4

Data and Data Systems to Support the Infrastructure

Data are a core infrastructure component—important at every step of implementing evidence-based programs (EBPs) and informative to other components of the infrastructure, from enumerating the workforce and identifying shortages to demonstrating the value of prevention to funders. The prevention infrastructure needs both surveillance and monitoring data systems that are regularly updated to determine not only the prevalence and incidence of mental, emotional, and behavioral (MEB) disorders but also any changes in trends (e.g., new and emerging substances, new routes of administration, change in demographics of those affected). Communities and their partners need data about the MEB outcomes they are trying to enhance and the drivers of those outcomes, as the latter are the focus of prevention and health promotion interventions. Those drivers include structural (e.g., discrimination), environmental (e.g., high retail availability of substances), and social-interpersonal (e.g., family conflict, community violence) factors. These data can be used at federal, state, tribal, regional, and local levels to identify strengths and areas of concern; point to factors to intervene on for the purpose of improving outcomes; and monitor and evaluate results of policy, program, and practice interventions.

At the federal level, the implementation of the Foundations for Evidence-Based Policymaking Act of 2018 included guidance from the Office of Management and Budget (Vought, 2020) “to be used to continually improve the capacity of Federal agencies to generate evidence about effectiveness and implementation, identify areas for improvement of programs, policies, or organizations, and inform mission-critical decisions and policies.” This memo and related efforts provide a starting point

for considering a whole-of-government data infrastructure for evaluating prevention, which could be replicated at, or at least inform, program evaluation efforts at the state and local level.

At the local level, data need to be available at sufficient levels of granularity to serve the needs of local communities. They also need to be available by subgroup, particularly those that continue to be marginalized and experience persistent MEB health disparities. As with other aspects of the infrastructure, data to meet these needs may be incomplete or not easily accessible to communities, including due to funding constraints.

This chapter describes the purposes of data collection and systems to support prevention initiatives, along with key questions and principles for developing and using data. It provides a brief overview of the data sources that inform the promotion of MEB health and well-being and related gaps; outlines federal/national, state, and local models; and discusses key issues related to extant data and their use. It also identifies some sources of technical assistance and promising approaches for developing and using data systems, especially by and for the communities in which these prevention activities and associated data collection and analysis would take place.

The Centers for Disease Control and Prevention (CDC) Public Health Approach provides a simple framework (slightly updated in Figure 4-1) for discussing how data are used at different points when implementing preventive interventions. First, decision makers at the state or community level define and monitor the problem. What are the rates of substance use disorder (SUD) or mental illness in the community or in a specific setting, such as a school district? What are the trends, and how do they compare to similar communities and to the state as a whole?

Identifying risk and protective factors requires asking what increases the likelihood a person or a group will experience poor outcomes, such as developing an SUD or mental illness. These factors include structural

drivers of MEB outcomes (e.g., health care access, poverty) and other aspects of the social-ecological context (see Chapter 2), such as adverse childhood experiences (ACEs). Sources of data to inform efforts to address ACEs or community and individual trauma may include population-level data available from the Behavioral Risk Factor Surveillance System (BRFSS) and data on physical or emotional trauma, including exposure to community violence (data such as homicide rates from the National Violent Death Reporting System) (CDC, 2024d; Swedo et al., 2023). Developing and testing prevention strategies generally takes place through research, but communities can also gather important data about implementation of programs in novel settings or populations or collect data that constitutes practice-based evidence. As a prevention strategy (program, communication campaign, etc.) is implemented, the implementation team and collaborators evaluate its effectiveness and the data collected from it (e.g., pre/post or other) can inform developers of the intervention, funders, and other communities considering the approach. A strong data system—as a key characteristic of a learning prevention infrastructure—is also important to feeding back into the evidence base about an intervention’s effectiveness or lack thereof in preventing MEB disorders.

Finally, data are crucial to ensure widespread adoption of a preventive intervention if the following and similar questions are answered affirmatively: Did the strategy achieve the desired effect? Did outcomes improve as a result of the intervention?

TYPES OF DATA SOURCES

There are five sources of data relevant to MEB health and related prevention efforts. These data sources and systems already exist. There are some limitations, such as a gap in local data, or difficulty accessing and using available data sources, but there are efforts to facilitate access to more locally relevant data that can be used, or built on, to support the data needs for delivering preventive interventions and to help communities work with data (PTTC Network, 2024).

1. Administrative data from a range of public systems, including health care, education, criminal-legal, employment, housing, and human services;
2. Electronic health records (EHRs), distinguished from administrative health care claims data;
3. Survey data;
4. Data from use of online platforms, especially social media; and
5. Qualitative data including observations, interviews, and focus groups with community members.

Administrative Data

Administrative data include public- and private-sector data, often from education, human services, criminal-legal, health care, and other systems (see Table 4-1). They are valuable because they generally cover large swaths of the population. They can capture critical domains of interest, such as risk factors (e.g., exposure to abuse) and outcomes (e.g., overdoses). Despite considerable challenges related to the interoperability of these systems, a number of initiatives have demonstrated that administrative data from different sources can be merged to study people’s experiences over time and across different systems. They can be used to monitor many outcomes of interest in close to real time, a decided advantage over many other sources of data. They can miss many important outcomes, however, such as psychiatric symptoms, suicidal behavior, or untreated substance misuse. They also provide information only on preventive interventions implemented through the health care system. The use of large administrative linked data sets can help evaluate the returns on public investments in prevention through policy. Linking behavioral health (BH) outcomes data to tax records and public benefit receipt can be used for marginal value of public funds analysis, which, compared to cost–benefit analysis, can offer policy makers additional insights about policy effects (Hendren and Sprung-Keyser, 2020). It is crucial to note that the purpose of such research is not to identify specific individuals or to work with identified data. Economic researchers and other social scientists use different methods to link data sets, and multistep processes for de-identification have been developed to link data without compromising confidentiality (Jutte et al., 2011). For example,

TABLE 4-1 Examples of Sources and Types of Administrative Data Relevant to MEB Health

NOTE: MEB = mental, emotional, and behavioral.

linkages are made between people’s health outcomes data and their tax records to inform research on the relationship between the Earned Income Tax Credit and a specific health outcome, such as anxiety. There are laws and regulations (e.g., Health Insurance Portability and Accountability Act (HIPAA) Privacy Rule) that are intended to safeguard people’s privacy and confidentiality. Although many types of data needed for the prevention infrastructure are anonymized or aggregate (e.g., population surveys), there are circumstances where that is not the case, and careful stewardship and compliance with relevant laws and regulations are crucial.

EHRs

EHRs are distinct from administrative claims data in that they include a richer source of quantitative and qualitative documentation of health care encounters. Primary care clinicians could more frequently collect measures of MEB health, such as the Personal Health Questionnaire, and enter them into the EHR. They also could take advantage of the Z-codes to note health-related social needs (HRSNs). Natural language processing and other informatics approaches could be used to link social needs data to public health and clinical data (NASEM, 2019b; Hossain et al., 2023). The addition of HRSN fields in EHRs could be helpful in implementing preventive interventions in a variety of populations and predicting unmet HRSNs that have a bearing on BH outcomes or even risk of outcomes such as suicide to respond with indicated prevention strategies (He et al., 2023). The Minnesota EHR Consortium offers a good example for how aggregated EHR data can fill gaps in surveillance (i.e., high-level, timely population information on key public health indicators). It is a voluntary collaboration that uses a distributed data model. Data are held and analyzed by health systems, but summary data are aggregated at one site and available to all partners in the consortium (Raths, 2021).

