A Systematic Review on the Effectiveness of the Cure Violence Approach

Abstract

The Cure Violence approach applies public health epidemic control strategies to reduce violence in highly impacted communities and countries. This paper conducts a systematic review to identify studies analyzing the effectiveness of the Cure Violence approach and provides an overview of their findings. A protocol was developed using the PRISMA guidelines. For this study, we examined findings related only to shootings and killings because these are the outcomes to which the Cure Violence approach is applied. We utilized PubMed database, Google Scholar, and the Cure Violence Global organizational files to look for relevant studies. The review identified 13 papers analyzing 27 program sites and offering 83 findings on the impact of the approach on either shootings or killings. Overall, 68.7% of the findings indicated a reduction in shootings or killings, with 32.5% demonstrating statistically significant decreases. Findings varied geographically, with Baltimore accounting for the highest proportion of sites reporting either an increase in shootings or killings or non-significant results. Outside of Baltimore, 95.8% of sites demonstrated reductions in violence, with 54.2% achieving statistical significance. Notable findings include a 52% reduction in killings in Chicago, a 63% reduction in shootings in New York City, and a 74% reduction in killings in Cali, Colombia. The context surrounding the Baltimore findings underscores the importance of contextual factors in determining program effectiveness. Key components influencing program effectiveness include targeting individuals and groups at highest risk, employing credible messengers to reach individuals and groups at highest risk, real-time violence interruption and conflict mediation, sustained case management, structured monitoring systems with rapid response mechanisms, continuous monitoring of impact on shootings and killings, and sufficient, sustained financial investment. Recommendations to researchers are presented regarding the importance of including contextual factors in future research and understanding the importance of their role in guiding communities in improving the effectiveness of violence prevention programs.

Introduction

Violence remains one of the most urgent public health challenges of our time. Globally, it ranks among the leading causes of death for individuals aged 15 to 49 ¹ and affects people in nearly all regions and ethnic groups. ²  In the United States (U.S.), firearm violence became the leading cause of death for people under the age of 20 in 2020. ³ For decades, violence has disproportionately affected the developing world and impoverished, marginalized populations. ⁴ In the United States, economically marginalized communities of color have historically experienced ⁵ and continue to experience violence at disproportionately high rates. ⁶

The Cure Violence model is one of the most prominent approaches to preventing violence. The Cure Violence model consists of 3 core components:

• Detecting and Interrupting Potentially Violent Events – Violence interrupters use their credibility, trust, and access to reach those at the highest risk, identifying conflicts before they escalate, mediating disputes, and following up to prevent relapses. This requires strategic planning, real-time conflict assessment, and rapid intervention in response to violent incidents.

• Providing Ongoing Behavior Change Support – Outreach workers build sustained relationships with high-risk individuals, engaging with them multiple times a week to provide mentorship, identify and address trauma, develop risk-reduction plans, and connect them to services such as education, employment, and mental health resources.

• Shifting Community Norms to Reject Violence – Through public education campaigns, community events, and coordinated responses to shootings, program staff actively work to change social attitudes toward violence, engaging a broad network of stakeholders in reinforcing nonviolent conflict resolution.

Over the past 25 years, Cure Violence Global (CVG) has developed, refined, and implemented an evidence-based approach that has been replicated in communities worldwide. This work has also contributed to the emergence of the broader field of community violence intervention (CVI), which includes various other models that often also employ outreach workers and violence interrupters.

There has been a policy shift recently, with a growing number of communities investing in CVI approaches. As these programs expand, it is increasingly important to understand not only their effectiveness but also the specific elements that work, for whom, and under what circumstances. Researchers play a critical role in advancing these efforts.

