Costs of Implementing Teen Dating Violence and Youth Violence Prevention Strategies: Evidence From 5 CDC-Funded Local Health Departments

Abstract

Objectives:

In 2016, the Centers for Disease Control and Prevention supported 5 local health departments (LHDs) to implement teen dating violence and youth violence primary prevention strategies across multiple levels of the social–ecological model and build capacity for the expansion of such prevention efforts at the local level. The objective of this study was to estimate the total cost of implementing primary prevention strategies for all LHDs across 3 years of program implementation.

Methods:

We used a microcosting analytic approach to identify resources and compute costs for all prevention strategies implemented by LHDs. We computed the total program cost, total and average cost per strategy by social–ecological model level, and average cost of implementation per participant served by the program. All costs were inflated via the monthly Consumer Price Index and reported in August 2020 dollars.

Results:

For 3 years of program implementation, the total estimated cost of implementing teen dating violence and youth violence primary prevention strategies was >$7.1 million across all 5 LHDs. The largest shares of program-related costs were program staff (55.9%-57.0%) and contracts (22.4%-25.5%). Among prevention strategies, the largest share of total costs was for strategies implemented at the community level of the social–ecological model (42.8%).

Conclusions:

The findings from this analysis provide a first look at the total costs of implementing comprehensive teen dating violence and youth violence primary prevention strategies and serve as a foundation for investments in local violence prevention funding for young people.

Understanding and preventing youth violence requires coordinated efforts between government and community-based organizations that integrate strategies and approaches for violence prevention.^1-3 Each day in the United States, >1000 adolescents and young adults are treated in emergency departments for physical assault–related injuries. ⁴ Physical assault accounts for 3.2% of all nonfatal emergency department visits among children and adolescents aged 10-14 years and 6.2% among adolescents and young adults aged 15-24 years. ⁴ Among fatal injuries, homicide is the third-leading cause of death among people aged 10-24 years. ⁴ In addition, in 2019 approximately 8.2% of US high school students reported experiencing sexual dating violence.^5,6 Youth violence is a substantial public health problem that affects not only young people but also their families, schools, communities, and society as a whole. The combined economic burden from fatal and nonfatal assault-related injuries among people aged 10-24 years is >$100 billion annually; this estimate is based on medical care, loss of work, value of statistical life (a monetary estimate of the collective value placed on mortality risk reduction), and loss of quality of life (the estimated monetary value of the quality of life impact from injury or illness).^7,8

Local health departments (LHDs) are well positioned to support prevention efforts because of their relationships with the communities they serve, partnerships across multiple sectors, and ability to engage in schools and other venues that directly serve young people.^2,9,10 In September 2016, the Centers for Disease Control and Prevention (CDC) issued a funding opportunity entitled “Preventing Teen Dating and Youth Violence by Addressing Shared Risk and Protective Factors” (hereinafter, the 1605 program). ¹¹ The 1605 program was designed to support LHDs’ capacity and readiness to prevent teen dating violence and youth violence and to expand existing prevention efforts at the local level. LHDs that received funds were tasked with implementing prevention strategies that addressed shared risk and protective factors across multiple levels of the social–ecological model: a framework for prevention that (1) identifies risk factors for violence at the individual, relationship, community, and societal levels and (2) suggests that, to prevent violence, it is necessary to implement prevention strategies across multiple levels of the model. ¹² LHDs funded under the 1605 program implemented strategies at the individual or relationship level and at the community level (Table). LHDs were also encouraged to leverage resources beyond CDC funding and expand their existing violence prevention strategic plan by collaborating with a citywide multisector coalition. ¹¹ Five LHDs were funded through this effort.

Table.

