Contextual determinants of family-driven care implementation in juvenile justice settings

Abstract

Introduction

Engaging families in behavioral health services is a high priority for juvenile justice (JJ) systems and family advocacy groups. Family-driven care (FDC) enhances family engagement and decision-making power in youth behavioral health services, ultimately, improving youth and family mental health and substance abuse outcomes. Despite the benefits, there is limited guidance on how to integrate FDC into behavioral health care within the JJ system. Therefore, the goal of this study is to understand factors that promoted adoption of FDC the JJ context.

Methods

JJ staff and leadership across the state of Georgia participated in surveys and interviews to understand contextual implementation determinants related to the adoption of FDC. Between November 2021- July 2022, 140 JJ staff participated in the survey from 61 unique JJ organizations. In addition, 16 staff participated in follow-up key informant interviews to explain quantitative findings.

Results

Based on a mixed methods analysis, JJ agencies were more likely to implement FDC if they had the following characteristics: (1) presence of site leaders that were strongly committed to family engagement, (2) a shared understanding that family engagement was a top priority, (3) staff training related to family engagement, (4) external partnerships with organizations that serve families, (5) a workplace culture that was supportive of innovation, and (6) presence of family engagement programs that were easier (or more feasible) for staff to implement.

Discussion

This mixed methods study underscores the importance of strengthening these 6 inner and outer setting implementation determinants when preparing to integrate FDC into JJ systems. Findings are used to promote the adoption and delivery of this high priority intervention in a state-level JJ system.

Introduction

Justice-involved youth (JIY) are disproportionately at-risk for behavioral health (BH) conditions, including substance abuse and mental health disorders. Approximately, 70% of JIY meet criteria for at least one BH condition (Shufelt & Cocozza, 2006), compared to about 13% in the general adolescent population (Merikangas et al., 2010). Specifically, 50% of JIY have substance use disorders (Wasserman et al., 2010), 47% of have disruptive or conduct disorders, 34% have anxiety disorders, and 19% have mood disorders (Shufelt & Cocozza, 2006). Additionally, about 30% of JIY meet criteria for post-traumatic stress disorder (Dierkhising et al., 2013) and 14% have attempted suicide in their lifetime (Wasserman et al., 2010). Many JIY also have co-occurring disorders, with one study estimating that 79% of JIY who have one mental health disorder actually meet criteria for two or more psychiatric disorders and 60% meet criteria for three or more disorders (Shufelt & Cocozza, 2006).

Among JIY, untreated BH conditions can lead to negative consequences, such as suicidal ideation (Wasserman et al., 2010), trauma exposure (Wasserman et al., 2010), elevated sexual risk behaviors (Teplin et al., 2003), reductions in academic achievement (Arthur et al., 2015), and rearrests and reincarceration throughout the life course (Schubert et al., 2011). Despite disproportionate risk, only 20% of JIY needing mental health care initiate treatment, and less than 10% in need of substance use service initiate treatment during or contact with the JJ system (Burke et al., 2015; Wasserman et al., 2021).

Family engagement (e.g., including families in services and decisions related to the care of their child) is critical to promoting treatment involvement and positive BH outcomes among JIY. In the juvenile justice (JJ) setting, families play multiple essential roles such as offering emotional support, choosing appropriate treatments, reinforcing positive behaviors, and participating in family-based therapies (Osher et al., 2008; Paik, 2017). They also provide practical support like scheduling and transporting youth to treatment appointments (Osher et al., 2008; Paik, 2017; Nelson et al., 2024). Involving families in their child’s BH services enhances treatment initiation, sustains participation and improves overall BH outcomes for JIY (Hornberger & Smith, 2011). Because of these improved outcomes, expert health panels, JJ reform organizations, and family-advocacy groups have highlighted the critical need for family participation within JJ system services (AACAP, 2003; Arya, 2013; Justice for Families, 2012; National Federation of Families for Childrens Mental Health, 2008; National Institute on Drug Abuse, 2014; OJJDP, 2013; Shanahan & diZerega, 2016; Vera Institute of Justice, 2014).

SAMHSA and the Federation of Families developed a framework, Family-Driven Care (FDC), which provides guidelines for engaging families in BH services within youth-serving organizations, such as JJ agencies (National Federation of Families for Childrens Mental Health, 2008). In family-driven systems, families have a primary decision-making role in the care of their children and a decision-making role in the organizational policies and procedures governing care for all children in the system (Osher et al., 2008). FDC is comprised of 10 overarching principles; for instance, “families and JJ staff embrace the concept of sharing decision-making and responsibility for care outcomes”, “families provide direction on decisions that impact funding, policies, services, and support, and “families engagement in peer support activities to reduce isolation and disseminate information”. The FDC principles can be operationalized using evidence based FDC strategies such as collaborative treatment planning, family advisory boards, and family peer support groups (Osher et al., 2008; Piper et al., 2024b). Additional details of FDC principles and strategies can be found in Fig. 1. Overall, FDC improves family and child outcomes, including increased family service engagement, increased family service satisfaction, improved family relationships, and improved child BH (Dunst & Trivette, 2009; Horwitz et al., 2010).

Fig. 1.

Principles of Family-Driven Care

Over the past decade, JJ systems across the U.S. have made substantial reforms which facilitated their inclusion of families and the adoption of FDC principles. Instead of focusing on restrictive confinement, JJ systems are now working to improve youth long-term success by increasing their involvement in services and strengthening their family and community support systems (Nellis et al., 2009; Prison Policy Initiative, 2019). In fact, there has been a 60% decrease in the number of youth placed in locked facilities between the years 2000 and 2017, and most JIY are now placed at-home within their families and communities (Prison Policy Initiative, 2019). These reforms created a culture shift within the JJ system (from a punitive culture to a healing culture) and increased interactions between the JJ system and the home/family environment (Prison Policy Initiative, 2019). Because of this paradigm shift, JJ system- family relationships are a top priority for JJ systems across the country (Office of Juvenile Justice and Delinquency Prevention, 2010), and over 35% of US JJ agencies now have formal policies for increasing family engagement in service delivery (Robertson et al., 2019).