To date, there has not been a meta-analysis or systematic review of the Cure Violence approach itself. A recent preliminary meta-analysis of “street outreach” includes the Cure Violence approach along with several other models that have significant differences from the Cure Violence approach. ⁷ The Cure Violence approach, because it is based on epidemic control, has a specificity to how outreach is performed and conflicts are interrupted that distinguishes it from other models. Further, there are other elements of epidemic control included in the approach that distinguish the approach, including its focus on prevention of spread, changing norms, and daily monitoring. Therefore, we believe this warrants a separate analysis of the evidence specific to the Cure Violence approach.

Several other papers have presented reviews of the evidence on the Cure Violence approach that were not systematic reviews and did not include all studies on the Cure Violence approach.^8-11 Several of these reviews also mistakenly identify programs that were separate models from the Cure Violence approach, such as Pittsburgh’s One Vision One Life.^9-11 The evaluators of the Pittsburgh program noted that it did not include violence interrupters or work with people at the highest risk, essential elements of the Cure Violence approach. ¹² One review reported results for the Cali Cure Violence program for the entire community, rather than the results from the specific target area. ⁸

This paper presents evidence specifically on the Cure Violence approach and explores the question of whether the Cure Violence approach should be considered effective. This paper adopts a field-based, public health perspective, highlighting when the approach has been most effective, and seeks to identify the factors that lead to a program not achieving its desired outcomes. To achieve this, the paper examines the broader context surrounding the findings, highlighting key factors and offering practical guidance to equip researchers and practitioners with the tools to better support the communities they serve.

Overview and History of the Cure Violence Approach

CVG was founded in 1995 as a public health initiative applying epidemic control methods to reduce violence. Drawing on global experience in combating epidemics across Africa, Asia, and North America, ¹³ CVG’s approach adapts proven public health strategies, specifically those used in epidemic intervention, to interrupt and prevent the progression and spread of violence. ¹⁴

In 2000, CVG launched its first community site under the name CeaseFire in Chicago’s West Garfield Park neighborhood, one of the most violent in the U.S. at the time. Reductions in violence were apparent in the first month of the launch, and by the end of 2000, shootings had dropped 67%, prompting an expansion of the program in Chicago. ¹⁵  By 2008, CVG began offering intensive training and support to communities seeking to replicate its approach, starting in Baltimore ¹⁶ and New York City. ¹⁷ Its international expansion followed, first in Sadr City and Basra, Iraq, ¹⁸ and later in Honduras, Mexico, and across Latin America. ¹⁹ This approach has since been replicated in over 50 U.S. cities and more than 15 countries worldwide. ²⁰

Methods

Search Strategy and Selection Criteria

A protocol was developed using the PRISMA guidelines. For this study, we examined findings related only to shootings and killings because these are the outcomes to which the Cure Violence approach is applied.

The literature search was conducted on May 5, 2025, using the PubMed database, Google Scholar, and the Cure Violence Global organizational files to look for relevant studies examining the relationship between the implementation of the Cure Violence approach and levels of shooting or killings in a community. For the PubMed and Google Scholar searches, an initial search used the term “Cure Violence.” Additional searches used the names of known replications of the approach before 2012, when the name changed to Cure Violence. ²¹ These additional searches included the terms “ceasefire chicago,” “ceasefire AND ‘safe streets,’” “ceasefire AND aim4peace,” “ceasefire AND ‘save our streets,’” “ceasefire AND snug,” and “ceasefire AND ‘hanover park.’” For the search of the Cure Violence Global files, we identified the files categorized over time as research related to the Cure Violence approach. The papers were screened independently by the lead author, and no automation tools were used.

Excluded Studies

Two searches were conducted, yielding 18 papers from the PubMed database and 3449 from Google Scholar. Additionally, a review of Cure Violence Global internal files identified 1429 documents as potentially relevant papers on the approach. These internal files were in a folder where files analyzing the Cure Violence approach were set aside by Cure Violence Global staff. After initial screen for duplicates and nonrelevant papers, remaining records were screened using 3 inclusion criteria: (1) analysis of the implementation of the Cure Violence approach; (2) quantitative findings on the impact of the approach on shootings or killings; and (3) inclusion of both implementation analysis and control areas (Figure 1).