Survey Data

Many federal, state, and local public health and other government agencies collect population-level and longitudinal survey data, which is reported to the public in the aggregate, and can convey the state of MEB health and provide estimates of population-level changes over time. Communities need to be able to measure and demonstrate the effectiveness of their prevention efforts, and this requires not only cross-sectional data (such as the four elements collected for the Drug-Free Communities grants) but also longitudinal cohort data, which are often available for larger geographic areas, such as county, state, and nation. Limitations and challenges

of national surveys, such as the BRFSS, include the high cost of administering the survey, lack of estimates for smaller geographic areas, survey length (e.g., the BRFSS takes over 24 minutes), inadequate representativeness, and a “lack of community engagement in survey design, analysis, interpretation of results or dissemination” (Oregon Health Authority, 2021, p. 3). CDC’s public health data modernization effort is a promising sign that these challenges will be considered and addressed.

To gather state- and local-level BH data, some states that do not participate in the YRBS implement their own surveys addressing MEB outcomes and their predictors of risk and protective factors (Chang, 2022). Examples include Washington State’s Healthy Youth Survey (HYS), Pennsylvania’s Youth Survey, and similar surveys in at least 5 states that are administered online to middle and high school students (Chang, 2022). Data from these surveys can be aggregated at state, local, and other relevant geographic levels. In some states, local data can be disaggregated by sex, race, and additional demographic characteristics. Several states have invested in dashboards that facilitate understanding trends over time and comparing community and state averages. Because of their granularity, data from the HYS and related surveys can facilitate local prevention planning by identifying locally specific strengths and areas of need and opportune points of intervention.

Requiring a common core of data elements in state surveys could address gaps left by state-specific surveys. Periodic investments by Substance Abuse and Mental Health Services Administration (SAMHSA) could also facilitate survey updates that incorporate new knowledge about risk and protective factors, changes in focal outcomes, and other realities in an ever-evolving prevention and health promotion landscape. Surveys administered to middle and high school students in over a dozen states are a valuable source of local data on risk, protection, and BH outcomes. To maximize their value in prevention planning, implementation, and monitoring, surveys benefit from periodic updates and psychometric validation. Investments in these activities would ensure surveys reflect state and community priorities and the latest prevention science.

Digital Platform Data

Big data is an emerging category that may include data from apps and social media that could contribute to real-time surveillance of public health concerns (Aebi et al., 2021). New and emerging technologies may provide future opportunities to identify at-risk populations with real-time data and deploy web-based interventions, such as identifying and preventing misinformation, developing “targeted communications to promote [behavior] change” and prediction and early intervention (WHO, 2022). More

research is needed to assess the utility and validity of relatively newer data sources, such as Internet search data and other big data (Knipe et al., 2021; Vaidyanathan et al., 2022; Wang et al., 2022) (see Box 4-1).

Qualitative data

Qualitative data include observations (e.g., to record behaviors, using standardized methods), interviews (e.g., to learn about people’s experiences), and focus groups (e.g., to gather feedback, test out different options). Qualitative research methods can be useful in evaluating prevention programs, helping to assess progress and identify needed improvement (FRIENDS National Resource Center for CBCAP, 2009).

OVERVIEW OF EXISTING DATA SOURCES

Multiple sources of data and measures can help meet the needs of states, localities, and communities implementing preventive programs (see Table 4-2). The committee discusses the data elements needed by stage of

TABLE 4-2 Select Types and Examples of Data Relevant to MEB Health by Life Stage

^a For example, “Number and percentage of public schools providing diagnostic mental health assessments and treatment to students and, among schools providing these services, percentage providing them at school and outside of school, by selected school characteristics: School years 2017–18, 2019–20, and 2021–22.” (NCES, 2023b).

^b For example, “Percentage of students in grades 9–12 who reported that illegal drugs were made available to them on school property during the previous 12 months, by selected student characteristics: Selected years, 1993 through 2021.” (NCES, 2023c).

^c https://monitoringthefuture.org/about/ (accessed 12/13/24)

^d Those measures are (1) past 30-day use (the percentage of survey respondents who reported using alcohol, tobacco, or marijuana (prevalence of use) or misusing prescription drugs at least once within the past 30 days (prevalence of misuse)); (2) perception of risk (the percentage who perceived that the use of a given substance has moderate or great risk); (3) perception of parental disapproval (the percentage who perceived their parents would feel that regular use of alcohol (one or two drinks nearly every day) or use any tobacco or marijuana or misuse prescription drugs is wrong or very wrong); and (4) perception of peer disapproval (the percentage who perceived their friends would feel it would be wrong or very wrong for them to drink alcohol regularly (one or two drinks nearly every day) or use any tobacco or marijuana or misuse prescription drugs).

^e “The report addresses the lack of a universally accepted definition of well-being and the inconsistencies in measuring well-being across campuses” and “focuses on five key dimensions: Community and Belonging, Coping and Stress Management, Purpose and Meaning, Subjective Well-being, and Institutional Environment.” (ACHA, 2021). The survey has been piloted at nine universities with over 8,000 students, staff, and faculty. (Key findings included lower scores for students compared to staff and faculty across all scales of emotional well-being, higher levels of depression, loneliness, stress, and anxiety, and low perceptions of institutional support across all three groups.)

the life course. Data sources and metrics pertain to different levels of the socioecological model of health—illustrated by a wide variety of graphics that show the individual at the center of several concentric circles, with increasing influence and potential for impact in each larger circle: interpersonal relationship, community and institutional, and societal factors. If gathering data to inform an intervention for adolescents, for example, there are

• measures of risk and protective factors captured by different types of surveys, administrative, and other data (e.g., harmful drug use, interactions with criminal-legal system, suicide attempts);
• measures pertinent to interpersonal and family relationships (e.g., indicating support, connection);
• measures of the neighborhood level, such as physical/built environment factors that operate at the community level (e.g., related to alcohol and tobacco, neighborhood walkability, housing quality and housing instability);
• measures of institutional factors (e.g., teacher stress and job satisfaction, school climate) (NASEM, 2019a); and
• measures of the aspects of the social and economic environment associated with risk factors for BH disorders (e.g., poverty, firearm-related injuries, social support).

Federal agencies provide reports and data briefs that assemble relevant information by population groups and by state, county, planning area, ZIP code, and census tract. For example, the SAMHSA Behavioral Health Barometer, last published in 2020, provides key data from the 2019 National Survey on Drug Use and Health (NSDUH) and National Survey of Substance Abuse Treatment Services. Data are organized by youth, young adult, and adult age groupings (SAMHSA, 2020). CDC’s National Center for Health Statistics provides data briefs, such as No. 467,¹ which is based on data from the National Health Interview Survey. The CDC WONDER Online Databases provide vital (birth and death) statistics. The National Survey of Children’s Health is funded and directed by the Health Resources and Services Administration (HRSA) Maternal and Child Health Bureau and conducted by the Census Bureau annually. Data include items related to social factors, such as housing instability, which is correlated with a range of poor health outcomes. The National Health and Aging Trends Study is supported by the National Institute on Aging. A number of sources of data include all age groups, such as the multisource surveillance CDC National Violent Death Reporting System and the cross-sectional household National

___________________

¹ https://www.cdc.gov/nchs/data/databriefs/db467-tables.pdf (accessed January 17, 2025).