Similar to JJ systems nationwide, the Georgia Department of Juvenile Justice (GDJJ) has included family engagement as a central component of its strategic plan, reflecting a documented priority for JJ agencies throughout the state (Georgia Department of Juvenile Justice, 2021). In addition to system prioritization, families in Georgia also expressed interest in improving engagement: 100% of surveyed justice-involved families in the state wanted to be involved in developing their child’s treatment plan, and 98% expressed interest in participating in family programs (Forde & Schwartz, 2020). To address new priorities, GDJJ has been implementing elements of FDC over the past few years. Research revealed substantial utilization of certain family engagement strategies by JJ staff and agencies in the state, such as involving families in care decision making, alongside less frequent use of strategies like family peer support (Piper et al., 2024b). Overall, the implementation of FDC was highly variable across JJ agencies in Georgia (Piper et al., 2024b), mirroring a similar trend observed in other systems across the country (Robertson et al., 2019).

Within Georgia and nationally, implementation of FDC has been challenging, since there are numerous policy and practice barriers that hinder effective collaboration with families (Burke et al., 2014). According to a survey of justice correctional leaders, family engagement is the most challenging issue to implement in their systems (Center for Juvenile Justice Reform, 2008). For instance, JJ systems struggle to establish trusting relationships with families due to the punitive and coercive nature of the system (e.g., requiring compliance from youth and families), power differentials between staff and families, and a culture that historically minimized the role of families, blamed and shamed families for their child’s behavior, and excluded families from decisions (Arya, 2013; Piper et al., 2024a; Shanahan & diZerega, 2016).

To overcome implementation challenges in this setting, it is imperative to understand how JJ system contextual factors impact successful adoption of FDC. Prior research identified that JJ system contextual factors such as staff attitudes, system structure, policies, culture, resources, and external partnerships play a critical role in the uptake and adoption of new practices (Prendergast et al., 2017; Taxman et al., 2009). In this mixed methods study, we identified key determinants of FDC adoption across JJ agencies in the state of Georgia. Findings are used to develop context-tailored strategies to overcome implementation barriers and guide the successful scale-up of FDC in JJ agencies in the state.

Methods

Study design

The study employed an explanatory, sequential mixed methods design to explore FDC adoption and utilization in Georgia’s JJ agencies. Online surveys were administered to JJ staff (Nov 2021-Feb 2022), followed by qualitative interviews (Mar-Jul 2022) for deeper insights. The study protocol received approval from the University IRB [approval #00002068] and the JJ-affiliated research review committee. All study participants provided electronic or verbal consent prior to completing the survey or interview, respectively.

JJ system context

Each day, approximately 7,000 youth are served at the 78 community services offices and 25 detention facilities across the state of Georgia (Georgia Department of Juvenile Justice, 2021), where GDJJ provides strengths-based, evidence-based programs to improve youth BH and long-term success, including family-based programs and treatments. On an average day in GDJJ, 39% of youth are 17 years or older, 37% are 15 or 16 years old, and 14% are 14 and under. Most youth are male (70%), and 30% are female. Approximately 51% of youth are Black or African American, 39% are White, 7% are Hispanic, and 3% are another race/ethnicity. Youth can be placed in long-term secure custody, short-term incarceration, and/or community probation and diversion programs. Most JIY in Georgia have community placements (91%) and are living at home with their families (Georgia Department of Juvenile Justice, 2021).

Participant recruitment

JJ state-level leaders emailed an online survey to all eligible employees, including staff/practitioners with selected roles (e.g., community supervision, detention, education, reentry, and behavioral health staff) and leaders (e.g., division directors, administrators, and managers with organizational decision/policy-making authority). The survey, conducted through REDCap Software, took 10–15 min to complete. At the end of the survey, participants were asked to express their interest in a follow-up web-based interview. Out of 140 staff from 61 agencies who completed the survey, 30 participants were open to an interview. We purposefully selected participants from diverse roles and departments until data saturation was reached, for a total of 16 interviews. As an incentive, we donated $5 for each completed survey and $10 for each interview to a mental health charity. Participants provided electronic consent before completing the electronic survey and oral consent before beginning the interview.

Measures and data collection

Survey Outcomes

Outcome measures included (1) the adoption of FDC principles and (2) the adoption of FDC strategies (See Table 1 for items included in each measure and Cronbach’s alpha for each scale). The first measure gauged the perceived alignment with FDC principles established by SAMHSA and the Federation of families for Children’s Mental Health (National Federation of Families for Childrens Mental Health, 2008), and the second assessed the perceived adoption of evidence-based FDC strategies, using items from the Family System Engagement Index (Robertson et al., 2019). We adapted items from these scales to fit terminology used by GDJJ, so they would be understood by staff in this JJ system context. Both measures employed a 5-point Likert scale (1 = strongly disagree; 5 = strongly agree), with overall scores calculated as the mean of the items.

Table 1.