Figure 1.

Flow diagram of literature search and study selection for Cure Violence systematic review.

Initial abstract screening identified 92 studies for full-text review. Three of these could not be retrieved, either online or through attempts to contact the authors. A preliminary review of the remaining papers excluded 36 that did not present findings on program effects related to shootings or killings.

To assess whether studies evaluated an intervention consistent with the Cure Violence model, we relied on evaluators’ descriptions provided in the studies. Six papers were excluded on this basis, as they either did not implement the Cure Violence model or omitted core elements of the approach. With 4 papers, the program examined was not intended to be a Cure Violence approach, and their exclusion is straightforward.^22-25

With the remaining 2 studies, the program implementers cited the Cure Violence approach as an influence, but the evaluators noted substantial deviations from the basic elements of the approach. For example, Pittsburgh’s One Vision One Life program claimed to have implemented the Cure Violence approach, yet evaluators noted that key elements were omitted. Notably, the approach did not employ credible messengers in their program and focused on lower-risk individuals and situations, a total departure from the Cure Violence approach. ¹²  Similarly, Phoenix’s Truce project did not incorporate individualized risk-reduction plans or community mobilization efforts, both of which are essential to the Cure Violence approach as well. ²⁶

An additional 32 studies were excluded due to issues with the analysis. In 4 cases, Cure Violence was implemented alongside other interventions, but the effects of individual components were not disentangled.^27-30Ten studies were excluded for flawed analytical methods, including 5 cases where outcomes were analyzed for geographic areas that extended substantially beyond the program’s operational scope,^31-35 4 papers used problematic baseline data,^36-39 and 1 study had an incorrect start date. ⁴⁰

Nine studies presented outcome data on shootings and killings, such as a line chart or rates of shootings and killings but did not clearly indicate changes resulting from program implementation.^41-49 Finally, 9 studies reported the effects on shootings or killings but lacked control areas in the analysis.^15,50-57 Each study selected was conducted by researchers independent from the program, minimizing risk of bias in the studies. Data from the studies were collected independently by the lead author.

Evidence on the Effectiveness of the Cure Violence Approach

In total, 13 studies met all inclusion criteria, providing analyses of the implementation of the Cure Violence approach along with quantitative measures of its impact on shootings or killings^16,58-69 (Table 1). These studies collectively examined 27 program sites, plus 4 studies conducted multi-site analyses.^60,64,67,69 Additionally, some sites in Baltimore were assessed multiple times. Across these studies, a total of 47 site-level analyses produced 83 total findings, 43 findings on shootings and 40 findings on killings.

Table 1.

Summary of Findings From Evaluations by Country, City, Community Area.