Teen dating violence and youth violence prevention strategies implemented by LHDs funded by the 1605 program, United States, 2016-2019 ^a

Social–ecological model level
Strategy	Description	No. of LHDs implementing the strategy	Setting and focus of strategy	No. of participants reached by strategy
Individual and relationship level
Safe Dates 13	A school-based curriculum program that educates young people on how to identify and prevent teen dating violence	2	(1) Setting: public high schools Focus: 9th-grade cohorts for in-class training and 9th- to 12th-grade students in Safe Dates–focused club at each participating high school (2) Setting: public high schools Focus: 9th- to 12th-grade students, curriculum taught in class at each participating high school	5572
Fourth R 14	A skills-based program that teaches students how to promote healthy relationships and reduce conflict and risk behavior	1	Setting: public high schools Focus: curriculum implemented in health class in each participating high school	1043
Strengthening Families 15	A parenting skills and family relationship program to enhance family support and connectedness	1	Setting: family resource centers, local churches, and juvenile hall Focus: families ordered by court to participate and self-referred families	120
Coaching Boys Into Men 16	A program that engages boys through athletics and trains coaches to model and promote healthy relationships with their athletes	2	(1) Setting: public middle and high schools, community-based team sport leagues Focus: young people in team sports in grades 6-12 and young adults aged 18-22 years in team sports with participating community-based sport leagues (2) Setting: basketball leagues, public high schools Focus: young people in grades 9-12 participating in high school sports or community-based team sport leagues	884
Community level
Crime Prevention Through Environmental Design (CPTED) 17	An environmental design approach to promote safe use of spaces, reduce opportunities for disputes and violence, and promote positive behavior	3	(1) Setting: public high schools, community multiservice centers, and public middle schools in targeted zip codes Focus: young people aged 10-24 years residing in targeted zip codes, provided with CPTED advocacy toolkits and education (2) Setting: local churches and community organizations Focus: 9th- to 12th-grade students served by participating churches and community organizations who receive CPTED training (3) Setting: targeted communities in identified zip codes Focus: young people aged 10-24 years residing in targeted communities	9673
Safe Streets 18	A community-based program that uses trained outreach workers to mediate and de-escalate conflict. Outreach workers also link young people to support services	1	Setting: targeted communities within city school district Focus: students at local high schools in targeted communities who are identified for mediation and provided with Safe Streets training after mediation	233
Street and community outreach 19	Norm change approaches connect trained adults to mediate conflict, improve safe environments, and create youth-friendly spaces that strengthen social connections	1	Setting: targeted communities in 18 census tracts; strategies implemented within each community at public schools, neighborhood events, community libraries, YMCA, and Boys and Girls Clubs Focus: young people aged 10-24 years in targeted communities	18 290

Abbreviation: LHD, local health department.

^aPrevention strategies were funded by the Centers for Disease Control and Prevention’s funding opportunity “Preventing Teen Dating and Youth Violence by Addressing Shared Risk and Protective Factors” (aka 1605 program) during program years 1 to 3. ¹¹

The objective of this study was to evaluate the total cost of all resources used by LHDs to implement the 1605 program, including resources supported by CDC and those leveraged from other sources, to provide CDC and other partners with a total cost of program implementation and a foundation for budget justification, decision-making, and forecasting for investments in teen dating violence and youth violence prevention funding at the local level.

Methods

This study used a retrospective microcosting analytic approach to estimate the total costs associated with implementing strategies required of LHDs under the 1605 program funding announcement. Microcosting allows for an assessment of the economic costs of public health interventions and involves collecting detailed data on resources used for each activity and applying unit costs for those resources. ²⁰ This is a bottom-up approach that allows for precise estimates of activities that may not have accounting costs associated with them, such as in-kind contributions. ²¹ This approach has been found to enhance the usefulness of economic evaluations by providing a complete economic picture. Using a program cost perspective, we held an initial recipient meeting with each LHD in August 2019 to engage its 1605 program team in discussions about the program, and we reviewed each LHD’s annual performance reports from 2016 to 2019 to identify program resources. Given this information as well as information on standard resources from a previous cost study conducted for a CDC-funded program, ²² we identified 6 broad program cost categories to organize data collection: program staff, consultants, contracts, supplies, transportation, and administrative costs. We also identified contributions allocated to programs during each program year—both labor and nonlabor in kind (ie, volunteered or donated). We identified all resources used for program implementation, calculated quantities for all resources, and applied unit costs to calculate total costs. We evaluated all resources for the prevention strategies implemented by each LHD during 3 years of program implementation.