Description of variables and measures

Variable	Description	Survey Items		Cronbach’s Alpha	Source of Measure
CFIR Constructs
Intervention Characteristics: Relative Advantage	The advantages of family-driven care compared to current family engagement initiatives.	1. Family-driven care would be more effective than our current family engagement practices.		N/A	Kegler et al., 2018
Intervention Characteristics: Complexity	Perceived difficulty of implementing family-driven care.	1. Implementing family-driven care seems easy to do.*		N/A	Sales et al., 2021 and Kegler et al., 2018
Intervention Characteristics: Cost	Costs associated with implementing family-driven care.	1. Family-driven care would be too expensive to implement at my agency.		N/A	Sales et al. 2021
Individual Characteristics: Negative Attitudes	Staff’s negative and stigmatizing attitudes about families and family engagement.	1. Families want to be involved in their child’s care.* 2. Staff are too busy to involve families in the care process. 3. Working with families makes my job harder. 4. Engaging with families can be stressful. 5. Families are the cause of their child’s delinquent behavior. 6. Families are difficult to work with. 7. Families do not listen to staff recommendations.		0.81	Self-developed. Based on a review of the literature on stigmatizing and negative attitudes about family engagement.
Inner Setting: Culture	Norms, values, and basic assumptions of the juvenile justice agency.	1. Staff at all levels openly talk about what is and isn’t working. 2. Most staff in my agency are willing to change how they do things in response to feedback from others. 3. It is hard to get things to change in our agency.* 4. I can rely on the other people to do their jobs well. 5. Most of the people who work in our agency seem to enjoy their work. 6. Difficult problems are solved through face-to-face discussions. 7. Staff regularly take time to reflect on how we do things. 8. After trying something new, we take time to think about how it worked. 9. People in this agency operate as a real team.		0.88	Fernandez et al., 2018
Inner Setting: Implementation Climate	The extent to which family-driven care will be supported and expected within their juvenile justice agency.	1. Individuals at my agency would approve of family-driven care. 2. Individuals at my agency value new types of family engagement practices. 3. Individuals at my agency are open to new types of family engagement practices. 4. Individuals working at my agency are flexible enough to integrate new family engagement practices.		0.89	Fernandez et al., 2018
Inner Setting: Compatibility	How family-driven care fits with existing workflows	1. The juvenile justice system should only provide care to youth (not families).* 2. Family-driven care is well-suited to the juvenile justice system. 3. Family-driven care is compatible with the needs of families and youth at my agency.		0.73	Sales et al., 2021 and Kegler et al., 2018
Inner Setting: Relative Priority	Staff’s shared perception of the importance of the family engagement.	1. Family engagement is a top priority at my agency.		N/A	Fernandez et al., 2018
Inner Setting: Access to Knowledge/Information	Access to information and knowledge about family engagement at the juvenile justice agency.	1. Individuals in my agency have received sufficient training on family engagement practices.		N/A	Sales et al., 2021
Inner Setting: Leadership Engagement	Commitment, involvement, and accountability of leaders and managers to family engagement.	1. Management at my agency would be supportive of family-driven care.		N/A	Sales et al., 2021
Inner Setting: Available Resources	The level of resources that can be dedicated to family-driven care in the agency.	1. We have enough staffing to make family-driven care work. 2. We have sufficient financial resources to make family-driven care work. 3. We have enough physical space to make family-driven care work. 4. We have sufficient training to make family-driven care work.		0.73	Fernandez et al., 2018
Outer Setting: External Partnerships	The degree to which the agency is networked with other external organizations that support family engagement.	1. Individuals in my agency are connected with external organizations that provide family services and supports.		N/A	Sales et al., 2021
Outer Setting: Peer Pressure	Competitive or mimetic pressure to implement family engagement initiatives.	1. Other juvenile justice agencies will be adopting new family engagement practices in the next year. 2. Agencies that promote family engagement practices are seen as leaders in juvenile justice. 3. Agencies that promote family engagement practices are seen are held in high esteem.		0.75	Sales et al., 2021
			Outcomes
Outcome 1: Adoption of Family-Driven Principles	Staff’s perception of their facility’s alignment with the 10 family-driven principles.	In my agency, we … 1. Families and staff embrace the concept of sharing decision-making and responsibility for outcomes. 2. Families are given accurate, understandable, and complete information necessary to set goals and make informed decisions about the right services and supports for their children. 3. All youth have a biological, adoptive, foster, or surrogate family voice advocating on their behalf. 4. Families are provided opportunities to engage in peer support activities and connect with other parents/guardians of justice-involved youth. 5. Staff advocate for the needs and preferences of families and youth. 6. Families provide direction on funding, policies, and service options. 7. Management allocates staff, training, support and resources to promote family engagement practices. 8. Staff work to remove family barriers to engagement and participation. 9. Staff embrace, value, and celebrate the diverse cultures of their youth and families. 10. Staff continually advance their own cultural and linguistic responsiveness, so that the needs of all families are appropriately addressed.		0.89	National Federation of Families for Children’s Mental Health, principles of family-driven care
Outcome 2: Adoption of Family-Driven Strategies	Staff’s perception of their facility’s current level of implementation of family engagement strategies.	In my agency, we … 1. Systematically identify members of each youths family unit. 2. Involve families in treatment planning (e.g., family-group decision making or family-group conferencing). 3. Encourage families to provide formal feedback on system processes (e.g., through family surveys and/or family town halls). 4. Invite family representatives to serve on advisory boards or policy-making committees. 5. Provide opportunities for family members to participate in support groups. 6. Educate family members on the procedures and policies of the juvenile justice system. 7. Provide parenting skills programs. 8. Refer to parenting skills programs. 9. Provide family-based mental health services (e.g., family counseling, therapy, mental health treatment). 10. Refer to family-based mental health services (e.g., family counseling, therapy, mental health treatment). 11. Provide family-based substance use services (e.g., family substance use prevention or treatment). 12. Refer to family-based substance use services (e.g., family substance use prevention or treatment). 13. Provide family-based HIV/STI services (e.g., family-based HIV/STI prevention or treatment). 14. Refer to family-based HIV/STI services (e.g., family-based HIV/STI prevention or treatment). 15. Receive formal training on diversity and inclusion (e.g., cultural humility training, racial sensitivity training, and/or unconscious bias training). 16. Provide flexible scheduling to accommodate families. 17. Assist families with transportation needs. 18. Assist families with childcare needs.		0.92	Adapted from Family System Engagement Index (Robertson et al., 2019)

* Indicates that the item was reverse coded

Survey primary predictors

Primary predictors included selected implementation-focused determinants from the Consolidated Framework for Implementation Research (CFIR) (Damschroder et al., 2009). CFIR provides a menu of 39 constructs that can be used as a practical guide for systematically assessing potential barriers and facilitators in preparation for implementing a new intervention, such as family-driven care (FDC). CFIR is a multi-level framework that assesses implementation determinants across multiple domains including inner setting (e.g., agency) factors, outer setting (e.g., societal and community) factors, individual (e.g., staff) characteristics, and characteristics of the intervention (e.g., FDC). Because CFIR has 39 constructs, it is challenging (and impractical) to apply all constructs in a single study, so evaluations typically focus on a subset of CFIR constructs. Damschroder and colleagues (2009) recommended selecting constructs based on their likelihood of being a potential barrier (or facilitator) to implementation and/or having sufficient variation across the units of analysis (e.g., JJ agencies). Ultimately, we selected 13 CFIR constructs based on conversations with justice system professionals and based on a thorough literature review of potential determinants of FDC implementation in various child-serving settings. Specifically, we measured inner setting factors including climate, culture, compatibility, priority, access to knowledge, leadership engagement, and resources. Outer setting factors included were external partnerships and peer pressure. We also measured individual characteristics, including attitudes, and characteristics of the intervention, including relative advantage, cost, and complexity. CFIR constructs were measured using scales applied in prior studies (Ehrhart et al., 2014; Helfrich et al., 2009; Fernandez et al., 2018; Sales et al., 2021; Kegler et al., 2018). We adapted items from these scales to fit terminology used by GDJJ, so they would be understood by staff in this JJ system context. See Table 1 for details on CFIR construct definitions, items, and Cronbach’s alphas. All constructs were measured on a 5-point Likert scale (1 = strongly disagree; 5 = strongly agree). Overall scores were calculated by averaging the items. JJ stakeholders reviewed and provided feedback on the measures before survey administration.