Health Interview Survey. The National Institute on Drug Abuse (NIDA) also provides several data resources, including the ongoing research study Monitoring the Future,² and the National Drug Early Warning System.³

The NIH All of Us Research Program offers a unique opportunity to learn about the health of millions of Americans and their experiences with specific risk factors and diagnoses. In 2023, the program launched new areas of mental health research. For example, researchers examined “what kinds of social support have the greatest positive impact on mental health and are most protective against depression,” and potential future areas for research include exploring prevention strategies and early risk factors (NIH, 2024).

Public Health Surveillance Systems

Public health surveillance is a critical tool in keeping communities healthy and improving their health and well-being, and its uses are most evident with regard to infectious diseases, such as COVID-19. Public health surveillance relies on clinical care settings, ranging from primary care to hospital emergency departments, reporting infectious diseases to public health agencies. In recent years, other means of passive surveillance have emerged, including wastewater testing. During the early part of the COVID-19 pandemic, some public health agencies began such wastewater surveillance. Surveillance of BH status takes place through several avenues, including emergency department (could include firearm injuries, suicide attempts), 988 call, and pharmacy data. CDC’s Drug Overdose Surveillance and Epidemiology System collects data about nonfatal overdose through syndromic surveillance by 46 states and the District of Columbia and discharge data from 34 states and DC. It also identifies information about overdose anomalies or outbreaks and changes in trends.

The Council of State and Territorial Epidemiologists has developed a list of 18 indicators for surveillance of substance misuse and mental health (MH) (CSTE, 2019). This indicator set draws on eight data sources: mortality data (death certificates), hospital discharge and emergency department data, the BRFSS, the YRBSS, prescription drug sales (opioids), state excise taxes for alcohol, the National Highway Transportation Safety Administration Fatality Analysis Reporting System, and SAMHSA’s NSDUH (NHTSA, n.d.). These sources provide information about people, policies, and market data, such as drug sales. Public health surveillance data also provides insight into shifts in behavior, signals changes in products, and also helps detect an acute event.

___________________

² https://monitoringthefuture.org/ (accessed December 15, 2024).

³ https://nida.nih.gov/research-topics/trends-statistics/national-drug-early-warning-system-ndews (accessed December 15, 2024).

Measurement Systems That Draw on Survey and Other Data

Nonprofit, academic, and government organizations manage these data-sets and often create “dashboards” that provide a synopsis of the health of various communities. These organizations include America’s Health Rankings, which is supported and overseen by the United Health Foundation and the American Public Health Association and provides yearly state rankings on specified measures; the County Health Rankings⁴ at the University of Wisconsin—Madison Population Health Institute, which provides select measures for all U.S. counties; the City Health Dashboard⁵ at NYU Langone Health, which provides dashboards for 750 of the largest cities; and the National Neighborhood Indicators Project (NNIP), an Urban Institute coordinated network of independent organizations in over 30 cities. NNIP’s vision is that “by democratizing information, they could give residents and community organizations a stronger voice in improving their neighborhoods” (NNIP, 2024). In addition to these crosscutting sets, data sets are also produced by state and local government agencies (e.g., public health, metropolitan planning organizations, school districts) and hospitals and health systems. These metrics can inform communities about how cities or counties are performing on key metrics, counties compare to one another, their states compare to others, and their local indicators compare to their state averages.

In 2023, the U.S. Department of Health and Human Services (HHS) People and Places Thriving initiative adopted the Vital Conditions for Health and Well-Being framework, which lists seven domains that reflect key contributors to health and well-being: belonging and civic muscle, lifelong learning, meaningful work and wealth, humane housing, basic needs for health (safety), thriving natural world, and reliable transportation. These roughly map to the Healthy People 2030 framework that lists five categories of social determinants of health (SDOH): education access and quality, economic stability, social and community context, neighborhood and built environment, and health care access and quality.

The need for coordination and data integration across different government agencies, including tribal governments, and systems has been highlighted by National Academies and other reports. One National Academies report recommended that federal agencies collaborate with state and local counterparts and private-sector partners in philanthropy and business to “develop an integrated plan for longitudinal data collection and coordination and analysis of federal surveys, administrative data, and vital statistics that provides a comprehensive approach to measuring and tracking child and adolescent MEB health” (NASEM, 2019a, p. 6). Another asserted that

___________________

⁴ https://www.countyhealthrankings.org/ (accessed December 15, 2024).

⁵ https://www.cityhealthdashboard.com/ (accessed December 15, 2024).

“digitization in social care lags behind . . . health care” and listed as barriers to integration “a lack of digital infrastructure, data standards, and modern technology architecture shared between and among organizations, as well as digital privacy and security concerns” (NASEM, 2019b). These barriers affect data sharing across a range of relevant systems and domains, including health care, public health, human services, and education. Solutions include legal arrangements and application program interfaces that can reduce risks to data owners (NASEM, 2023a), such as interfaces like those that allow ride-share providers to link to data from geomapping software on a user’s smartphone.

FEDERAL AND NATIONAL DATA INITIATIVES AND GUIDELINES

Recent and ongoing examples of efforts to establish infrastructure for data exchange, linkage, or integration have relevance to promoting MEB health and well-being. These include high-level coordination within HHS to set standards, certify systems, and build a support infrastructure. The HHS Office of the National Coordinator for Health Information Technology (ONC) and CDC jointly lead the federal public health interoperability strategy (ONC, 2023; CDC, 2024b).

In 2022, the HHS Assistant Secretary for Planning and Evaluation launched the CHILDREN initiative to help public child welfare and Medicaid agencies “develop sustainable, integrated data systems with linked data, to support care coordination and oversight of prevention services and congregate care services” (Mathematica, n.d.-b; Greenfield et al., 2023). The first stage addresses data integration between the child welfare and Medicaid agencies in the District of Columbia, Iowa, Oregon, and Wyoming (Mathematica, n.d.-b). The initiative is intended to help states and tribes implement the Family First Prevention Services Act (FFPSA),⁶ which reformed how funding under Title IV-E of the Social Security Act could be used to help keep children out of foster care (ACF, n.d.; Greenfield et al. 2023).

Between 2020 and 2021, the Robert Wood Johnson Foundation convened a National Commission to Transform Public Health Data Systems, which identified key challenges to the data infrastructure: limited funding, lack of federal–state coordination, and systemic hurdles (RWJF, 2021). It called for collecting better data stratified by population groups and geographic levels to provide better understanding of health disparities, the establishment of an interagency data council to “improve measures to assess equity and racial justice and bring together different agencies to

___________________

⁶ Family First Preventative Services Act of 2017, HR 253, 115^th Congress (accessed December 15, 2024).

create interoperable social and public health data,” and ensuring “that community input is represented in data collection, interpretation and decision making” funding “to systems that are standards based and interoperable” (RWJF, 2021, p. 2).

