Do Hospital-Based Violence Intervention Programs Really Reduce Community Violence? New Evidence From Boston

Community violence remains one of the most urgent public health challenges in the United States. For young people in urban areas, homicide continues to rank among the leading causes of death, while nonfatal firearm and stabbing injuries occur at even higher rates. Survivors of violence often face a long list of challenges after discharge, including posttraumatic stress disorder, physical disability, housing instability, unemployment, and disrupted education. Even more concerning, individuals who survive violent injuries face roughly double the risk of violent reinjury or future violence perpetration in the years that follow.

Hospital-based violence intervention programs, commonly known as HVIPs, were developed to interrupt this cycle. These programs engage survivors of violence at a critical moment, often while they are still hospitalized, and connect them with advocates, social services, and long-term support. Over the past decade, HVIPs have expanded rapidly across the United States, with more than 60 programs now operating nationwide.

Despite this growth, a central question has remained unresolved. Do HVIPs actually reduce future violence, or do they primarily improve social and health outcomes without measurable effects on reinjury or retaliation? A major new study published in Annals of Internal Medicine provides some of the strongest evidence to date, suggesting that HVIPs can reduce community violence, but only when participants remain deeply engaged over time.

A Landmark Study From Annals of Internal Medicine

In January 2026, researchers published a large observational study titled Effects of a Hospital-Based Violence Intervention Program on Community Violence in Boston, Massachusetts: A Target Trial Emulation. The study evaluated Boston Medical Center’s Violence Intervention Advocacy Program, or VIAP, one of the longest-running HVIPs in the United States.

The research team analyzed data from young adults aged 16 to 34 who survived gunshot or stabbing injuries and were treated at Boston Medical Center between 2013 and 2022. Using linked hospital records, trauma registries, death records, and police data, the investigators tracked violent reinjury and violence perpetration outcomes for up to three years after the initial injury.

What sets this study apart is its use of a method known as target trial emulation. Because randomized controlled trials are often unethical or impractical for HVIPs, which are considered standard care in many hospitals, the researchers designed their analysis to closely mimic what an ideal randomized trial would look like, using real-world data.

Why Evaluating HVIPs Is So Challenging

Evaluating violence prevention programs is notoriously difficult. Violence-related outcomes are relatively rare events, which means studies require large sample sizes and long follow-up periods. At the same time, not all eligible patients choose to engage with HVIP services, and many who do engage participate only briefly.

Randomizing patients to receive or not receive HVIP services raises ethical concerns, since these programs provide essential support services such as housing assistance, mental health referrals, and medical follow-up. As a result, most HVIP research relies on observational data, which can be vulnerable to bias.

The Boston study addressed these challenges by carefully defining eligibility criteria, treatment strategies, follow-up periods, and outcomes in advance. It then applied advanced statistical techniques to reduce bias related to differences in who did and did not engage with the program.

Two Levels of Engagement Were Tested

The researchers evaluated two distinct HVIP engagement strategies.

The first was any engagement, defined as having at least one interaction with a VIAP advocate within one month of the injury. This reflects the minimum level of contact that many HVIP participants experience.

The second was sustained engagement, defined as initiating contact within one month and engaging in more than four of the first eight weeks after injury. According to program staff, this level of engagement allows advocates to build trust, identify deeper needs, and deliver more intensive support.

Both strategies were compared with a control group consisting of individuals who did not engage with the HVIP within the first month.

Who Was Included in the Study

A total of 1,328 young adults met the study’s eligibility criteria. Most participants were Black or Hispanic, male, and in their early twenties, reflecting the populations most affected by community violence in Boston. More than half were treated for gunshot wounds, and many lived in neighborhoods marked by high levels of racialized economic segregation.

Importantly, fewer than half of eligible patients engaged with the HVIP within the first month. Among those who did, only about 10 percent achieved sustained engagement. This finding alone highlights one of the biggest challenges facing HVIPs nationwide: maintaining consistent participation among clients with complex and often unstable life circumstances.