Survey control variables

Participant demographics included age, gender, race, ethnicity, education level, role, years of experience, and caseload. Organizational characteristics included agency type (i.e., detention, community supervision, administrative). Agency catchment area characteristics included county poverty rate, insured rate, and racial/ethnic composition (U.S. Census Bureau, 2020). We also categorized agency catchment area as urban (i.e., large central metro, large fringe metro, or medium metro) or rural (i.e., small metro, micropolitan, noncore) based on 2014 urban-rural classifications from the National Center for Health Statistics (NCHS) (Ingram & Franco, 2014).

Qualitative interviews

Semi-structured interviews evaluated barriers and facilitators to integrating FDC into JJ systems. The guide was developed using implementation-focused constructs from CFIR based on guidance on CFIR website- www.cfirguide.org (Damschroder et al., 2009); specifically focusing on constructs that were identified as significant in the survey analysis. Questions explored staff perceptions of inner setting and outer setting contextual factors that may influence implementation of FDC. The questions also covered staff-perceived barriers to family engagement, staff attitudes towards working with families, and their perceptions of the FDC framework. Before beginning interviews, JJ stakeholders reviewed and provided feedback on the guide. The interviews, which ranged from 25 to 45 min, were conducted by female qualitatively trained (graduate level) researchers (KP and AJ), audio recorded, and transcribed verbatim.

Analysis

Quantitative survey analysis

Hierarchical linear modeling was used to analyze the survey data. A 2-level hierarchical model with a random intercept was used to account for the correlation among respondents from the same agency. Each agency had one to seven participants who participated in the survey. The analyses were performed using SAS 9.4, and significance was considered at p < 0.05.

Implementation climate was highly correlated with many CFIR constructs; therefore, it was excluded from model building, due to issues of multicollinearity. Missing data ranged from 0 to 11 responses per variable, and multiple imputation was employed using full conditional specifications based on the remaining survey items.

For the hierarchical linear regression, level 1 (i.e., individual level) variables included CFIR factors and demographic characteristics, and level 2 (i.e., organizational level) variables included agency characteristics. The following analyses were conducted for each outcome (i.e., FDC principles and FDC strategies). First, we ran the unconditional means model, calculating the ICC to assess organizational-level variation the outcome. The subsequent model included only CFIR factors (primary predictors) with a p-value < 0.25 in the bivariate analysis. In the next model, we added CFIR factors along with control variables (individual demographic and agency characteristics) that had a p-value < 0.25 in the bivariate analysis. Lastly, we simplified the previous model by removing insignificant factors (p-value > 0.05). While all models are reported in the results, the parsimonious model was prioritized for data interpretation. Multilevel analyses employed restricted maximum likelihood estimation (REML).

Qualitative and mixed methods analysis

Using MAXQDA v22.4.1, we applied standard qualitative data analysis techniques, including transcript reading, codebook creation, coding, and consensus meetings (Hennink et al., 2011). The codebook was deductively developed, drawing from CFIR constructs, and code definitions and inclusion/exclusion criteria were adapted from guidance on the CFIR website (www.cfirguide.org). Two analysts individually coded transcripts, meeting biweekly to resolve coding disagreements through discussion. To organize findings, memos summarizing key themes for each CFIR factor were generated based on analyst discussions. In this analysis, we focused on CFIR factors that were significant in regression analyses, to explain survey relationships. Results were shared with JJ stakeholders to ensure proper interpretation.

Results

Survey results

Overall, 140 JJ employees from 61 different agencies in Georgia participated in the survey. On average, participants were 47 years old, most were female (71.1%), and most identified as Black/African American (57.5%) and non-Hispanic (95.1%). Approximately, 37% held a graduate degree and 45% worked at their agency for more than 10 years. Participants held roles including line staff (e.g., correctional officers and probation/parole officers) (n = 50, 35.7%), behavioral and social service providers (n = 26, 18.6%), educators (n = 26, 18.6%), case managers and reentry planning (n = 20, 14.3%), and leadership (n = 18, 12.9%). Participants worked within community supervision (73, 52.1%), detention (48, 34.2%), and administrative (19, 13.6%) settings, which were in both urban (56.4%) and rural locations (43.6%). In unadjusted bivariate analyses, most CFIR constructs were significantly correlated with both outcomes (Table 2).

Table 2.

Survey Participant descriptive characteristics and Bivariate statistics (n = 140)

		Outcome 1: Family-Driven Principles	Outcome 2: Family-Driven Strategies
	Mean (SD) or Count (%)	Unadjusted β (SE) a	Unadjusted β (SE) a
CFIR Constructs
Intervention Characteristics: Relative Advantage	3.57 (0.83)	0.03 (0.07)	0.11 (0.08)
Intervention Characteristics: Complexity	3.07 (0.85)	-0.25 (0.07)***	-0.20 (0.07)**
Intervention Characteristics: Cost	2.70 (0.77)	-0.03 (0.08)	0.04 (0.08)
Individual Characteristics: Negative Attitudes	2.66 (0.63)	-0.33 (0.09)***	-0.27 (0.09)**
Inner Setting: Culture	3.33 (0.71)	0.54 (0.07)***	0.37 (0.08)***
Inner Setting: Implementation Climate	3.80 (0.70)	0.65 (0.06)***	0.49 (0.07)***
Inner Setting: Compatibility	3.87 (0.69)	0.29 (0.08)***	0.26 (0.09)**
Inner Setting: Relative Priority	3.73 (0.88)	0.49 (0.05)***	0.43 (0.05)***
Inner Setting: Access to Knowledge/Information	3.12 (0.90)	0.39 (0.06)***	0.30 (0.07)***
Inner Setting: Leadership Engagement	3.95 (0.72)	0.47 (0.07)***	0.43 (0.08)***
Inner Setting: Available Resources	2.71 (0.72)	0.36 (0.08)***	0.31 (0.08)***
Outer Setting: External Partnerships	3.71 (0.99)	0.31 (0.05)***	0.29 (0.06)***
Outer Setting: Peer Pressure	3.55 (0.59)	0.16 (0.10)	0.12 (0.11)
Outcomes
Outcome 1: Family-Driven Principles	3.73 (0.65)	-	-
Outcome 2: Family-Driven Strategies	3.44 (0.69)	0.70 (0.06)***	-
Demographic Characteristics
Age	47.37 (9.95)	-0.01 (0.01)	-0.001 (0.006)
Gender
Male	30 (23.4%)	Ref	Ref
Female	91 (71.1%)	-0.08 (0.14)	0.03 (0.14)
Race
Black/African American	69 (57.5%)	-0.13 (0.12)	0.11 (0.13)
White/ Caucasian	51 (42.5%)	Ref	Ref
Ethnicity
Hispanic/Latinx `	6 (4.9%)	-	-
Not Hispanic/Latinx	117 (95.1%)	-	-
Education Level
No Graduate Degree	79 (56.4%)	Ref	Ref
Graduate Degree	47 (37.3%)	-0.13 (0.12)	-0.21 (0.13)
Caseload	13.85 (20.13)	0.005 (0.003)	0.003 (0.003)
Years with Agency
<5 years	38 (27.3%)	-0.04 (0.14)	0.002 (0.14)
Between 5–10 years	38 (27.3%)	-0.24 (0.14)	-0.18 (0.14)
10 + years	63 (45.3%)	Ref	Ref
Role
Leadership	18 (12.9%)	-0.26 (0.17)	-0.15 (0.20)
Behavioral/Social Provider	26 (18.6%)	0.002 (0.17)	-0.21 (0.18)
Case Manager	20 (14.3%)	0.08 (0.18)	-0.15 (0.18)
Educator	26 (18.6%)	-0.01 (0.16)	-0.30 (0.18)
Line Staff	50 (35.7%)	Ref	Ref
Agency Characteristics
Agency Type
Administrative	19 (13.6%)	-0.19 (0.17)	-0.07 (0.27)
Detention	48 (34.3%)	-0.09 (0.13)	-0.20 (0.14)
Community Supervision	73 (52.1%)	Ref	Ref
Urbanicity
Urban	79 (56.4%)	Ref	Ref
Rural	61 (43.6%)	0.05 (0.12)	0.02 (0.13)
% Poverty rate in county	17.6 (5.76)	0.001 (0.01)	0.001 (0.01)
% Uninsured in county	16.0 (2.17)	0.02 (0.03)	0.02 (0.03)
% White/Caucasian population in county	51.4 (18.5)	0.004 (0.003)	0.002 (0.004)
% Hispanic population in county	7.0 (4.7)	-0.007 (0.01)	0.005 (0.01)
% Black/African American population in county	37.6 (18.0)	-0.003 (0.003)	-0.003 (0.004)