Country	City	Community area	Killings	Sig.	Shootings	Sig.	Year of publication
Colombia	Cali	Charco Azul	−74.0%	*	—		2020
		Comuneros	−32.0%	*	—		2020
Trinidad	Port of Spain	Multiple areas	—		−38.7%	*	2018
United States	Charlotte	Beatties Ford	−76.2%		—		2023
United States	Chicago	Auburn Gresham	−52.0%	*	−42.0%	*	2009
		Englewood	41.0%		−27.0%		2009
		Logan Square	−16.0%		−40.0%	*	2009
		Rogers Park	−74.0%		−66.0%		2009
		Southwest	−92.0%		−35.0%	*	2009
		West Garfield Park	−43.0%		−42.0%	*	2009
		West Humboldt Park	−14.0%		−42.0%		2009
United States	New York City	Crown Heights	—		−6.0%	*	2013
		East New York	—		−50.0%	*	2017
		South Bronx	—		−37.0%	*	2017
		Multiple sites	−18.0%	*	—		2015
		Multiple sites	—		−11.6%		2025
United States	Philadelphia	North Philadelphia West	—		−30.0%	*	2017
United States	Baltimore	Bellair-Edison	103.0%		−21.0%		2023
		Belvedere	−40.0%		459.0%	+	2023
		Brooklyn	27.0%		85.0%		2023
		Cherry Hill	−56.0%	*	−34.0%	*	2012
			−28.8%		−4.7%		2018
			−21.0%		10.7%		2022
			−39.0%	*	−18.0%		2023
		Ellwood Park	—		−34.0%	*	2012
			95.6%		−3.6%		2018
			112.2%	+	13.4%		2022
		Franklin Square	−18.0%		−84.0%	*	2023
		Lower Park Heights	−15.8%		−19.1%		2018
			−21.1%		−19.3%		2022
			−48.0%	*	−29.0%	*	2023
		Madison-Eastend	170.0%		−44.0%	*	2012
			78.6%		96.2%		2018
			69.0%		152.6%		2022
		McElderry Park	−26.0%	*	22.0%		2012
			−36.3%	*	19.0%		2018
			−30.8%		25.6%		2022
			−28.0%	*	2.0%		2023
		Mondawmin	75.0%		40.8%		2018
			75.9%		27.0%		2022
			−24.0%		−11.0%		2023
		Penn North	−53.0%	*	−51.0%	*	2023
		Sandtown-Winchester	−9.9%		12.5%		2018
			−8.6%		15.4%		2022
			44.0%		−53.0%		2023
		Woodbourne-McCabe	−15.0%		−47.0%	*	2023
		Multiple sites	10.0%		−23.0%	*	2023

Note. Criteria for inclusion: findings on shootings or killings, findings on significance, and use of controls.

^*Significant findings on reductions in shootings or killings.

⁺Significant increase in shootings or killings.

Of these findings, 68.7% (n = 57) indicated a reduction, including 67.4% (n = 29) of shootings and 70% (n = 28) of killings. Overall, 32.5% (n = 27) of findings demonstrated statistically significant decreases, including 39.5% (n = 17) of shootings and 27.5% (n = 11) of killings. Findings varied geographically, with Baltimore accounting for the highest proportion of sites reporting either an increase in shootings or killings or non-significant results. Outside of Baltimore, 95.8% of sites demonstrated reductions in violence, with 54.2% achieving statistical significance. By contrast, in Baltimore, 57.6% of findings showed reductions in shootings or killings, with only 23.7% reaching statistical significance.

When examining shootings specifically, excluding Baltimore, all other Cure Violence sites reported reductions, with 73.3% demonstrating statistical significance. Among the 4 sites that recorded reductions without reaching significance, 3 were from the 2009 evaluation of the Chicago program. Regarding killings, 84.6% of sites, excluding Baltimore, reported decreases, though only 30.8% achieved statistical significance. Given the relatively low frequency of homicides at the community level, detecting statistically significant reductions in killings presents a methodological challenge.

In Baltimore, 4 studies assessed the impact of the program across 13 implementation areas, including one that examined citywide effects.^16,63,66,67 These studies generated 59 findings on shootings and killings, with 23.7% reaching statistical significance. Notably, one study analyzed the program’s effects during and after the civil unrest that followed the death of Freddie Gray, an African American man who died while in police custody, sparking protests over police brutality. Two other studies examined impact during the COVID-19 pandemic, which was a period of heightened community tension and structural instability.

Beyond the U.S., evaluations in Latin America have also demonstrated the effectiveness of the Cure Violence approach. In Port of Spain, Trinidad and Tobago, an evaluation of the Project Reason program in the Laventille community reported a statistically significant 38.7% reduction in shootings. ⁶⁴  Similarly, in Cali, Colombia, two program sites experienced statistically significant reductions in killings of 74% and 32%, respectively. ⁶⁵  These findings are particularly notable given the unique challenges of implementing and evaluating violence prevention programs in this region.

In addition to the statistically rigorous studies, 9 additional studies or reports documented reductions in shootings and killings but did not include control communities for the analysis (see Table 2).^15,50-57 These studies analyzed 13 individual program sites and 4 multi-site interventions, yielding a total of 28 findings. Among these findings, 85.7% indicated a decrease in violence. The SNUG (guns spelled backwards) program sites in New York State accounted for all of the instances of an increase. ⁵⁰

Table 2.