The Key Findings on Violence Reduction

The study’s results were striking and nuanced.

For individuals who had any engagement with the HVIP, there was no meaningful reduction in violent reinjury or violence perpetration over one, two, or three years of follow-up. In other words, minimal contact alone did not appear sufficient to reduce future violence.

In contrast, individuals who sustained engagement with the program experienced substantially lower rates of community violence. At one year, the cumulative incidence of violent reinjury or perpetration was about 4.5 percent in the sustained engagement group, compared with 8.7 percent in the control group. At three years, the difference widened further, with sustained participants showing roughly half the risk of violence compared with those who did not engage.

Although the confidence intervals were wide due to the relatively small number of sustained participants, the pattern consistently favored intensive engagement. The estimated risk reductions ranged from approximately 47 percent to nearly 60 percent across different follow-up periods.

Why Dosage Matters in Violence Intervention

One of the most important lessons from this research is that dosage matters. HVIPs are not magic bullets that work automatically after a single interaction. Their effectiveness appears to depend heavily on the intensity and continuity of engagement.

Participants in the sustained engagement group had far more contact with advocates, with a median of more than 15 service interactions compared with just three among those with any engagement. They also had significantly more needs identified and met across critical areas such as employment, education, legal support, injury recovery, and family stability.

Some of the most challenging needs to address, including education and employment, were met at higher rates among sustained participants. These structural supports may help explain why long-term violence outcomes improved in this group.

Implications for Public Health and Policy

This study carries important implications for policymakers, hospital systems, and community organizations investing in violence prevention.

First, it suggests that HVIPs can reduce future violence, but only if programs are designed and resourced to support sustained engagement. Simply offering HVIP services is not enough if most participants disengage early.

Second, it highlights the importance of removing barriers to participation. Survivors of violence may face competing demands, trauma-related symptoms, fear of retaliation, or mistrust of institutions. Programs that invest in credible messengers, flexible scheduling, and long-term relationship building may be more successful in achieving high-dose engagement.

Third, the findings underscore the need for HVIPs to be integrated into broader community violence intervention ecosystems. Risk of reinjury and retaliation is highest in the first few months after injury, sometimes before sustained engagement can take hold. Partnerships with street outreach teams and community-based mediators may help fill this critical gap.

Limitations to Keep in Mind

While the study is among the strongest observational evaluations of an HVIP to date, it is not without limitations. Because participation was voluntary, unmeasured factors such as motivation or readiness to change may have influenced who sustained engagement and who did not.

The outcomes data also relied on hospital and police records within the Boston area. Violence events that occurred outside the city or were not reported may have been missed. Additionally, the results reflect a single, well-resourced program and may not fully generalize to all HVIPs.

Nevertheless, the use of target trial emulation, long follow-up periods, and comprehensive data sources strengthens confidence in the findings.

A Turning Point for HVIP Research

The Boston Medical Center study represents a turning point in how hospital-based violence intervention programs are evaluated. It moves the conversation beyond whether HVIPs work at all and toward a more precise question: under what conditions do they work best?

The answer, based on this evidence, is clear. HVIPs have real potential to reduce community violence, but only when participants receive sustained, intensive support. Future research and funding should focus on improving engagement, addressing readiness, and integrating HVIPs into comprehensive violence prevention strategies.

If communities want to break the cycle of violence, this study suggests that depth, not just reach, may be the key.

Source

Jay J, Pino E, Georges M, et al. Effects of a Hospital-Based Violence Intervention Program on Community Violence in Boston, Massachusetts: A Target Trial Emulation. Annals of Internal Medicine. Published January 27, 2026. DOI: 10.7326/ANNALS-25-01678.

Disclaimer

This article is for informational and educational purposes only. It summarizes findings from a published medical research study and does not constitute medical, legal, or policy advice. Interpretations are based on the original authors’ reported results, and readers should consult the full article in Annals of Internal Medicine for detailed methodology and limitations.