Effects of a Hospital-Based Violence Intervention Program on Community Violence in Boston, Massachusetts : A Target Trial Emulation.

Abstract

Hospital-based violence intervention programs (HVIPs) are widespread, but their effectiveness for violence prevention is unclear. To determine the effects of Boston Medical Center's HVIP on future violence outcomes among violently injured young adults. Target trial emulation using observational data. Boston, Massachusetts. Young adults aged 16 to 34 years who survived a shooting or stabbing. Target trials of 2 treatment strategies using the same eligibility criteria, time zero, and control group were emulated: 1) any treatment: engaging with the HVIP within 1 month of injury and 2) sustained treatment: initiating within 1 month and engaging more than 4 of the first 8 weeks. Combined measure of violent reinjury or violence perpetration at 1, 2, and 3 years, using hospital and police data. 1328 patients met criteria; 565 (42.5%) initiated within 1 month. Of these, 58 (10.2%) sustained engagement. In the any-treatment analysis, estimated cumulative incidence was roughly equal between the treatment and control strategies at 1, 2, and 3 years. In the sustained engagement analysis, treatment was associated with considerably lower cumulative incidence (4.5% [95% CI, 1.1% to 9.3%] at 1 year; 5.1% [CI, 1.1% to 9.3%] at 2 years; 6.4% [CI, 1.4% to 12.9%] at 3 years) versus the control strategy (8.7% [CI, 6.6% to 10.0%] at 1 year; 12.3% [CI, 10.2% to 14.5%] at 2 years; 14.3% [CI, 11.8% to 16.6%] at 3 years), with corresponding risk reductions of 47.6% (CI, -19.8% to 86.7%), 58.5% (CI, 21.6% to 91.2%), and 55.3% (CI, 4.9% to 90.2%). Confidence intervals were wide. Despite our target trial emulation approach, results could be confounded by unmeasured factors associated with program engagement. Although HVIPs can improve long-term violence outcomes, these effects seem to require intensive participant engagement. Fund for a Safer Future.