Summary of Findings for Cure Violence Studies Without Controls.

Country	City	Community area	Killings	Shootings	Year of publication
Mexico	Culiacan	Multiple areas	−84.2%	—	2023
South Africa	Cape Town	Hanover Park	−47.4%	−78.3%	2013
			−37.5%	−32.1%	2014
United States	Albany	Multiple areas	−46.3%	−18.0%	2013
	Annapolis	Eastport	−80.0%	−100.0%	2024
	Charlotte	Beatties Ford	−75.0%	−4.5%	2023
			—	−52.0%	2024
	Chicago	Auburn Gresham	—	−80.0%	2010
		Logan Square	—	−84.0%	2010
		Rogers Park	—	−62.0%	2010
		Southwest	—	−78.0%	2010
		West Garfield 1	—	−82.0%	2010
		West Garfield 2	—	−40.0%	2010
		West Humboldt Park	—	−41.0%	2010
	New York City	East New York	38.9%	1.0%	2013
		Harlem	14.0%	−23.5%	2013
	Niagra Falls	Highland	−39.5%	3.5%	2013
	St. Louis	Multiple areas	−43.0%	—	2022
	Yonkers	Multiple areas	−47.4%	−53.9%	2013

Overall, the breadth of evidence from multiple locations and methodological approaches suggests that the Cure Violence model consistently contributes to reductions in violence, particularly in shootings. While findings in Baltimore were more inconsistent, results across other sites, including those in international contexts, underscore the model’s adaptability, and effectiveness in diverse settings. These outcomes provide valuable insights for policymakers, practitioners, and researchers seeking to implement evidence-based violence prevention strategies.

Understanding the Baltimore Findings

Baltimore was the first city outside Illinois to implement the Cure Violence approach, with CVG leadership playing a central role in its establishment. From the outset, an evaluation team was engaged to assess both implementation fidelity and program impact across 4 sites. This rigorous study found statistically significant reductions in violence across all sites, with 3 locations reporting declines in shootings and 2 recording a reduction in killings. ¹⁶ Beyond reductions in violence, the evaluation also highlighted improvements in community norms and participant outcomes: 88% of participants secured employment, 95% enrolled in a school or GED program, and 100% received assistance in resolving family conflicts. Overall, 80% of participants reported experiencing positive life changes since joining Safe Streets.

However, the program’s effectiveness varied by site, offering critical insights into conditions necessary for successful implementation. Cherry Hill, a geographically distinct community in South Baltimore, demonstrated strong implementation and substantial impact, with a statistically significant 53% reduction in killings and a 34% decrease in shootings. In contrast, the 3 East Baltimore sites – McElderry Park, Elwood Park, and Madison-Eastend – faced greater implementation challenges. Elwood Park and Madison Eastend both experienced significant reductions in shootings as well as increases in killings, highlighting key factors that influenced program outcomes.

In McElderry Park, the first site established in East Baltimore, there was initially a 53% reduction in killings, including a 22-month period without a homicide. However, when funding constraints forced the program to expand without providing adequate resources for the expansion, its impact declined. New sites in Elwood Park and Madison-Eastend were introduced, but rather than receiving funding for separate infrastructure and leadership, they were forced to share an office, site director, and outreach supervisor. As a result, resources originally designed for one site were stretched across 3 locations, and neither new site was able to establish an office within its own geographic target area. Following this expansion, McElderry Park experienced an increase in killings, demonstrating that a dilution of resources weakens program effectiveness, as might be expected.

Additionally, the Madison-Eastend site launched amid an ongoing and escalating group conflict that predated the program and flared up just as the program started. Within 6 months, this conflict escalated into a mass shooting that left 12 people injured. Given the program’s limited resources and the scale of violence in the area, staff faced significant challenges in intervening at a pace necessary to catch up with and de-escalate tensions.

