What Are the Risks?

Named Program Failures and Near-Collapses

Funding instability and program discontinuity: The Health Alliance for Violence Intervention’s primary advocacy focus has been sustainable funding, which reflects the field’s experience with grant-dependent programs that face reduction or closure when initial funding expires.¹ No named HVIP has fully closed and reopened in the documented record.

Undercapacity as a failure mode: The University of Chicago model runs roughly 20 specialists around the clock for a Level 1 Trauma Center with high gunshot volume. Dwayne Johnson, the VRP’s director, describes this as the staffing level required to actually reach patients consistently.² Most programs do not approach that scale.

Workforce Challenges

The workforce required for effective HVIPs is unusual and constrained. Programs need specialists who carry credibility in the communities from which patients come, typically people with lived experience of violence, incarceration, or the same social networks as the patient population. They also need, in most program models, some combination of clinical training, case management skills, and the emotional stability to engage repeatedly with patients in crisis and with the grief and anger of patient families.

That combination is rare. The University of Chicago requires roughly 20 specialists for around-the-clock operations serving Chicago’s South Side trauma patients. Dr. Randi Smith describes the same constraint at Grady: recruiting staff who have both community credibility and the professional skill to navigate hospital systems, insurance processes, housing applications, and legal service connections is a limiting factor on program scale.³

The University of Chicago’s VRP launched a vicarious trauma support program in 2022 specifically to address worker burnout and vicarious trauma among violence recovery specialists who work daily with patients who have been shot and families of homicide victims.⁴

Selection Bias and Evidence Limitations

The 98% acceptance rate documented at Grady Hospital sounds like a near-universal sample, but acceptance of initial services does not guarantee sustained engagement. Programs have not consistently documented how many enrolled patients complete the full sustained engagement that the Boston University Annals study found necessary for durable outcomes.

The selection concern runs in two directions. Programs may be enrolling patients who were already somewhat motivated to change, producing better outcomes than an unselected population would show. Alternatively, programs that deploy strong credible messenger staff may be reaching the highest-risk patients precisely because those patients respond to someone who has walked their path. That would mean program outcomes may not be replicable by programs with weaker community-embedded staff.

Most HVIP outcome data is program-reported and not independently verified. The Indianapolis comparison (3% vs. 8.7% without the program) and the VCU five-year figure (3.6% recidivism) are among the more credible Tier 2 data points, but they remain self-reported. The Johns Hopkins RCT and the Boston University target trial emulation are the only independent evaluations in the field, and both have population limitations (youth-only for the RCT; one program for the Boston study).⁵

The Long-Term Outcomes Gap

Most documented outcomes measure reinjury within the first year. The Boston University study extended to three years and found sustained risk reductions for patients with sustained engagement. But whether the reinjury reductions documented in year one persist through years five and ten is unknown. Violence prevention effects could decay as program contact ends and patients return to the same structural conditions (poverty, neighborhood violence, limited economic opportunity) that produced the injury in the first place.

The Boston University researchers documented that sustained engagement produces better three-year outcomes than brief contact, but noted that whether effects persist beyond three years remains undocumented.⁶

Scope Limitations: Who Is Not Reached

Rural and community hospital patients: Programs operate almost exclusively in Level 1 Trauma Centers in major urban areas. A gunshot victim admitted to a rural hospital, a community hospital without an HVIP, or a suburban trauma center outside a major metro has no access to the services, regardless of reinjury risk. Virginia’s statewide expansion to 12 hospitals is the clearest model for addressing this gap; it remains an exception rather than a norm.⁷

Patients who decline: Even at programs with exceptional acceptance rates, some patients decline initial engagement. The patients who decline are not necessarily those at lowest risk. Active refusal at the bedside may reflect the same distrust of institutions (including hospitals) that characterizes the highest-risk patients in the highest-violence networks.

Homicide survivors and family members: The family network of a homicide victim faces parallel retaliation pressure that programs focused on injured survivors do not address. Programs that expand to serve bereaved families are doing qualitatively different work and need different workforce competencies. Most programs do not formally include this population.

The pre-hospitalization window: If a patient is shot and does not go to the hospital (dying at the scene, refusing EMS transport, or being brought to a hospital without an HVIP), the intervention point is missed entirely. HVIP’s model depends on hospital admission as the activation trigger, which means patients who never reach a participating hospital are not reachable through this model.

Institutional and Structural Risks

Hospital culture resistance: Programs that are championed by a single physician or administrator are vulnerable to leadership transition. When the physician champion leaves, retires, or takes a different role, the program can lose its institutional patron and face budget cuts or scope reduction. Sustainable programs are ones that have built broad medical staff support, board-level endorsement, and funding independence from any single champion’s authority.

The “not our job” institutional default: The cultural shift that HVIPs require, from treating the wound to treating the person, is not complete in most hospital systems. Programs that have achieved the shift have done it through sustained demonstration of outcomes, not through a single policy decision. Programs that launch without that cultural groundwork face ongoing internal resistance.