Note: *<0.05, **<0.01, ***<0.001

^a Estimates based on Generalized Linear Mixed Model with random intercept, unadjusted for other variables

Outcome 1: implementation of FDC principles

In the parsimonious model (Table 3), factors significantly related to implementation of FDC principles included relative priority (β = 0.21, SE = 0.06, p < 0.001), culture (β = 0.22, SE = 0.07, p < 0.001), access to knowledge and information (β = 0.13, SE = 0.05, p < 0.05), leadership engagement (β = 0.21, SE = 0.06, p < 0.001), complexity (β=-0.10, SE = 0.05, p < 0.05), and caseload (β = 0.004, SE = 0.001, p < 0.05). Based on the unconditional means model (not shown), the ICC is 0.03, indicating that about 3% of the variation in the outcome is at the agency level.

Table 3.

Model building for outcome 1: family-driven Care principles (n = 140)

	Model 1 (Primary Predictors)	Model 2 (Full Model)	Model 3 (Reduced) b	Model 4 (Parsimonious) c
Fixed Effects a	β (SE)	β (SE)	β (SE)	β (SE)
Complexity	-0.11 (0.05)*	-0.10 (0.07)*	-0.11 (0.05)*	-0.10 (0.05)*
Negative Attitudes	-0.04 (0.07)	-0.03 (0.07)
Culture	0.22 (0.07)**	0.21 (0.07)**	0.22 (0.07)***	0.22 (0.07)***
Compatibility	-0.01 (0.07)	0.01 (0.07)
Relative Priority	0.18 (0.06)**	0.17 (0.06)**	0.18 (0.06)**	0.21 (0.06)***
Access to Knowledge	0.9 (0.05)	0.11 (0.06)	0.12 (0.05)*	0.13 (0.05)*
Leadership Engagement	0.17 (0.07)*	0.16 (0.07)*	0.17 (0.06)**	0.21 (0.06)***
Available Resources	0.05 (0.06)	0.05 (0.06)
External Partnerships	0.09 (0.04)	0.09 (0.05)	0.09 (0.04)
Peer Pressure	0.08 (0.07)	0.11 (0.07)	0.12 (0.07)
Caseload		0.004 (0.002)*	0.004 (0.002)*	0.004 (0.002)*
Years with Agency
<5 years		0.14 (0.10)	0.14 (0.09)
Between 5–10 years		-0.15 (0.09)	-0.14 (0.09)
10 + years		Ref	Ref

Note: *<0.05, **<0.01, ***<0.001; Estimates based on Generalized Linear Mixed Model with random intercept, adjusted for other variables in the model

^a Variables with p < 0.25 in the unadjusted models (Table 2), were included in the model-building process

^b Removed variables with p > 0.25 from full model

^c Only includes significant variables

ICC: about 3% of the variability in the outcome is at the agency level

Outcome 2: implementation of FDC strategies

In the parsimonious model (Table 4), factors significantly related to implementation of FDC strategies, included relative priority (β = 0.31, SE = 0.06, p < 0.001), leadership engagement (β = 0.22, SE = 0.08, p < 0.01), and external partnerships (β = 0.12, SE = 0.06, p < 0.05). Participant role was also statistically significant, with case managers being least likely to perceive implementation of FDC strategies (Table 4). Based on the unconditional means model (not shown), the ICC is 0.10, indicating that about 10% of the variation in the outcome is at the agency level.

Table 4.

