Exploring programme implementation of a US paediatric hospital-based violence intervention programme by injury mechanism.

Background: Hospital-based violence intervention programmes (HVIPs) support recovery following interpersonal violence, with most patients participating following firearm injuries. There remains insufficient understanding of who HVIPs serve and how programmes are implemented, especially among paediatric patients. We sought to describe the implementation of a paediatric HVIP and examine relationships between HVIP implementation metrics and mechanism of injury (firearm vs non-firearm).

Methods: We conducted a retrospective cohort study using HVIP records from 2018 to 2023, identifying 2021 patients (8-18 years) treated at our urban paediatric trauma centre in the northeastern USA, 416 of whom enrolled in our HVIP. We conducted bivariate and multivariate regression analyses examining differences in patient-level characteristics and HVIP implementation by injury mechanism.

Results: Most patients (94%) experienced non-firearm injuries. Firearm-injured patients were more likely to enrol in our HVIP (adjusted OR=4.01, 95% CI 2.64 to 6.14) than non-firearm-injured patients, though non-firearm-injured patients represented 85% of HVIP participants. In comparison to non-firearm-injured patients, firearm-injured patients had longer programme duration (adj_B=43.73, 95% CI 4.84 to 82.63 days), more documented encounters (adj_B=16.30, 95%CI 3.44 to 29.16) and more recovery goals (adj_B=3.37, 95%CI 1.21 to 5.52). HVIP goal resolution and graduation rates did not significantly differ by mechanism.

Conclusion: Our study identified metrics to describe HVIP implementation among paediatric patients with diverse injury types and documented consistent HVIP engagement, retention and outputs for those with firearm and non-firearm injuries alike. Our work suggests the relevance, and acceptability of HVIPs for paediatric patients with diverse injuries and offers a framework for process measurement in future implementation, outcome and impact evaluations.