Data Collection

We developed a spreadsheet-based cost collection tool (eFigure 1 in Supplemental Material) to collect information from LHDs on each implemented 1605 program strategy. The tool was designed to ensure collection of standard cost elements despite differences in implementation among LHDs. Using the tool, LHDs reported on (1) all sources of funding received for 1605 program implementation and (2) the established cost categories and all in-kind labor and nonlabor contributions received for implementation of 1605 program strategies for each program year. To appropriately allocate costs, we collected details on the distribution of these costs among each strategy implemented by the LHD, including all strategies in the social–ecological model framework, and activities that supported collaborative efforts with a multisector coalition. ¹⁰ Each LHD’s 1605 program director completed the tool with support from an administrative or financial officer, reporting resources and costs from September 1, 2016, through August 31, 2019. LHDs were provided with a user guide (eFigure 2 in Supplemental Material) that had detailed definitions of each cost category.

This guide assisted with data reporting and helped to ensure consistent reporting among all 1605 programs. To reduce errors and ensure high-quality data, 2 of the 5 programs pilot-tested the cost collection tool and provided feedback, which allowed for edits to be applied to the final instrument. Collection of data was retrospective; however, several features were implemented to ensure that data collection methods were standardized across all LHDs, including webinar trainings and ongoing technical assistance. All data were reviewed and verified by 2 authors (K.L.R, G.F.M.), who served as the lead economists for the study.

Cost Estimation

We estimated costs for labor by using annual salaries and the percentage of total full-time equivalent employee time allocated to 1605 program strategies. For in-kind labor, we estimated the total value of labor volunteered using the Bureau of Labor Statistics’ Metropolitan and Nonmetropolitan Area Occupational Employment and Wage Estimates. ²³ Each wage estimate was adjusted for the annual average fringe benefit rate (for the given year) for state and local government employment. ²³ We aggregated the in-kind labor costs for each strategy by multiplying the estimated hourly rate by the number of annual hours performed per strategy. For nonlabor in-kind resources, we estimated the value of each resource using published market rates. For example, donated material items were valued at the average price in the market for each item, and donated event space was valued at the fair market rental rate by venue size for the metropolitan area in which the LHD was located. For all other nonlabor resources, we applied unit costs to each resource to generate estimates of the total cost for each program year.

Data Analysis

Costs were aggregated and analyzed for each program cost category and each prevention strategy across all LHDs for the period indicated. Strategies were categorized by strategy level of the social–ecological model (ie, individual/relationship level vs community level), as we identified that individual/relationship-level strategies were similar in implementation across all LHD sites; likewise, community-level strategies were similar in their implementation across all LHD sites. We estimated the total program cost, total and average cost per strategy by level of the social–ecological model, and average cost per participant served. The number of participants served annually by the program was provided in annual progress reports submitted by each LHD to CDC. All costs were inflated via the monthly Consumer Price Index ²⁴ and reported in August 2020 dollars. We did not conduct a sensitivity analysis. The cost study protocol was reviewed through CDC’s institutional review board process to determine if human subjects research applied, and the study was determined to be exempt from human subjects review because it was a nonresearch cost evaluation.

Results

Total funding and in-kind contributions received varied per LHD, and other sources of funding varied widely (Figure 1). One LHD reported no additional sources of funding received, and another LHD identified 36.5% of its total funding coming from other sources across the 3 years of the study. Other funding sources included additional federal funding, state grants, city and county funds, and private grants. In-kind contributions varied widely. One LHD received no in-kind contributions, while another LHD cited 12.1% of its resources contributed as in kind across the 3 years of the study. Volunteer labor was the primary component of in-kind contributions.

Figure 1.