In 2022, CDC launched its Public Health Data Strategy, and in 2024, it partnered with the Association of State and Territorial Health Officials, National Network of Public Health Institutes, and Public Health Accreditation Board to establish three Public Health Data Modernization Implementation Center Programs (Mathematica, n.d.-c). The effort includes “working with state, tribal, local, and territorial health departments to accelerate data exchange to reduce burden and improve public health threat detection and health outcomes across populations” (Mathematica, n.d.-a).

Better national coordination on prevention, such as the mechanisms recommended in the Governance and Partnerships chapter, could involve bringing together representatives from relevant agencies that play a role in preventing MEB disorders to extend or adapt interoperability standards to their data to enable integration that can better inform cross-sector work and meet the shared aims of improving MEB health.

State and Local Models and Examples

In addition to these national initiatives, state- and local-level efforts have brought together different data sources to inform decision making and the provision of health care and social services. The examples are not a comprehensive list but rather illustrate some ongoing efforts and innovations in states as geographically and demographically varied as Iowa, Oregon, California, New Jersey, Connecticut, and Minnesota and communities such as Hennepin County, Minnesota, and San Diego, California.

State Government–Academic Partnerships

The learning community training and technical assistance program at University of Pennsylvania Actionable Intelligence for Social Policy supported a partnership between Early Childhood Iowa and Iowa State University to form the state’s Integrated Data System for Decision Making and a comprehensive statewide needs assessment in 2019 with a grant from the Administration for Children and Families. The Iowa Department of Health and Human Services has conducted a HRSA-supported assessment of family support, home visiting, and community risk, using the state’s home-visiting data system and other types of data (e.g., child care subsidy, Early Head Start and Head Start, K–12 education) to ascertain “strengths

and gaps among Iowa’s home-visiting services for families with young children” (I2D2, 2021).

The Center for Evidence-Based Policy at Oregon Health and Science University operates the Oregon Child Integrated Dataset (OCID, 2020). The effort includes a dashboard that integrates the following indicators: health (birth, weight, and Medicaid well-child visits, ages 3–6); child welfare (foster care placement, early childhood, child maltreatment, early childhood); and education (kindergarten assessment, approaches to learning and early literacy, third grade assessment reading and math, ninth grade on track to graduate, student homelessness, school attendance).

In 2018, New Jersey enacted legislation that established the Integrated Population Health Data Project at Rutgers University. The project facilitates research and allows academics to support the state government in improving the health, safety, security, and well-being of state residents (Rutgers, n.d.). Its five foci are opioid misuse, maternal and infant health, SDOH, COVID-19 and other public health emergencies, and high-value care.

The University of Delaware prepares the annual State of Delaware Epidemiological Profile for substance use, MH, and related issues for the State Epidemiological Outcomes Workgroup, which promotes “the use of behavioral health data for prevention, strategic planning, decision-making, and evaluation.” (University of Delaware, 2024). The profile draws on national data sources (NSDUH, BRFSS, and the Household Pulse Survey), Delaware School Survey and the state’s YRBSS, data from the state’s divisions of public health and of substance abuse and MH, the state’s department of safety and homeland security, and crisis text line (Delaware State Epidemiological Outcomes Workgroup, 2023). The state uses Cantril’s Ladder as a measure of well-being. This simple, self-reported measure asks respondents to place their state of being on a 10-step ladder to indicate where they stand in relation to the best possible life, with rungs from Suffering to Struggling to Thriving.

Data to Drive Resource Allocation and Evaluation

Data play a critical role in Washington State’s Community Prevention Wellness Initiative (CPWI), a community- and school-based effort to prevent SUD (Mariani, 2024). Because of limited public prevention dollars, the Division of Behavioral Health and Recovery (DBHR), the single state agency for administering SAMHSA block grant dollars, prioritizes communities with higher need and greater risk for substance use problems. Eligibility is determined by a risk index developed with local data about crime, truancy, BH problems, and other factors. DBHR uses data from Washington State’s HYS to evaluate the results of CPWI efforts.

Guidance for State-Level Health Information Sharing

California established the State Health Information Guidance, which supports information sharing among collaborators, including for BH. In April 2023, it expanded to include 22 scenarios derived from real user stories, which clarify how state laws govern the exchange of information about MH and SUD (Center for Data Insights and Innovation, n.d.).

Academic and Local Government Partnership

The University of Arizona has developed a dashboard that provides 36 specific health and SDOH measures for southern Arizona counties. It also allows Arizona cities to compare themselves with peer cities on key BH metrics, including number of poor MH days per month, percent of population with depression, percent of population who drink to excess, and SDOH measures, such as housing cost burden and poverty (Making Action Possible, n.d.).

Health Systems Data Integration Partnership

In Minnesota, the state’s 11 largest health systems collaborate in its Electronic Health Record Consortium, which was launched in March 2020 to address gaps in traditional public health surveillance and support public health response to COVID-19. Hennepin County adapted it to inform the county’s efforts to address substance use–related harms. The consortium collects timely health care use data that informs public health authorities and health care organizations about nonfatal overdoses and other relevant data, in addition to highlighting demographic patterns. One of its projects is Health Trends Across Communities, which provides a data dashboard with 24 priority health topics and technical assistance resources (MN EHR, n.d.). The consortium uses a master data use agreement (also known as a “data sharing agreement”) that is amended for each new project (MN EHR, 2024).

State Government Data Effort

In Connecticut, DataHaven is a partner of the NNIP, and its largest program is the DataHaven Community Well-Being Survey, which is a source of information about quality of life, public health, economic development, and civic vitality for Connecticut communities and others in the region (DataHaven, n.d.). The survey “uses probability sampling to create highly reliable local information that is not available from any other public data source” (DataHaven and Siena College Research

Institute, 2024). Additionally, DataHaven produces a Community Index and the Connecticut Town. In 2023, DataHaven and its partners released the 2023 Community Well-Being Index at the State Capitol. A 2021 DataHaven report commissioned by the Quinnipiac Valley Health District explored root causes for a 40 percent increase in fatal overdoses in the New Haven area and recommended prevention strategies (Davlia, 2021).

These examples illustrate the breadth of capabilities that exist across different types of partnerships to support the collection and reporting of data from multiple sources to inform BH-related and other prevention efforts of states and localities. Although using, sharing, and integrating existing, new, and emerging data sources has challenges, especially with identified data, there are robust guidelines and frameworks to guide data sharing and integration, as well as methods for integrating multiple sources and systems (e.g., health care, housing, education, child welfare) not only for efficient service delivery to individuals but also to optimize the planning, implementation, and sustainment of population-level prevention strategies across a variety of settings and along the life course.