Funding structure mismatch: The Kimball mechanism applies directly to HVIPs. Trauma hospitals are paid for procedures — surgeries, ED visits, inpatient days. Violence recovery specialists provide services that are not billable under most payer structures. The sustained case management work, community relationship-building, family engagement, and transportation to follow-up appointments: none of these generate revenue. The only documented sustainable reimbursement pathway is Connecticut’s Medicaid model for community services to do violence intervention outreach, and even that model has not been replicated in most states.⁸

This means that HVIPs depend on a combination of hospital institutional support, municipal public health investment, federal grants, and philanthropic funding — each of which can be withdrawn independently. Programs without a dedicated, sustainable funding stream are perpetually one budget cycle away from reduction.

ARPA expiration: Many programs that launched or expanded between 2021 and 2024 used American Rescue Plan Act (ARPA) funding. ARPA funds must be spent by the end of 2026. Programs that relied on ARPA for core operations without building a post-ARPA funding plan are facing sustainability pressure in 2025 and 2026.

Criticized Claims and Evidence Gaps

Reinjury comparisons across different time windows: The most dramatic outcome comparison in the field — Minneapolis’s 41% historical baseline versus 3% first-year rate — involves comparing a five-year cumulative statistic to a first-year rate. These are not equivalent measurements, and the comparison, while directionally meaningful, overstates the magnitude of change if read as an apples-to-apples comparison.⁹ Programs and advocates who cite this figure without this caveat are presenting it in a misleading frame.

The “three times less likely to be arrested” finding: This finding originates from Cooper, Eslinger & Stolley (2006), a study of adult repeat victims of violence on parole or probation — not the youth population most HVIP programs serve. The Johns Hopkins Cheng et al. (2008) RCT with youth aged 10-15 found trends toward reduced misdemeanor activity and aggression, but most results did not reach statistical significance, and the sample was small (113 with usable data). A 2022 Johns Hopkins review of seven RCTs found that “overall evidence of reduced risks for violence was mixed” and that RCTs “were underpowered, and all but one were vulnerable to selection bias.”

Return on investment figures: Cost-benefit calculations like the $82.5 million in health care costs avoided in Virginia, or the $82,765-$4 million in individual cost savings cited in research reviews, are estimates based on assumptions about what would have happened without the programs. These figures have the internal logic of avoided-cost calculations, but they are not independently verified and should be presented as projections rather than established outcomes.⁷

Safety Challenges in the Hospital Setting

Trauma centers are not neutral environments. Patients arriving after a shooting may be accompanied by family members or associates who are processing grief, fear, and anger, sometimes with weapons. The hospital itself can become a locus of conflict, particularly when patients and the individuals who shot them are in the same emergency department simultaneously.

HVIP programs have operated in this environment without documented safety incidents in the peer-reviewed or news record, but the potential is real and shapes how programs must be designed. Workers who engage patients and families in the immediate aftermath of a shooting are not in a safe, controlled clinical environment. They are in a setting where emotions are raw and the circumstances of the injury remain active. Experienced programs account for this in training, in the paired-staff model some use for initial contact, and in ongoing supervision protocols.

The hospital security apparatus provides some protection, but HVIP workers (particularly credible messengers who work outside the clinical hierarchy) can find themselves in ambiguous positions when tension escalates in a patient family interaction. No documented program has reported a formal safety incident involving HVIP staff, but program directors who have operated for years describe it as a constant management consideration rather than a resolved risk.¹⁰

The Community Hospital and Rural Gap in Detail

The coverage gap deserves more than a bullet point. In major metro areas with Level 1 Trauma Centers, the 30-50% national reinjury rate for untreated gunshot patients is mitigated (imperfectly, in the small fraction of eligible patients currently served) by the HVIP programs that exist. Outside those centers, it is not mitigated at all.

Rural areas with elevated rates of gun violence — including parts of Appalachia, the rural South, and Indigenous communities in the Mountain West — rarely have access to HVIP services because they lack both the Level 1 Trauma Center infrastructure and the community organization ecosystems that urban programs connect to. A patient in a rural trauma center has no warm handoff to an HVIP specialist because no such specialist exists within their region.

Virginia expanded from one program at VCU to 12 programs across the state, with Dr. Michel Aboutanos’s VCU program serving as the training model. The Virginia Mercury reported that the expansion required sustained state investment and hospital-level capacity that most states have not provided.¹¹

The Post-Discharge Environment Problem

Dr. Randi Smith’s program at Grady has documented that two-thirds of 14-24-year-old gunshot victims face food insecurity, housing instability, and financial precarity.¹² Programs operating in those conditions are connecting patients to resources while the patient’s post-discharge environment remains constrained. The Boston University researchers noted this structural limit: their findings may not generalize across all HVIP contexts because “there is no single, agreed-upon package of services for all HVIPs.”¹³