criteria, 120 adults (median age 37.5 years (interquartile range (18 to 61)) and 84% males (n = 101)) diagnosed with an acute displacedmid-shaft fracture were recruited, and randomized to either the operative (n = 60) or nonoperative (n = 60) treatment group. This randomized controlled, partially blinded trial followed patients for 12 months following initial treatment. Functionality was assessed by the Constant score (CS) (assessor blinded to treatment) and Disability of the Arm, Shoulder and Hand (DASH) score. Clinical and radiological evaluation, and review of patient files for complications and reoperations, were added as secondary outcomes. Results At 12 months, 87.5% of patients (n = 105) were available for analysis. The two groups were well balanced based on demographic and fracture-related characteristics. At six weeks of follow-up a significant difference in DASH score (p < 0.001) was found in favour of operative treatment. The functionality at 12 months of follow-up based on CS and DASH was excellent in both groups (CS > 90 points and DASH < 10 points) with no significant difference (p = 0.277 for DASH and p = 0.184 for CS) between the two groups. The risk of symptomatic nonunion was significantly higher in the nonoperative group (p = 0.014), with a relative risk of 9.47 (95% confidence interval (CI) 1.26 to 71.53) in this group compared to the operative group. The number-needed-to-treat to avoid one symptomatic nonunion was 6.2. Initial treatment and age were factors significantly associated with nonunion in a logistic analysis. There were26% in both groups (n = 14 in operative group and n = 15 in nonoperative group) who requiredsecondary surgery, with most indications in the nonoperative group mandatory due to nonunion compared to most relative indications in the operative group requiring intervention due to implant irritation. Conclusion Superiority was not identified with either an all-operative or all-nonoperative approach. The functionality at short term (within six weeks) seems igreater following operative treatment but was not found at one year. The risk of nonunion is significantlyhigher with nonoperative treatment. However, an all-operative approach to lower the nonunion risk may result in unnecessary surgery and is not recommended.