Privacy Standards

Privacy issues are largely relevant to health care contexts and, given that a considerable proportion of prevention services are delivered at the community level and not in clinical settings, there will be minimal risks to individual privacy. When health care providers share data with social services providers, such as using a single EHR to help serve individuals more seamlessly (e.g., case management, housing, food assistance) and an integrated data warehouse for analysis and reporting, the data collection and sharing are governed by both federal and state laws and policies (Owen, 2014). In educational settings, if a school is by the Health Insurance Portability and Accountability Act of 1996 (HIPAA) because it employs a health care provider that bills Medicaid electronically, health information maintained under “education records” is governed by Family Educational Rights and Privacy Act of 1974, and the school is not required to comply with the HIPAA Privacy Rule, though it must comply with the HIPAA Transactions and Code Sets Rule (HHS and ED, 2019). In many cases, however, health care systems can share deidentified data “to help public health departments and policy makers monitor and respond to emerging mental health issues” (Purtle et al., 2020). Also, health care systems along with public health agencies can advocate for evidence-supported public policy changes that can improve MEB health and well-being.

Building Community Capacity to Use Data

Some components of the systems needed to consistently collect and collate community data are already in place, but challenges remain. They include lack of skilled personnel to collect and analyze data, barriers to integration and interoperability, difficulty accessing data, and cost of some private-sector sources. Sources of data may include government agencies at all levels, including local health departments, schools, and area agencies on aging. Data collection (primary or secondary data) is necessary but not sufficient for communities to track their progress.

The prevention infrastructure in communities unfolds in a variety of settings, two of which are multisector community coalitions and primary care settings, beginning with community health centers and federally qualified health centers. Those working in and with communities to identify existing and collect new data inevitably need to partner with a variety of organizations, ranging from academic institutions to cooperative extensions.

Communities, partners, and prevention leaders and workers need not only the contextual data described (e.g., how a community or region compares to others and the averages in that state) but also hyperlocal or granular data. Local data (e.g., for census tract or zip code) are available from a variety of partners (public health agency, hospital, community health center, etc.) and sources. However, some data (e.g., program implementation and outcomes) must be collected and analyzed locally by the implementers of evidence-based interventions or their collaborators. These latter include data about local needs and assets, local capacity and other institutional or system features, and intervention- and implementation-specific features, such as reach and penetration and fidelity to design.

Community Coalitions

Data from the Office of National Drug Control Policy about the 2022 cohort of Drug-Free Communities–supported coalitions showed that “For all coalitions since inception, past 30-day use rates decreased significantly across all substances at both the middle and high school levels, evidence that DFC coalitions are meeting the goal of preventing youth substance use. That is, there were significant decreases in past 30-day use across substances. This same pattern held true for the FY 2021 cohort” (Drug-Free Communities Support Program and Executive Office of the President of the United States, 2023, p. 37). These data show that community coalitions can help improve key BH outcomes.

CADCA (formerly the Community Anti-Drug Coalitions of America) outlines the kinds of data collection or assessments that coalitions need to conduct to inform their work: quantitative, qualitative, and resource

assessment. Data resources include surveys (available nationally, such as YRBSS, or at the state level and potentially also locally from partners and collaborators that collect data in the community) and archival or secondary data from other organizations, such as arrest data from local law enforcement or substance use treatment data from the state health department (CADCA, n.d.). Qualitative data collection methods include key informant interviews, focus groups, listening sessions, town hall or community meetings, observation, and environmental scans. Resource assessment includes identifying assets (broadly conceived), capacity, and a range of resources such as existing programs, initiatives, coalitions, laws and policies, and funding streams.

A growing body of research examines the relationship between different aspects of community coalition work and change in the outcomes of interest. One study of coalitions across multiple countries found that “perceptions of their community and their pride in community are closely linked to the way they talk about their life satisfaction and mental well-being and may be a key pragmatic measure for initiative success” (Powell et al., 2024). These data have been used to show that community coalitions can help improve key BH outcomes. Box 4-2 provides an example of a coalition’s use of different kinds of data to inform planning.

Green and colleagues (2024) provide a noteworthy example of a finding about local capacity—an oversaturation of SUD prevention coalitions in a county in Montana, which indicated a risk of inefficient use of resources and diluted focus (Green et al., 2024). Researchers used the metric “program saturation,” estimated with SAMHSA’s Calculating an Adequate System Tool algorithm. Coalition Check-Up is another resource for coalitions.

Data Equity and Data Ownership

Data equity considerations are needed to inform all aspects of data collection, use, and reporting. The Office of Management and Budget’s report on the implementation of Executive Order 13985 Study to Identify Methods to Assess Equity: Report to the President states that “a broad range of assessment frameworks and data and measurement tools have been developed to assess equity, but equity assessment remains a nascent and evolving science and practice.” (OMB, 2021, p. 14). Beyond measuring the implementation and outcomes of prevention programs, it is important to recognize that data belong to the communities from which they are collected, and communities need support in using their own data for planning and evaluation purposes.

The CDC Foundation has developed five equity principles for using public health data in partnership with the communities from which the data come: recognizing and defining systemic factors, paying attention to equity in language and action, making space for cultural modifications, developing a shared data agreement, and facilitating community governance of “the collection, ownership, dissemination and application of their own data.” (CDC, 2024). Existing data may have biases and gaps, and the Rutgers Policy Lab notes that holding data producers accountable for data equity requires considering how data are collected, analyzed, interpreted, and shared and also “encourages further inspection into potential racial bias of research instruments, publication’s role in the reinforcement of stereotypes, and marginalized communities’ ability to control and access their own data. It also cautions against data misuse and inaccurately broad generalization. Data equity also considers issues regarding power and privilege between researchers and their targeted populations and concerns that harmful decisions might be justified through data” (Spiegel, n.d.). Implementation considerations are discussed in Box 4-3.

Using Data for Community Action

Several important resources inform variable and measure selection and data use and implementation considerations. The Prevention Technology Transfer Center (PTTC) Network has put forward a checklist of six data areas aligned with the SAMHSA Strategic Planning Framework to help communities identify data gaps: consequences, consumption, target populations, intervening variables (risk and protective factors), prevention resources, and community readiness (PTTC Network, 2021). For MH prevention work, other assessment data would be substituted for consequences and consumption.

___________________

⁷ The report found that a “broad range of assessment frameworks and data and measurement tools have been developed to assess equity, but equity assessment remains a nascent and evolving science and practice” and made four recommendations on this topic:

Recommendation 5: The Office of Management and Budget (OMB) should require the Census Bureau to facilitate and support the design of sampling frames, methods, measurement, collection, and dissemination of equitable data resources on minimum OMB categories—including for American Indian or Alaska Native, Asian, Black or African American, Hispanic or Latino/a, and Native Hawaiian or Pacific Islander populations—across federal statistical agencies. The highest priority should be given to the smallest OMB categories—American Indian or Alaska Native and Native Hawaiian or Pacific Islander.

Recommendation 6: The Office of Management and Budget (OMB) should update and ensure equitable collection and reporting of detailed-origin and tribal affiliation data for all minimum OMB categories through data disaggregation by race, ethnicity, and tribal affiliation (to be done in coordination with meaningful tribal consultation), including populations who self-identify as American Indian or Alaska Native, Asian, Black or African American, Native Hawaiian or Pacific Islander, and Hispanic or Latino/a.

Recommendation 7: The Centers for Disease Control and Prevention should coordinate the creation and facilitate the use of common measures on multilevel social determinants of racial and ethnic health inequities, including scientific measures of racism and other forms of discrimination, for use in analyses of national health surveys and by other federal agencies, academic researchers, and community groups in analyses examining health, social, and economic inequities among racial and ethnic groups.

