Adapting Cure Violence for Southwest Philadelphia: A Pilot Implementation Evaluation of the Penn Community Violence Prevention Program.

BACKGROUND: Community violence interventions (CVIs) aim to reduce firearm-related harm through strategies that center community engagement, credible messengers, and tailored outreach. Despite increasing adoption, few studies use implementation science methods to assess the implementation of adapted CVI models in new contexts. This pilot evaluation describes the implementation of the University of Pennsylvania Community Violence Prevention (PCVP) program, an adaptation of Cure Violence (CV) implementation strategies co-developed with community stakeholders. We assessed implementation outcomes and short-term participant-level impacts to inform future scale-up. METHOD: Using a type 3 hybrid implementation-focused design, we conducted a mixed-methods evaluation over an 18-week pilot period. Guided by the RE-AIM framework and Proctor's implementation outcomes, we examined reach, fidelity, acceptability, and appropriateness. Administrative data, staff-reported checklists, and geospatial outreach analyses were triangulated with survey data from 18 high-risk participants, informed by the COM-B model. Quantitative data were analyzed descriptively and using Pearson correlations; qualitative responses underwent rapid thematic analysis. RESULTS: The PCVP team (two full-time, two part-time outreach workers, and one manager) engaged 38 participants, reaching 75% of target caseloads and covering 70% of shooting incidents within two blocks and two weeks. Fidelity was high for case management and community engagement, but low for violence detection/interruption due to staffing limitations. Participants reported high acceptability and appropriateness of the intervention (M = 4.39, SD = 0.98), with COM-B scores indicating strong engagement across capability, opportunity, and motivation domains (M = 25.64/30). Correlations revealed significant links between social/physical opportunity and motivation (r > .70, p < .01). CONCLUSIONS: This study demonstrates the feasibility and acceptability of a community-adapted CVI model. High acceptability and engagement support continued implementation, though limitations in fidelity to violence interruption underscore the need for targeted staffing and funding. Context-sensitive adaptation, robust community partnerships, and longer-term implementation research are essential for sustaining and scaling CVI efforts.