Total allocated funding and in-kind contributions reported per local health department as part of the 1605 program, United States, 2016-2019. The percentages reported for “funding from other sources” and “in-kind contributions” were unweighted averages of each value during the 3 years of reported funding. Prevention strategies were funded by the Centers for Disease Control and Prevention’s funding opportunity “Preventing Teen Dating and Youth Violence by Addressing Shared Risk and Protective Factors” (aka 1605 program) during program years 1 to 3. ¹¹

The total cost of resources used to implement the 1605 program for 3 years of implementation was estimated at >$7.1 million (Figure 2). For all LHDs in aggregate, the largest program cost categories across the 3 years were program staff (55.9%-57.0% of total program costs) and contracts (22.4%-25.5%). Contracts awarded to entities varied from community and youth-serving organizations tasked with strategy implementation and outreach to local universities tasked with monitoring and evaluating activities. Of the overall program costs across the 3 years, supplies constituted 7.6% to 10.0% and administrative overhead 8.0% to 12.4%. Costs for consultants and transportation were <1.0% across the 3 years.

Figure 2.

Total aggregated costs per year and percentage of costs per year, by program category, as part of the 1605 program, United States, 2016-2019. Program costs exclude valuation of in-kind contributions. Consultant and transportation costs were combined because they accounted for <1.0% of total costs. Prevention strategies were funded by the Centers for Disease Control and Prevention’s funding opportunity “Preventing Teen Dating and Youth Violence by Addressing Shared Risk and Protective Factors” (aka 1605 program) during program years 1 to 3. ¹¹

Among strategies and activities, total costs for implementing community-level strategies were estimated at >$3.0 million (42.8%), followed by individual/relationship-level strategies at nearly $2.5 million (35.5%), and activities that supported multisector coalitions at >$1.5 million (21.7%). Overall, strategies implemented at the community level were more costly to implement than those at the individual and relationship level (Figure 3). The average cost per participant served decreased annually, with individual/relationship-level strategies costing $699 per participant served in year 1 and $239 per participant served in year 3 (Figure 4). As compared with costs for individual/relationship-level strategies, costs for community-level strategies were substantially higher per participant served in year 1; however, the decrease over time was much larger. In year 1, the average cost was $3673 per participant served, and by year 3 the average cost was $58 per participant served. Overall, annual average costs per participant decreased as the number of participants served increased.

Figure 3.

Total cost of implementation per year, stratified by social–ecological model level, as part of the 1605 program, United States, 2016-2019. The total cost at the strategy level includes valuation of in-kind contributions and excludes costs for activities supporting multisector coalitions. Prevention strategies were funded by the Centers for Disease Control and Prevention’s funding opportunity “Preventing Teen Dating and Youth Violence by Addressing Shared Risk and Protective Factors” (aka 1605 program) during program years 1 to 3. ¹¹

Figure 4.

Average cost per participant and total participants served per year by the 1605 program, stratified by social–ecological model level, as part of the 1605 program, United States, 2016-2019. Average costs at the strategy level include valuation of in-kind contributions and exclude costs for activities supporting multisector coalitions. Prevention strategies were funded by the Centers for Disease Control and Prevention’s funding opportunity “Preventing Teen Dating and Youth Violence by Addressing Shared Risk and Protective Factors” (aka 1605 program) during program years 1 to 3. ¹¹

Discussion

Preventing teen dating violence and youth violence requires comprehensive strategies that address shared risk and protective factors across multiple levels of the social–ecological model. LHDs working to implement such strategies often have limited resources and capacity. The 1605 program was designed to support LHDs and leverage their existing capacity to implement strategies in their communities. The focus of the program is on preventing violence before it occurs and addressing multiple needs in a community, while considering the context of the neighborhood and larger environment where young people live. Our study estimated the total cost to implement the 1605 program by measuring all resources used for program implementation (resources funded by CDC, leveraged from other funding sources, and provided as in-kind contributions).

