High Variation in Surgical Rates for Pediatric Clavicle Fractures and the Influence of Socio-economic Factors

Background: Historically, treatment strategies for pediatric midshaft clavicle fractures were non-operative with a brace or sling. Increasingly, operative treatment has been offered to children based on evidence from adult trials or retrospective reviews in children. Therefore, the decision to operate is guided by low levels of evidence. The purpose of this study is to describe the treatment variation of midshaft clavicle fractures and to evaluate which patient or hospital factors are associated with surgery.

Methods: Patient encounters with a diagnosis of midshaft clavicle fracture between the ages of 0-18 from 2013 to 2019 were collected from the Pediatric Hospital Information System (an administrative database of billing and clinical data). Exclusion criteria included absolute indications for surgery (polytrauma or open fractures) or incomplete data. Data obtained included procedure/CPT codes, admit/discharge dates, gender, age, race/ethnicity, and insurance status. Regression analyses were used to determine which factors were associated with operative treatment.

Results: A total of 39,222 patients with midshaft clavicle fractures were identified from 40 hospitals, with 1254 (3.19%) of these patients receiving operative treatment. Hospitals treated a median of 92.5 (IQR: 54.75, 171) fractures per year. Operative rates varied significantly (p = 0.0001) by hospital, ranging from 0.36% to 11.68% (Median: 4.29%; IQR: 2.06%, 5.79%). Of patients treated operatively, 99% were at least 10 years old. In considering patients 10 years old or older, operative and non-operative treatment groups were found to differ significantly in their distributions of gender, race/ethnicity, insurance status, and age (p < 0.0001 for all). Furthermore, patients receiving operative treatment had significantly increased odds of being male (OR: 1.418; 95% CI: 1.213, 1.667), white (OR: 1.924; 95% CI: 1.707, 2.171), and having private insurance (OR: 1.693, 95% CI: 1.500, 1.912). Significantly increased odds of operative treatment were observed for older age (OR: 1.487; 95% CI: 1.442, 1.534).

Conclusion: There is a high level of variation by hospital for treatment of clavicle fractures. Older, male, white, non-Hispanic children, and children with private insurance are associated with higher operative treatment rates. High treatment variation for clavicle fractures result from limited evidence to guide clinical decision making. The high treatment variation suggests that children may be undergoing unnecessary procedures or they are undertreated. The decision to treat operatively may be influenced by non-clinical socioeconomic factors. Prospective research on treatment outcomes is needed and this study further motivates a randomized controlled trial for clavicle fractures in children.