The evaluation found a direct correlation between the number of violent interruptions conducted and reductions in killings, reinforcing the critical role of violence interruption as a key mechanism for reducing violence. While the study did not specifically assess hiring practices or training quality, these findings suggest that ensuring properly trained and well-equipped violence interrupters is essential to maximizing the impact of the model. In some cases, programs that have claimed to implement the approach have not adhered fully to these essential components, potentially affecting their outcomes.

Another key factor influencing program outcomes was the presence or absence of infrastructurally strong community-based organizations. Cherry Hill benefited from robust institutional support from local organizations, which played a crucial role in program implementation. In contrast, the East Baltimore sites lacked comparable infrastructure, limiting their ability to sustain violence prevention efforts. These findings underscore the critical role of community partnerships in ensuring program success and the need to strengthen these organizations when resources permit.

Since the 2012 evaluation, 3 studies have assessed the Safe Streets program, each examining the approach during periods with significant external factors affecting violence, notably the civil unrest following the killing of Freddie Gray and the COVID-19 pandemic.^63,66,67 Many of these evaluations documented a program operating under severe funding constraints, with some sites lacking outreach workers entirely or relying on a minimal number of violence interrupters. These variable conditions in Baltimore illustrate the challenges of implementing the Cure Violence approach without full fidelity to the approach, reinforcing the necessity of consistent investment, staffing, and structural support to achieve sustainable reductions in violence.

Discussion

For the past 25 years, CVG has been at the forefront of developing, refining, and expanding a public health approach to violence prevention. Through rigorous research and extensive fieldwork, CVG has built a strong evidence base demonstrating the effectiveness of its model. This work has not only validated the Cure Violence approach but has also seeded the broader community violence intervention (CVI) field, which now includes numerous applications employing outreach workers, violence interrupters, and the basic tenets.

The results from Baltimore offer an essential lesson in the importance of analyzing, understanding, and reporting on the context in which a program is implemented and its impact on the levels of violence. In medical research, new treatments and interventions are often tested under lab or other more controlled conditions, allowing researchers to isolate their effects better. In contrast, violence prevention programs, including those using the Cure Violence approach, are implemented in highly complex environments, often in under-resourced communities where external factors shape program outcomes. Unlike more fixed or stable conditions, real-world implementation is subject to systemic challenges such as political opposition, inconsistent funding, lack of community infrastructure, and broader social conditions, all of which can affect program effectiveness. And approaches often require modification to increase their effectiveness, which research can help to identify.

Too often, discussions of violence prevention effectiveness focus on whether an approach “worked” or “did not work,” without accounting for key contextual and implementation factors. No evidence-based intervention, whether in public health, education, or violence prevention, has demonstrated universal success in every implementation. Variability in outcomes is not necessarily an indication of a flawed model; it may instead reflect differences in fidelity, external conditions, and available resources. Even vaccines, which may have a high efficacy rate in controlled trials, face barriers in real-world application due to vaccine distribution challenges, public perception, or inconsistent dosage. Likewise, violence prevention programs require sufficient dosage, accurate and proper implementation, monitoring, and structural support to achieve their intended impact. In other words, the impacts are dependent on implementation and contextual matters, even when the approach has been highly effective in other locations.

When evaluating outcomes, researchers could also ask deeper questions about implementation and context: Was the model implemented as designed? Were the necessary resources in place? What external factors influenced the program’s effectiveness? Understanding and reporting on these elements is critical to understanding results, and even programs with demonstrated success benefit from continuous assessment and refinement for their next phase of work. Identifying areas for adjustment or improvement, such as increasing funding, expanding staffing, strengthening political support, and enhancing wraparound services, can be very helpful to the field as well as to the community adapting the approach to its particular circumstances.

Research and evaluation should help contribute to the growing body of evidence on what works, where, and why. The goal is to continuously refine and strengthen interventions based on real-world implementation experiences, ensuring that the field of violence prevention remains data-driven, adaptable, and responsive to community needs.