Recommendation 8: Congress should increase funding for federal agencies responsible for data collection on social determinants of health measures to provide information that leads to a better understanding of the correlation between the social environment and individual health outcomes. (NASEM, 2023b)

Communities and their collaborators in prevention require a parsimonious data set that describes outcomes of interest (e.g., substance use, psychological well-being, suicide rate) and measures that can inform them about sociodemographics, assets and resources, and data on risk (e.g., exposure to ACEs, such as abuse and violence) and protective (e.g., supportive family relationships) factors. Communities need to be able to use data to track and compare needs and outcomes against their own historical data (to assess impact, course correct), compare themselves against other communities, and determine the effectiveness of prevention efforts.

A key characteristic of effective data collection and integration efforts is that they occur in the context of multisector and community partnerships. The “Learning Health Care Communities” (Natafgi et al., 2021) model describes the relationship between the learning health care system (data are turned into knowledge that shapes patient care) and the learning health care community (health measures create awareness that help improve community health). Other public health and social or community change models exist that can inform and guide community-partnered MEB health promotion. One example is Mobilizing for Action through Planning and Partnership (MAPP and MAPP 2.0) from the National Association of County and City Health Officials that includes community partner assessment and both quantitative (community status) and qualitative (community context) assessments. Another is the collective impact model, which includes shared measurement systems as one of five conditions of collective success (with common agenda, mutually reinforcing activities, continuous communication, and backbone support organization) (Piff, 2021; Tamarack Institute, 2017). Communities also can learn from partnering with organizations about what measures and data are available to them (e.g., health department, schools, and area agencies on aging) to inform their planning, selection of interventions, implementation, and evaluation.

To support community data infrastructure, government agencies at all levels can provide funding for data sharing, analysis, and collection. While HHS agencies have a history of temporarily funding resources that provide local, small-area data, including on SDOH health elements, and mechanisms that allowed peer comparisons and collaboration, such federal efforts have fluctuated over time.

DATA ISSUES (CHALLENGES AND OPPORTUNITIES)

Some key issues related to data and measures in the prevention infrastructure for MEB health and well-being include the following:

1. The need to build a health equity focus into data and measurement systems themselves;
2. Community level—the role of communities in deciding what matters most and collecting and using data; frequency of collection and associated concerns, such as community burden/fatigue;
3. Related infrastructure and resource needs (i.e., workforce and funding); resources to address technical hurdles for integration and interoperability;
4. Regulatory and governance (e.g., data sharing and innovation); ownership and sovereignty; privacy and confidentiality when individual-level data are collected or needed;
5. Lack of hyperlocal/granular or small-area data; and
6. Inconsistencies in data.

Building an Equity Focus in Data Systems

Data are collected and used in the process of assessing and responding to the MEB health needs of populations and communities, so MEB health disparities are an important consideration. Data and measures may seem neutral in terms of their provenance, means of collection, and meaning, but that is not the case. Rather, they reflect values and norms and decisions about what matters and what does not, and they may hide or obscure important truths. Nancy Krieger described data as a two-edged sword: edge 1 is “no data, no problem,” and edge 2 is “problematic data, big problem” (Krieger, 2021, p. 2). The former refers to the nonuse of data related to populations marginalized on the basis of race or ethnicity. That is exemplified in the inadequacy of data documenting the impact of COVID-19 in minoritized communities, notably in American Indian, Alaska Native, and Native Hawaiian populations (Douglas et al., 2021; Urban Indian Health Institute, 2021). The latter refers to the use of “problematic data in harmful ways” (Krieger, 2021, p. 2).

Community level

Communities have often been the locus for and subject of data collection. A long history exists of data collection (and research) approaches that exploit this resource, may be unaccountable to the community, and may not adequately engage or partner with, community members and organizations (Emmons et al., 2023). The Patient-Centered Outcomes Research Institute (PCORI) provides a good model for sharing research data with community members who contributed (PCORI, n.d.). Data to support MEB health promotion can be generated with the community and used to respond to its needs and priorities. As more community coalitions and organizations work with partners, including academic institutions, to implement preventive interventions and generate data, locally developed insights and implementation experiences could inform other communities, and efforts are needed to ensure the wide dissemination of information about intervention impact and other practice-based evidence.

Related infrastructure and resource needs

Financial resources are a challenge in establishing, operating, and sustaining data systems or data collection efforts. Braiding and blending of funding is often needed for data systems, given that many funding sources often lack robust support for data collection and measurement efforts (SAMHSA, 2024). But a patchwork of funding streams also has consequences, such as limiting data sharing, such as if data are associated with categorical programs that have narrowly constituted funding. There are workforce implications for data collection, too. For clinical, human services, education, and other types of metrics, data collection needs could overtax service providers.

Regulatory and Governance Issues

Solutions exist for most or all of these issues and challenges, including models and guidance, tools such as data use agreements, application programming interfaces, and other approaches that can be used to safeguard privacy and confidentiality whenever individual-level data are involved. An oft-raised concern is the perceived barriers to data sharing across systems and sectors, such as between state (or local) agencies or states and the federal government. The guidelines provided by the National Academies report Toward a 21st Century National Data Infrastructure include the following: “Data sharing is incentivized when all data holders enjoy tangible benefits valuable to their missions, and when societal benefits are proportionate to possible costs and risks” (NASEM, 2023a, p. 6).

Inconsistent or lack of state reporting

For some key metrics, there are inconsistent processes among states for collecting the same data. This may affect the ability to assess the effects of state policies on MEB outcomes (e.g., using quasi-experimental approaches that that compare states that do and do not adopt a preventive policy). For example, the counting of suicide deaths differs in county coroner-only states and coroner-only and state medical examiner states (Fernandez and Jayawardhana, 2024).

As noted above, some states opt out of the YRBSS and use their own youth survey, which in some cases may have data elements that align with the YRBSS. This is important to allow comparisons. However, there are cases, such as in Florida, where participation in a national state-level survey—the YRBSS—has been discontinued (Dollard 2023). That may impede the longitudinal tracking of MEB health, a crucial source of information for planning, implementation, and evaluation. James (2023) notes that unlike other federal surveys, states and local jurisdictions are only asked to use 60 percent of the survey, which “provides a lot of flexibility to state and local jurisdictions, and several states have exercised that flexibility to remove questions related to sexual orientation, gender identity, and sexual behaviors. The ability to adapt the YRBSS to meet local needs is another reason the complete withdrawal of a state is deeply concerning.”

Need for small area data

One gap in the data infrastructure needed to inform MEB disorder prevention efforts is a federal community resource. HHS has had several efforts to create such a data repository, including the Community Health Status Indicators (CHSI) project, which operated for several years, with some interruptions, and the Health Indicators Warehouse. Both were ended in 2017. Phillips and colleagues (2021) outline the history of CHSI, which allowed counties to compare themselves to their peers (in 88 different strata that clustered counties according to several categories of similarity). The tool was moved from HRSA to CDC and terminated in 2017, along with several other federal community-level data resources: the Health Indicators Warehouse, Health Data Interactive, and BRFSS multiyear data roll-ups. In 2018, the National Committee on Vital Health Statistics partnered with the 100 Million Healthier Lives (100M Lives) network of communities, health coalitions, nonprofits, health care organizations, and others, in an effort that yielded the Well-Being in the Nation measurement framework and vetted measure sets. Five principles guided the collaboration:

• Codesign;
• Continual alignment (with Healthy People, with the County Health Rankings, etc.);

• Continual testing in the field with those who would use it;
• Flexibility and balance in measure selection, adaptation, and use; and
• A focus on creating a living library of measures (Saha et al., 2020, pp. 644–45).