Model building for outcome 2: family-driven strategies (N = 140)

	Model 1 (Primary Predictors)	Model 2 (Full Model)	Model 3 (Reduced b)	Model 3 (Parsimonious c)
Fixed Effects a	β (SE)	β (SE)	β (SE)
Relative Advantage	0.02 (0.07)	0.003 (0.07)
Complexity	-0.11 (0.06)	-0.08 (0.06)	-0.10 (0.06)
Negative Attitudes	-0.001 (0.09)	-0.03 (0.10)
Culture	0.01 (0.08)	0.01 (0.08)
Compatibility	0.03 (0.08)	0.07 (0.09)
Relative Priority	0.25 (0.08)**	0.25 (0.08)**	0.28 (0.06)***	0.31 (0.06)***
Access to Knowledge	0.01 (0.07)	0.05 (0.08)
Leadership Engagement	0.19 (0.09)*	0.21 (0.09)*	0.22 (0.08)**	0.22 (0.08)**
Available Resources	0.03 (0.08)	0.02 (0.08)
External Partnerships	0.12 (0.06)	0.10 (0.06)	0.12 (0.06)*	0.12 (0.06)*
Graduate Education		0.001 (0.12)
Years with Agency
<5 years		0.10 (0.13)
Between 5–10 years		-0.13 (0.12)
10 + years		Ref
Role
Leadership		-0.28 (0.19)	-0.21 (0.16)	-0.23 (0.16)
Behavioral/Social Provider		-0.31 (0.18)	-0.23 (0.14	-0.23 (0.14)
Case Manager		-0.20 (0.15)*	-0.31 (0.14)*	-0.35 (0.14)*
Educator		-0.41 (0.23)	-0.23 (0.14)	-0.24 (0.14)
Line Staff		Ref	Ref	Ref
Agency Type
Administrative		0.06 (0.29)
Detention		0.18 (0.21)
Community Supervision		Ref

Note: *<0.05, **<0.01, ***<0.001; Estimates based on Generalized Linear Mixed Model with random intercept, adjusted for other variables in the model

^a Variables with p < 0.25 in the unadjusted models (Table 2), were included in the model-building process

^b Removed variables with p > 0.25 from full model

^c Only includes significant variables at p < 0.05 level

ICC: 10% of the variability in the outcome is at the agency level

Qualitative interview results

In total, 16 JJ employees (7 leaders and 9 staff representing reentry services, community services, detention, behavioral health, and education divisions) from 10 unique agencies in Georgia participated in follow-up interviews. Significant implementation determinants identified in the survey covered multiple CFIR domains including Intervention Characteristics (i.e., complexity), Inner Setting Factors (i.e., culture, priority, access to knowledge and information, and leadership engagement), and Outer Setting Factors (e.g., external partnerships). To explain how these significant CFIR constructs operated to facilitate/hinder adoption of FDC, qualitative results for each construct are summarized below.

Complexity

Many participants described family engagement as a complicated process involving cooperation between numerous stakeholders including parents, youth, community-based organizations, and the justice system:

Working with the juvenile has not always been the issue…It’s more so that there’s a lack of collaboration with the different stakeholders, to try to get the parents’ buy in, or get the parents to participate in that child’s treatment, to make sure that they’re rehabilitated. A lot of times, like I said before, we’re able to hold the juveniles accountable for not attending or participating in treatment, but we know that a lot of times, that the parents’ support and buy in is key. (ID:#1, Leadership, Probation)

Participants believed that although services and programs are available for families, gaining families’ cooperation, trust, and willingness to participate in services was an obstacle outside of their control:

Sometimes parental engagement it’s not always there. We’ll have the kid that wants to do the right thing, but they don’t have the resources or support they need because the parent is unable to provide or to be honest, in some cases, willing to assist. The improvements that I would like to see with family engagement are not necessarily things that are controllable by the department. I think that if we could have more parental engagement or be able to hold sometimes the parents more accountable than we are, then that would increase that tenfold. Because like I said, we’ll have the kid that is willing to participate, but sometimes it’s not always possible to get the parents to do their part. (ID: #5, Staff, Community Services)

In addition to lack of stakeholder cooperation, participants also acknowledged the numerous obstacles that made FDC challenging in the juvenile justice setting, including service accessibility, transportation constraints, and competing family responsibilities. Ultimately, engagement varies based on each family’s unique circumstance, and one participant described not knowing what the “magic thing” is to make family engagement work in the juvenile justice setting:

How to engage the families is always a trick question. Do you engage them with offering incentives, do you engage them if it’s on the weekend, at night? And we know that they’re faced with dealing with their life issues. They may be busy trying to make ends meet and put food on the table. They may be dealing with a lot of different things that trump what we are trying to accomplish, so we just don’t know what that magic thing is. (ID: #2, Leadership, Reentry Services)

Culture

JJ Leaders discussed how they worked to change the culture of the department- from a focus on incarceration to a focus on treating the underlying causes of justice involvement through BH care and family-driven services:

We knew we had to change the culture of an organization from just incarcerate – You know, incapacitating you and holding them and to really making sure that we were concentrating on also trying to provide service with an end state in mind of doing what we can to help that youth improve their outcome and hopefully not recidivate. Staff have to care about who they serve – providing services to and making sure that it’s not just going through the motions, making sure that it’s looking at things with the end in mind and also making sure whatever we do, we understand why we are doing it. And so just making sure that our culture and staff to have empathy and care about the people they’re serving, because it is a service. (ID: #15, Leadership, Reentry Services)

This culture change was embraced by many JJ agencies across the state, and staff discussed the passion, care, and empathy shown by colleagues and the ways they go above and beyond to provide care to families and youth: I’ve seen it from the years I’ve been here, everybody that’s in this field has a passion for it. It’s not for the money. We do it because we love working with the kids and their family. (ID: #1, leadership, Community Services)

Priority

Both leadership and staff overwhelmingly described family engagement as their top priority because it is integral to their ability to help youth succeed. Setting this priority changed the culture across agencies in Georgia and promoted family engagement programming, as described by the following participants:

Family engagement is such a priority that it is one of our major platforms on our mission statement. So very heavily such because we serve kids. You cannot serve kids and not involve the parent and be successful. It has to be an integral part of what we do. (ID: #6, Leadership, Behavioral Health)

Family engagement is the number one priority. We always want to include the families and let them know what’s going on in their in their child’s life, one step at a time and letting them know and making them aware … because it’s very important, they are youth, they realize that they make mistakes, that’s the reason why they’re here. But the number one reason is to keep family involved, keep it a positive atmosphere, and not judgmental, but helping them through the process to just become better, stronger. (ID: #9, Staff, Education)

Leadership engagement

Participants discussed how their leaders, especially executive leaders, were highly committed to and supportive of family engagement and are actively leading innovation across the state:

Our current [executive leadership] recognizes that we stand at the intersection of law enforcement, but also child’s welfare and is trying to effect a culture change that supports that understanding. [Family engagement] is a priority for our executive leadership, a high priority. It’s a priority for my office. The degree to which that is filtering down is definitely connected to agency effects at culture change. So it’s an area of growth for us, but our leadership has identified that as growth that they wish to make, that we need to make. (ID: #7, Leadership, Behavioral Health)

Staff also discussed how they and their local supervisors are committed to leadership’s mission to improving family engagement on all levels:

My [leaders and supervisors]- all of those people are very keen with us engaging with families… I would say everyone is leading the change, because [leaders are] asking us to do activities, we have a log that we are accountable for, and so that shows that our leaders want us to implement different strategies. And so of course on our level, we’re going to implement them. So I think everyone is on board with it. (ID: #14, Staff, Education)

Access to knowledge and information

JJ leaders discussed their constant search for new information and training on family engagement. Though trainings related to family engagement are limited, leadership tried to make all educational opportunities available to staff across the state to promote dissemination of family-focused programs:

I know there are [need for training], because I don’t believe we ever know it all and we ever got everything we need. We took advantage of really exposing our staff throughout the system to a lot of webinars and education… So we’re constantly not just assessing what’s out there, but we’re also trying to make sure we provide those things that we have, that we can give to the general public, as well as the other staff throughout the agency. (ID: #15, Leadership, Reentry)

However, JJ staff were unaware if they received family engagement-specific trainings, and they requested more widespread training on family engagement across agencies: [We need] more training and more awareness and getting to know the family and their needs and every child where they come from. (ID: #9, Staff, Education)

External partnerships

Participants discussed the importance of external partnerships- including behavioral health providers, social service providers, and schools- in delivering family driven services. Because JJ sites are unable to provide all family services on site, external partners are critical to fulfilling their family engagement mission; though, external partnerships vary based on location (e.g., rural versus urban) and availability within communities:

We’re not only trying to hold [families] accountable, but we are trying to return them as law abiding and productive citizens. We try to surround them with as many opportunities, as many resources, as many services as we can to help them. We’re always trying to figure out what’s going on with the family. Our probation officers going to the schools, they meet with the counselors, the teachers. You know, we make referrals for mental health services. We have other resources that are pro-social, not necessarily in terms of treatment…that we can refer, tap into. We’re constantly trying to seek partners that will help bring some services in. (ID: #2, Leadership, Reentry Services)

Discussion

Youth involved in the juvenile justice (JJ) system are particularly at-risk BH conditions, yet very few initiate and engage in treatment services. Families play a critical role in mitigating this need-treatment gap among JIY, and engaging families in JIY’s BH services is one of the most effective strategies to promote treatment initiation and long-term positive BH outcomes. However, there is minimal context-specific guidance integrating family engagement frameworks (like FDC) into BH care in the JJ setting. To fill this gap, we conducted surveys and interviews with JJ staff and leadership in the state of Georgia to understand multi-level implementation determinants of FDC principles and strategies. Specifically, we assessed implementation determinants across multiple levels of CFIR including individual (i.e., staff) characteristics, intervention (i.e., FDC) characteristics, inner setting (e.g., agency) factors, and outer setting (e.g., community/societal) factors. Based on a mixed methods analysis, JJ agencies were more likely to implement FDC principles and strategies if they had the following characteristics: (1) presence of leaders that were strongly committed to family engagement, (2) a shared understanding that family engagement was a top priority, (3) staff training and education related to family engagement, (4) a workplace culture that was caring towards families and supportive of innovation, (5) presence of family engagement programs that were easier (or more feasible) for staff to implement, and (6) external partnerships with organizations that serve families. This study underscores the importance of strengthening these 6 determinants when preparing to implement FDC in JJ systems.

Leadership engagement and relative priority were two CFIR inner setting implementation determinants that were highly salient in our findings in both quantitative and qualitative data. These determinants were consistently significant across regression models and significant for both outcomes (e.g., implementation of FDC principles and strategies). They were also consistently endorsed by all interview participants as key factors that facilitated the uptake of family engagement programming in JJ agencies across the state. Participants discussed how JJ leaders were instrumental in creating a culture that was accepting, nonjudgmental, and supportive of families. Leaders also disseminated information on family engagement strategies, set priorities around family engagement, and modeled the principles of FDC in their interactions with families. These findings are consistent with literature identifying the critical role of leaders in promoting organizational change, gaining employee acceptance of innovations, and changing organizational culture (Thoms, 1996). Prior interventions within the JJ setting have also noted how leadership commitment and priority setting are critical first steps to implementing and disseminating BH interventions for justice-involved kids and families (Elwyn et al., 2017; Fernandez et al., 2018; Jolivette & Nelson, 2010; Rocque et al., 2014). JJ systems that are adopting FDC frameworks should not only consider executive leadership buy-in, but also incentivize and reward site-level leaders and ensure measurable family engagement priorities and objectives are implemented across all detention and community supervision agencies.

Along with leadership engagement and priorities, agency culture was another inner setting factor related to the implementation of FDC principles. Agencies that were more accepting of innovations, had better teamwork and staff cohesion, and were passionate about working with youth and families were more likely to implement principles. Many interview participants discussed how a culture of empathy and care was necessary for this intervention to be successful in the JJ setting. Although culture change is notoriously challenging (especially in bureaucratic systems) (Kanter, 2003), participants discussed how JJ leaders changed the culture of the system over the past decade. Specifically, they implemented state-level policy reforms, which transformed the system from focusing on incarceration to now focusing on promoting youth and family long-term success. This is consistent with JJ systems across the U.S. that reformed their agencies to address the underlying causes of delinquency (e.g., behavioral health, family functioning) through linkages to community-located treatment, prevention services, and diversion programs (Nellis et al., 2009; Prison Policy Initiative, 2019). Findings from our surveys and interviews suggest this culture change (from a punitive to a healing focus) was an important precursor and facilitator of the adoption of FDC principles and strategies. This aligns with prior research that identified a relationship between positive organizational culture and higher quality services and more effective outcomes for children and families involved in child welfare systems (Glisson & Green, 2011; Glisson & Hemmelgarn, 1998). In mental health and social service organizations, a positive culture affects whether the most effective and innovative service protocols are adopted, if they are implemented with fidelity, and whether they are sustained (Glisson & Green, 2011; Hoagwood et al., 2001).

Another important determinant of implementation of FDC principles was access to knowledge and information (e.g., staff training and education) on family engagement. Participants identified family engagement training as a current gap. Although leaders are attending conferences and webinars that discussed the importance of family engagement; there are no widespread family engagement trainings available and/or mandatory for staff across all agencies. Although family engagement trainings are available for other child-serving systems such as healthcare and schools (Kovacs Burns et al., 2014; Smith & Sheridan, 2019), trainings for the JJ setting are limited. Development of training and skills development programs are a critical next step for advancing family engagement in JJ systems across the country. Additionally, our study identified that perceived implementation of FDC strategies and principles varied by staff role and caseload, indicating that awareness and utilization of strategies may differ based on role (e.g., line staff, behavioral health provider, case manager). Not only does education on family engagement need to be disseminated to staff across all divisions, but training may also need to be tailored to staff’s unique roles and responsibilities.