While all 5 sites in this study received 1605 program funding, each site may have had additional leveraged funds that are used in these programs. The amount of funding from additional sources differed by LHD and demonstrated the importance of determining all lines of funding when pursing cost research. For example, 1 LHD reported receiving no additional funding, while another LHD received nearly half of its total funding from non-CDC sources. Additionally, 1 LHD had no in-kind contributions, while other LHDs did receive in-kind support. Some of these differences may be due to the benefit of some LHDs having received previous CDC funding (ie, STRYVE [Striving to Reduce Youth Violence Everywhere] and Dating Matters) to support youth violence and teen dating violence prevention activities. Some relationships may have already been formed from the STRYVE and Dating Matters initiatives that facilitated access to additional funding and resources.

Although we found variation in other funding sources, the distribution of annual costs across program categories was comparable across LHDs, with program staff and contracts constituting the largest share. By examining program costs annually, we were able to show where resources were being allocated and whether any changes in allocation were observed during the 1605 program funding cycle. Our study also determined that the largest program costs for implementing teen dating violence and youth violence prevention strategies were consistent with a previous cost analysis of the Dating Matters teen dating violence prevention model, which identified staffing as the largest percentage of program costs, at 50% of the total cost of implementation. ²⁵ While program staff and contracts were the costliest components, investing in staff and establishing contracts with community and youth-serving organizations and educational institutions were essential for addressing the multiple risk factors for youth violence prevention and ensuring LHD program sustainability. The findings showed that strategies implemented at the community level of the social–ecological model were more costly to implement than individual/relationship-level strategies. This difference in cost may be attributed to the community-level strategies being broader and taking more time to implement than the individual/relationship-level strategies. Strategies such as Crime Prevention Through Environmental Design, ²⁶ Safe Streets, ²⁷ and Street and Community Outreach ²⁸ are multifaceted community capacity–building efforts that rely on partnerships with key community organizations. While these community-level strategies were more time intensive and costly than individual/relationship-level strategies, implementing them served more people over time, and the average cost per participant decreased as the strategy expanded in reach.

Limitations

This study had several limitations. First, data were collected retrospectively, which increased susceptibility to recall bias. Programs may not have been able to recount all resources, expenditures, and in-kind contributions during the previous 3 years. We were not able to account for this level of potential recall bias in our estimates. As such, amounts for some cost categories and in-kind contributions were likely underestimated. Future cost analysis, if conducted prospectively, can minimize the likelihood of this type of bias. Second, data on the number of participants reached for the first year represented the first 6 months of the funding period for LHDs as required in the 1605 program funding announcement. As a result, the analysis might not have fully captured the number of participants served in the first year. However, subsequent years did account for participants served for a full year of reporting. Third, while we acknowledge uncertainty related to the costs reported by each LHD, we had no basis for which to estimate the size or direction of this uncertainty; therefore, we did not conduct a sensitivity analysis. Fourth, we did not estimate costs before the start of the CDC 1605 program funding cycle; as such, we could not estimate costs for any youth violence prevention efforts that may have been implemented before the LHDs were funded by CDC. Additionally, activities being conducted prior to CDC funding were inconsistently reported across LHDs, with some LHDs reporting no such activities prior to 1605 program funding. As such, we focused this study on costs of strategies after 1605 program funding was awarded. We recognize that some LHD costs may underestimate the actual cost of violence prevention efforts being conducted at LHDs. Finally, we examined only costs and did not consider outcomes associated with implementation of strategies. A subsequent cost-effectiveness analysis of various strategies is a next step to inform decision-making about which strategies may result in larger reductions in teen dating violence and youth violence relative to their cost.

Conclusion

Our study estimated the total costs to implement comprehensive primary prevention strategies through LHDs to ensure population-level impact for preventing violence among young people. For all strategies implemented across the social–ecological model, LHDs became more cost-efficient over time in serving program participants. The findings from this analysis provide a foundation for budget justification, decision-making, and forecasting for investments in teen dating violence and youth violence prevention funding at the local level. Further analysis to examine differences in how LHDs deliver strategies can provide additional context when assessing potential scale-up of the 1605 program to other LHDs. Opportunities to leverage funding are also important for the sustainability of public health programs; thus, an opportunity exists to further evaluate the factors that facilitate success in leveraging funding to provide additional technical assistance and support for LHDs.