The resulting measure sets (of overall well-being, of the well-being of people, and the well-being of places) are intended to “help communities access data that are already being connected,” “connect what matters to create equitable well-being” and “drive their own change” (WIN Network, n.d.). They include nine core measures in three sets: well-being of people (e.g., Cantril’s Ladder; life expectancy at birth); well-being of places (e.g., child poverty; healthy community indexes aligned with the framework); and equity (e.g., differences in well-being; years of life lost; income inequality; high school graduation rates; demographic variables to use in a standard way for equity analysis) (Saha et al., 2020).

The Federal Data Strategy is expected to create access to federal data assets that could help inform policy making in compliance with the provisions of the Evidence Based Policymaking Act of 2018 (Phillips et al., 2021). Its implementation includes the Population-Level Analysis and Community Estimates (PLACES) initiative, which provides “model-based estimates across 36 health measures, including seven disability measures added in 2023, for every county, city, and census tract in the U.S.” (Wiltz et al., 2024). PLACES develops small-area estimates using BRFSS data, and data from the U.S. Census and American Community Survey.⁸

This is a good starting point, although more could be done to support the work of communities. Phillips and colleagues (2021) noted that PLACES “has limitations as a health planning tool as the imputed data can be at odds with locally collected data and cannot support tracking of intervention-related changes” (Phillips et al., 2021, p. 1871). They added that the importance of PLACES “for local health assessment and intervention efforts may be the platform and its functionality on which needed community data elements could be loaded” (Phillips et al., 2021, p. 1871). The result of the National Committee on Vital and Health Statistics collaboration with 100MLives, Philips and colleagues note, offers “a vetted framework for matching community measures to federal (and other) data assets, defining an automated routine for analyzing relevant data sets, and offering the analytic outputs publicly and equitably” (Phillips et al., 2021, p. 1871). Although PLACES may have some limitations, it can serve as an important tool for communities and could be strengthened or expanded to increase its usefulness. Earlier community data efforts could inform its continued development.

___________________

⁸ https://www.cdc.gov/places/methodology/index.html (accessed December 13, 2024).

RECOMMENDATION 4-1: The Centers for Disease Control and Prevention (CDC) should sustain, enhance, and regularly update Population Level Analysis and Community Estimates (PLACES) as a data tool that communities can access for locally relevant, granular (i.e., census tract and ZIP code) data and the ability to compare themselves to peers. CDC should enhance PLACES in collaboration with the Substance Abuse and Mental Health Services Administration to add measures relevant to mental, emotional, and behavioral health and population well-being, and support functionalities to PLACES that allow community partnerships to layer their own data on PLACES data for their planning and evaluation efforts.

Communities will require funding to support data infrastructure and staffing, and identifying data elements will need to involve community expertise at every step. For example, the surveys of middle and high school students in over a dozen states are a valuable source of local data on risk, protection, and BH outcomes. Investments in these activities would ensure surveys reflect state and community priorities and the latest prevention science.

RECOMMENDATION 4-2: The Substance Abuse and Mental Health Services Administration, Centers for Disease Control and Prevention, and other federal agencies that provide resources for community-based prevention of behavioral disorders should include specific support for data infrastructure in all relevant grant programs, including funding for acquiring relevant data, data integrity and privacy, new data collection, data sharing, collaboration with relevant public- and private-sector partners, and obtaining training and technical assistance as needed.

Measuring Well-Being

HHS has made well-being a central concept in its Healthy People 2030 initiative’s vision and adopted Overall Health and Well-Being Measures (OHMs), which it describes as “broad, global outcome measures intended to assess the Healthy People 2030 vision. OHMs can be used to summarize and evaluate progress toward achieving Healthy People objectives” (HHS, n.d.). HP2030 includes eight OHMs organized in three tiers: well-being (a measure of life satisfaction in the current year), healthy life expectancy, and summary mortality and health.

One noteworthy addition to the measurement of well-being is the development of an Indigenous measure of overall health and well-being (or wicozani), the Wicozani Instrument, by a Dakota community (Peters et al., 2019). This nine-item validated self-report measure applies Indigenous

epistemology. Peters and colleagues (2019) noted that “[w]hile there is diversity among Native communities, the Wicozani Instrument may appeal to many Native communities because health is defined from an Indigenous perspective, in that it is viewed through a holistic lens and relies upon the understanding of relationality and interdependence between physical, mental, and spiritual health” (Peters et al., 2019).

NASEM panels have also made the case for using self-reported summary measures of population well-being as important signals as to how people in the nation and at state and community levels are doing. Such measures, by their nature, draw on peoples’ mental and emotional assessment of their lives in the context of their social, economic, and environmental circumstances. One such measure is Cantril’s Ladder (NASEM, 2020). It is used internationally, including by the Organisation for Economic Cooperation and Development, and has been demonstrated to have predictive validity for health outcomes. Scores of 7 and above indicate “thriving,” while more than 4 and below 7 is considered “struggling”—with twice the number of sick days as those with higher well-being—and a score of 4 or below is considered “suffering” (WIN Network, n.d.). The most recent data on the United States averaged 6.7, with young people’s scores bringing the average down.⁹

The committee asserts that measures of well-being, combining individual and structural-level factors, are critical constructs related to MEB health. A measure of population well-being would also provide a more expansive way to track and demonstrate progress, complementing specific national measures, such as for deaths of overdose and suicide, and framing a positive high-level target for the prevention infrastructure.

Such measures can be used to periodically ascertain, track, and report on population well-being and integrated into Healthy People and related state and tribal population health reports. The committee believes that a summary measure of well-being is needed at all levels of government that can be disaggregated by subpopulation to examine and track population-level inequities and progress in overcoming them.

RECOMMENDATION 4-3: To identify and adopt measures of population well-being that allow the nation to track progress and report on mental, emotional, and behavioral health, the Office of the Assistant Secretary of Health, National Center for Health Statistics, and the Substance Abuse and Mental Health Services Administration should convene and collaborate with relevant partners.

___________________

⁹ Use cursor to view United States in https://ourworldindata.org/grapher/happiness-cantril-ladder (accessed December 15, 2024).

The measure(s) would be disaggregated by socioeconomic factors, race and ethnicity, age, and geography.

Relevant partners may include the HHS Office of the National Coordinator of Health Information Technology and national public health organizations, such as the Association of State and Territorial Health Officials and the National Indian Health Board.

The HHS Strategic Plan, to which ONC, SAMHSA, CDC, and NIH are contributors, includes Objective 4.4: “Improve data collection, use, and evaluation, to increase evidence-based knowledge that leads to better health outcomes, reduced health disparities, and improved social wellbeing, equity, and economic resilience” (HHS, 2024). In 2024, the Office of Management and Budget finalized revisions to Statistical Policy Directive No. 15 (SDP 15), which updates standards for maintaining, collecting, and presenting federal data on race and ethnicity. A National Academies report on federal policy effects of race, ethnicity, and tribal health equity discussed an early version of SDP 15 that was posted to the Federal Register in January 2023 (NASEM, 2023b).