In addition, participants who perceived family engagement as being complex were less likely to implement FDC principles. Participants may think FDC is complex because the components and operationalization of the intervention are not well specified or well understood by staff. Training and education may help reduce perceive complexity by increasing staffs’ skills and capability to perform family engagement strategies. Interviews also revealed that staff perceived family engagement as being complex because there were numerous barriers to working with families including lack of family willingness to engage as well as other logistical barriers such as family availability, scheduling conflicts, transportation, and competing family responsibilities. Other studies have documented similar barriers: in a survey of justice-involved families, the most common barriers to engagement included transportation (42%), distance to treatment centers/facilities (41%), time (37%), and cost (35%) (Justice for Families, 2012a). Families may also choose not to participate in services due family’s feelings of stigma, shame, and frustration with their child’s behavior, minimization of their families’ health needs, and families’ stress, trauma, or other behavioral health conditions that pose obstacles to engagement (Hock et al., 2015; Tambling et al., 2022). Therefore, staff training on how to help families overcome barriers may help agencies and their staff navigate this complexity. For instance, in one study, adolescent mental health providers were trained to support families’ ability to initiate treatment by addressing transportation, financial, scheduling concerns, and other barriers. They worked with families to develop a plan to overcome barriers prior to the start of services, which led to higher therapy initiation and engagement compared to the control group (McKay et al., 1996).

Lastly, the presence of external partnerships was related to the implementation of FDC strategies. JJ systems often rely on partnerships with external organizations to provide appropriate medical and social services for youth and families, which also increases the complexity of service provision in the JJ setting. Improving the delivery of family-driven services may rely on enhancing cross-system collaboration and linkages with organizations that extend the family engagement mission of the JJ system (Cocozza & Skowyra, 2000). Although research aimed at increasing cross-system collaboration between JJ agencies and community organizations is lacking, studies have shown that use of local change teams (or interagency workgroups) can successfully improve collaborations between agencies, formalize partnership roles and processes, and improving interagency communication (Belenko et al., 2013; Bowser et al., 2018; Hoffman et al., 2012; Huang et al., 2020; Hurlburt et al., 2014). These processes can also help to mitigate family barriers to care by streamlining linkages and referrals between agencies, and future studies should develop and evaluate strategies to improve these cross-sector linkages.

Specifically in Georgia, next steps include developing a formal implementation plan to improve FDC adoption across JJ staff and agencies in the state. Although prioritization of FDC and leadership buy-in are strong at the state-level, engaging site-level leaders and champions is an important next step to ensure consistent implementation across agencies. Because many of the implementation determinants (e.g., leadership engagement, priority, and culture) assessed in this study were generally strong, we will prioritize developing implementation strategies to address weaker determinants. For instance, addressing access to knowledge and information through provision of family engagement education/skills trainings may be a feasible next step to improve implementation of FDC principles and strategies. Ensuring trainings are accessible to all staff roles and all divisions may help improve standardization of implementation across agencies. To address concerns about FDC complexity, trainings should also be designed to help JJ staff and practitioners navigate family barriers (e.g., logistical challenges and lack of willingness to engage) and interpersonal barriers (e.g., staff-family trust and communication) to engagement. In addition, development of measurable family engagement metrics for JJ systems is an important next step, to not only evaluate the effectiveness of trainings, but to also evaluate the effectiveness of new family engagement policies, initiatives, and practices. Longer-term implementation strategies should focus on building eternal partnerships with family-serving organizations that can provide family programs and services.

Limitations of this study include the utilization of non-probability sampling for both the surveys and interviews. JJ staff and leaders who volunteered for the survey may be more attuned to family engagement issues than those who did not participate. Additionally, the department employees over 3,400 individuals, but we only captured perspectives from 140 JJ staff and leaders, so findings may not represent the larger population of JJ employees in Georgia. However, our survey captured FDC adoption data from over 60 agencies across Georgia (including detention and community services settings) as well as data from employees from different sectors of the JJ system, including probation, education, behavioral health, reentry services, and leadership, which increased the diversity of perspectives included in this assessment. Our mixed methods approach further strengthened the findings, since there was strong cross-validation between the survey and interview results. In addition, because the contexts that shape the delivery of family engagement strategies and behavioral health services vary across jurisdictions, results may not apply to other states with different policies and organizational structures. Another key gap of this assessment is the lack of data and perspectives from families and youth involved in the system. Among the limited literature on this topic, there is a paucity of research that centers the voices of justice-involved families to understand their needs and recommendations for improving relationships between home and the JJ system.

Conclusions

In conclusion, involving families in service and organizational decisions can improve the long-term outcomes of justice-involved youth. Historically, JJ systems have not been inclusive of family voices; however, advocacy efforts from families and justice reform organizations have increased the prioritization of family engagement among JJ systems within the last decade. Scaling-up FDC principles and strategies across agencies requires attention to unique contextual factors including local leadership involvement, priorities, staff training and education, external partnerships, workplace culture, and program complexity. JJ systems that are transitioning to FDC should consider strengthening these inner and outer setting contextual factors to promote uptake. Findings from this mixed methods assessment will inform scale-up of family-driven programs in Georgia and may inform implementation of this high priority intervention in systems nationally.

Abbreviations

BH

Behavioral Health

CFIR

Consolidated Framework for Implementation Research

FDC

Family-Driven Care

FDC

Georgia Department of Juvenile Justice

JIY

Justice-Involved Youth

JJ

Juvenile Justice

Author contributions

KP, BWJ, CE, and JS conceptualized and designed the study and developed data collection instruments. DS and CSC led participant recruitment and provided feedback on data collection materials to ensure relevance to study population. KP and AJ surveyed and interviewed participants. KP and AJ coded and analyzed the data. All authors were involved in interpretation and dissemination of study findings. KP wrote the initial draft of this paper. All authors reviewed, edited, and approved the final manuscript.

Funding

This study was supported by the National Institutes of Health/National Institute on Drug Abuse F31DA053005 (PI: KP) and R25DA037190 subaward (PI: KP). The contents of this publication are solely the responsibility of the authors and do not necessarily represent the official views of the NIDA, NIH, or the participating universities or JJ systems.

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

The university-affiliated ethical review board (IRB ID: STUDY00002068) and the juvenile justice system research review committee approved this study. All participants provided consent before participating in study activities.

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.