Evaluations could therefore focus on identifying the key local factors that influence implementation success or challenges and ensuring that assessments account for both enabling conditions and potential barriers. A more precise analysis of these real-world complexities can provide actionable guidance to local organizations on what may be missing in implementation while also contributing valuable insights to the broader field of violence prevention.

Community Context Factors

External conditions within the target community significantly impact program effectiveness. Key contextual factors include:

• Availability of Resources for High-Risk Individuals – The presence of employment programs, education services, and mental health support can enhance the ability of violence prevention programs to facilitate long-term behavior change.

• Economic Conditions – High levels of poverty and unemployment increase the level of need in a community, requiring greater resources to sustain violence reduction efforts.

• History of Violence – Long-standing feuds or deeply entrenched cycles of retaliatory violence may require extended intervention periods before measurable declines in violence occur.

• Broader Social Disruptions – Natural disasters, pandemics, civil unrest, and major political events can introduce new stressors, affecting both community violence levels and program operations.

Programs must be assessed within these broader structural realities, recognizing that the success of an intervention is not determined solely by its internal components, but also by the external ecosystem in which it operates.

Implementation Factors

Effective violence prevention efforts require more than the existence of a model; successful implementation depends on stakeholder support, organizational capacity, and sustained investment.

• Stakeholder Engagement – Programs are most effective when political leaders, law enforcement, faith-based organizations, and community groups understand and support the approach. Opposition or disengagement from key stakeholders has, in several cases, undermined program success.

• Organizational Capacity – The agency responsible for implementing the model must have the ability to engage high-risk individuals, manage financial resources, and provide adequate support for staff. If an implementing agency lacks these capacities, program fidelity and impact can suffer.

• Adequate & Sustained Funding – Funding must be sufficient to support staffing levels, outreach efforts, and violence interruption activities. Short-term or inconsistent funding cycles can disrupt program continuity and weaken effectiveness.

CVG employs a structured implementation process that includes: selecting a target area, issuing an RFP (Request for Proposals), identifying a qualified implementing agency, conducting structured hiring panels, and building partnerships with hospitals, social service agencies, and justice system actors. Additionally, CVG provides ongoing training and technical assistance to ensure program fidelity and adaptability to local conditions.

Determinants of Success

Research on the Cure Violence model has identified key determinants of success that help explain why certain sites achieve significant reductions in violence while others encounter challenges. These include implementation fidelity, stakeholder engagement, funding stability, and broader environmental conditions.

One of the most crucial factors in program effectiveness is fidelity to the model. The Cure Violence approach is built around specific components that must be implemented in full to achieve the desired outcomes. At a minimum, programs must:

• Employ highly credible staff who have deep trust and access within the community.

• Focus on individuals at the highest risk of violence, ensuring that interventions target those most likely to be involved in violent conflicts.

• Interrupt high-risk conflicts through direct intervention and mediation.

• Provide ongoing case management and behavioral support for participants.

• Provide mental health and trauma-related supports for program staff and participants.

• Respond immediately to shootings and emerging violent situations.

• Disseminate public education messages that shift community norms around violence.

• Monitor and assess program progress using structured data collection and feedback systems.

When these components are not in place, whether due to underfunding, staffing shortages, or a lack of local political support, the intervention’s ability to produce significant reductions in violence is naturally compromised. Evaluating a program without first examining whether it was implemented correctly can lead to misleading conclusions about its effectiveness.

The broader community context in which the program operates can have a significant impact on effectiveness. External factors such as poverty, unemployment, long-standing group conflicts, and systemic disinvestment can influence violence levels, sometimes delaying observable reductions even when an intervention is functioning as designed. Large-scale events, including civil unrest, pandemics, and political instability, can further complicate evaluation efforts by introducing new variables that impact both violence trends and program operations. Effective research must account for these structural realities.