Elevating prevention leadership to the White House Domestic Policy Council and a more robust BH Coordinating Council in HHS (see Governance chapter Recommendation 6-1) could include an initiative to create linkages among multiple data sources across the federal government, perhaps building on the CDC PLACES initiative. Such an effort could bring together administrative data (EHR, other system data such as education, child welfare), survey data (e.g., BRFSS), surveillance (reporting), big data (e.g., social media), and other types of data (including social sensing index using social media inputs to measure MH status in a specific geographic location) (Park et al., 2024). As noted, some states are already working with integrated or linked sets of measures that draw data from different systems.

HELPING COMMUNITIES WORK WITH DATA

Implementing evidence-based interventions that are well suited to a community and setting requires that communities, community coalitions, and their partners undertake a process to (1) choose what outcomes are of interest, (2) design feasible data collection and reporting, and (3) implement a method for using data to inform feedback/progress. State public health or BH agencies can help. For example, Washington State conducts state-level surveys, supports communities conduct surveys, administers a surveillance system (Community Outcome and Risk Evaluation Information System), and provides other data resources (Mariani, 2024). New York State has a state epidemiological workgroup to oversee the collection of data aligned with the SAMHSA Strategic Prevention Framework (e.g., risk and protective factor data) and also meets regularly with other relevant agencies in

the state. Like Washington State, New York State works to make data more available and accessible, including to and through its community coalitions and state resource centers (Cunningham, 2024). Pennsylvania collaborates with counties in the process of needs assessment, using data primarily from the Pennsylvania Youth Survey (for most of the state) and YRBSS (Philadelphia and Pittsburgh), in addition to local surveys. The state works with multiple sources of data that have various limitations and caveats (e.g., NSDUH data are old and regional, many sources do not provide sufficiently local data) and has begun a recent effort to gather county-level evaluation (outcomes) data from counties.¹⁰

Technical Assistance for Communities

Although the prevention infrastructure is fragmented, robust resources are available to provide training and technical assistance to coalitions and organizations endeavoring to use data and indicators to inform their planning, implementation, and evaluation. Technical assistance is characterized by a focus on improving capacity of organizations or systems, and on targeted and tailored supports by subject matter experts (Scott et al., 2022). Although technical assistance is insufficiently reported on—and there is a lack of standard definition, objective measures, and reporting standards—existing research indicates that TA is more likely to be effective if it is intensive and there is fidelity of TA practices (Scott et al., 2022; Dunst et al., 2019; Anderson et al., 2021). TA sources include especially a range of partners, such as the PTTC Network (focused on SUD prevention) and previously a network of Mental Health Technology Transfer Centers (the MHTTC network was closed September 29, 2024). In place of the MHTTCs, SAMHSA has launched a new National Center for Mental Health: Dissemination, Implementation, and Sustainment, which will establish five bi-regional centers. PTTCs are available to provide technical assistance to public health departments, BH agencies, community coalitions, and other partners on prevention of substance use. The CDC-funded National Centers of Excellence for Youth Violence Prevention play a similar role, and some of CDC’s Prevention Research Centers have, over the years, focused on aspects of preventing BH disorders (CDC, 2024c).

Kingston and colleagues (2016) provide detailed discussion for how academic institutions partnering with communities can help them use data and evidence to assemble packages of programs that are appropriate for their specific youth violence prevention needs. For example, a University of Colorado Boulder team worked with the Montebello community in Denver

___________________

¹⁰ G. Kindt, personal communication, June 25, 2024.

to support them in using the Communities That Care model to establish community governance, review data about community risk factors, and consider EBPs (using an existing clearinghouse of EBPs). The university team gathered baseline data through parent and youth surveys and worked with neighborhood partners to prioritize a small set of risk and protective factors.

Virginia Tech and Iowa State University collaborated to create a model for data-driven community engagement and community-based research, the Community Learning Through Data Driven Discovery (CLD3) (Keller et al., 2018). One pillar is the Cooperative Extension System funded through the U.S. Department of Agriculture. Cooperative extension professionals work with university researchers to “translate their science-based research results into language and decision tools appropriate for targeted audiences” (Keller, 2018). Another pillar is the network of four Regional Rural Development Centers that “serve as sources of economic and community development data, decision tools, education, and guidance in rural communities” (Keller, 2018).

The University of Pennsylvania Community Engagement and Research Core could be helpful to communities in building their own approach to identifying and using data to inform action (CEAR, n.d.). Participating institutions can collaborate with community leaders “to understand community needs and improve community health” (NCATS, 2023).

CDC Prevention Research Centers (PRCs) can support communities in gathering data to inform their efforts. For example, the Fayette County Integrated Community Engagement Collaborative has collaborated with the local school district and West Virginia University’s PRC to collect data to inform “what can be done to reduce risk and protect children and adolescents” (ICE Collaborative, 2020, p. 4). With the help of their academic partner, the collaborative collected data through a 2023 high school and middle school surveys that drew on several sources, including the YRBSS, National Institutes of Health Monitoring the Future survey of youth substance use, and European School Survey Project on Alcohol and Drugs (ICE Collaborative, 2023a-b).

Using Community Evaluation and Outcomes Data to Inform Planning

The work of implementing EBPs to promote MEB health and well-being is cyclical (see Figure 4-2), with data such as needs assessments informing initial decisions about the outcomes of interest and selection of interventions and evaluation data indicating progress and informing next steps. It is important for community knowledge generated during implementation to be shared with others systematically, similar to the post-marketing surveillance (also known as Phase IV clinical trials) conducted once a drug begins to be widely prescribed and used. Communities collect data for their own local partnerships’ continuous improvement and monitoring, but there are

broader implications. The learnings and findings that result from implementation of EBPs outside of research settings need to be shared widely and make their way into the practice-based knowledge that informs other communities and could be used to spread and scale programs.

Data Systems Require Funding

The cost for a community data system will vary depending on community size, available in-kind resources (such as free data infrastructure support), and the data needs. Local data intermediaries are one type of model for the core of a community’s data system. Local data intermediaries serve as “mediator between data and local stakeholders—nonprofit organizations, governments, foundations, and residents . . . are data translators, educators, conveners, collaborators, and voices for change. They use data to describe their communities, and they empower communities to use data in their activities, from community building to advocacy and program

planning, to policymaking” (Hendey et al. 2016). Data intermediaries can be used to derive rough estimates for the cost of operating a data system in a community. The cost of launching a data intermediary will require at least one full time employee at $100,000–150,000 (Hendey et al, 2016).

Another model is the community data infrastructure, as found in the work of the National Neighborhood Indicators Project. An analysis of the membership of 35 partners shows that the range of budgets for partners in NNIP ranges from $200,000 to $604,000, with a median budget of $325,000 (Kingsley et al., 2015). That provides a rough figure for the cost of community data infrastructure. Multiplying $325,000 by 3,142 counties (U.S. Census figure), it would cost nearly $1 billion dollars for every county to have a community data infrastructure.

REFERENCES

This page intentionally left blank.