The responsibility for the success of these programs does not rest solely on the Cure Violence model or the organizations implementing it. Local leadership must play an active role in sustaining these approaches by:

• Championing the public health framework citywide

• Providing adequate and sustained funding

• Ensuring access to essential wraparound services

• Investing in ongoing training and technical assistance

• Monitoring and deploying timely responses

Without this level of commitment, the full potential of these interventions will remain unrealized.

The Expanded Role of Research: A Collaborative Effort

Violence prevention research is at a pivotal moment. The field requires a more nuanced understanding of the factors that help to shape program success. This requires not only examining whether an intervention reduces violence at a site at a point in time but also analyzing the conditions that enable or inhibit its effectiveness and identifying what is needed for more effective interventions. By maintaining a commitment to rigorous, context-aware evaluation, researchers can ensure that violence prevention strategies continue to evolve and improve based on real-world evidence.

A well-informed research agenda should prioritize methodological rigor, account for external influences, and refine evaluation standards to align with the public health paradigm. By embracing scientifically grounded, context-sensitive approaches, researchers can provide the insights necessary to strengthen policy, improve program design, and ultimately save lives on a global scale.

The research community has a vital role to play in ensuring the continued development and success of violence prevention programs. By refining methodologies, identifying best practices, and tailoring research to meet the unique needs of each community, researchers can help build a more robust, evidence-driven foundation for community violence prevention. Their work is critical to ensuring that public health-based, community-driven approaches to violence prevention remain viable, effective, and central to efforts to build safer, healthier communities.

Even stronger collaboration between leading research institutions, implementing organizations, and practitioners can fundamentally shift the approach to evaluating interventions with an established evidence base. When programs have been successfully replicated across diverse contexts, cities, and countries over time, the focus of evaluation should move beyond the binary question of “Does this work?” to a more context-specific inquiry: “Is this working here so far?,” “What factors influence success or challenges?,” “What modifications are needed to optimize effectiveness at this site further?,” and “What is needed for the next phase of work?”

By applying this iterative and context-driven evaluation framework, researchers and practitioners can work collaboratively to identify local barriers, refine implementation strategies, and strengthen intervention outcomes, ultimately ensuring that proven models achieve their full potential in every setting.

As the field of violence prevention continues to evolve, policymakers, researchers, and community leaders can work together to ensure that violence prevention efforts receive the necessary resources, infrastructure, and commitment to drive even more meaningful and substantial change. A fully realized and supported evidence-based approach to community violence intervention will not only save lives but also strengthen our communities, promote healing, and advance public health and safety on a global scale.

Limitations

The present study reviewed literature on the Cure Violence approach. This paper presents a limited view of the field of community violence intervention, as there are now a multitude of approaches, and this study only looks at one. Future research is needed to explore the other approaches and the factors that make each approach effective as well as ineffective.

Additionally, the present study was conducted by staff of Cure Violence Global. While this presents a bias to the analysis, we sought to limit any potential for bias by thoroughly listing all excluded studies with explanations and reporting on the findings of the reports without additional analysis. While this mitigates some potential bias, the potential for bias still exists in the framing of the findings in this paper.

Conclusion

The effectiveness of the Cure Violence approach in reducing shootings and killings is well-documented, with multiple evaluations across multiple communities, cities, and countries over the past 20 years demonstrating significant reductions in violence when the model is implemented with fidelity. This systematic review found 13 evaluations covering 27 sites and producing 83 findings on the impact of the Cure Violence approach on shootings or killings. Although studies in Baltimore found varied results, findings for other cities found predominantly large reductions in shootings and killings. These findings indicate that the Cure Violence approach is an effective strategy when implemented with fidelity and the necessary funding and support.

The results from Baltimore offer additional insights into factors that may affect program success. Key factors include a focus on people at the highest risk, use of highly credible staff, sufficient and sustained funding, ongoing data monitoring, strong management and supervision, timely responses to conflict, and structured training and technical assistance.