The Pilot Test of a Cross-Systems Behavioral Health Treatment Referral and Linkage Intervention for Youth on Probation

Abstract

Despite high rates of substance use, youth involved in the juvenile justice system are unlikely to be linked to the treatment services they need. Family Connect is a flexible, family-focused, linkage intervention developed to address multilevel barriers and increase youth engagement in care through the introduction of a linkage specialist. We describe the components of Family Connect and present findings from the intervention pilot test comparing 18 youth–caregiver dyads to 95 historical controls on referral, attending intake and initiating treatment. Results indicated preliminary support for Family Connect as an approach to increase cross-systems linkage and access to behavioral health care. Findings also suggested support for the feasibility of the intervention and indicated that justice-involved youth and their caregivers found the intervention to be acceptable. In addition to discussing our findings in the context of recent justice reforms, and the importance of improving access to treatment, we make recommendations to inform a future trial of Family Connect.

Youth involved in the juvenile justice system (YIJ) have substantially higher rates of substance use (SU) and disorders (SUD) compared with youth in the general population. In the United States, 28% of youth report any past-year SU, and over 10% have a SUD (Johnston et al., 2003; Merikangas et al., 2010). SUDs are even more common among YIJ: between 25% and 50% report a disorder (McClelland et al., 2004; Teplin et al., 2012, 2021; Yurasek et al., 2021), and an estimated 70% of arrested juveniles have prior drug involvement (Belenko & Logan, 2003; Zhang, 2004). Adolescent SU is a risk factor for SUD in adolescence (Winters & Lee, 2008) and adulthood (Stone et al., 2012), and SU problems in YIJ contribute to their continued involvement with the justice system (Hoeve et al., 2013; Teplin et al., 2021; Wibbelink et al., 2017) as well as other poor outcomes such as HIV/STIs (Elkington et al., 2008), violence (Elkington et al., 2015) and early mortality (Teplin et al., 2014). The well-established overlap between justice involvement and SU, as well as evidence to suggest treatment reduces recidivism in these youth (Cuellar et al., 2004), indicates that identifying and treating their substance use is a crucial public health concern (Cuellar et al., 2004; Henggeler & Sheidow, 2012; Hoeve et al., 2013).

A recent national survey of probation agencies across 20 states revealed that 64% screened for behavioral health problems (Scott et al., 2019). However, despite the availability of evidenced-based substance use treatment for adolescents, between 50% and 90% of YIJ with substance use treatment need do not receive services even after they are identified as in need of treatment (Johnson et al., 2004; Wasserman et al., 2021; White et al., 2019), suggesting cross-systems referral and linkage are key points for intervention. This is particularly true for some youth, such as those from racial/ethnic groups who are less likely to be referred to and receive services than White YIJ (Dalton et al., 2009; Maschi et al., 2008; White, 2019).

Probation departments have become critical justice settings in which to implement programs that close the treatment gap for YIJ with SU problems via cross-system linkage and enrollment in services. As an initial point of contact for over half of U.S. youth entering the juvenile justice system after arrest (Office of Juvenile Justice and Delinquency Prevention, 2020), probation departments are well-positioned to reach large numbers of youth who would otherwise remain disconnected from the treatment system. However, because probation agencies typically do not offer SU services (Scott et al., 2019), youth on probation typically must move from the justice system to the community behavioral health system for SU treatment. Thus, effective cross-system linkage is an essential component of service uptake for these youth.

There are numerous system/organizational-, staff-, and youth/family-level factors that cause YIJ to “get lost” in the process referral from the justice to the treatment system (Gopalan et al., 2010; Stiffman et al., 2001). Within probation settings, poorly developed or absent referral and linkage procedures and limited interagency collaboration can increase the likelihood that youth do not make the transition across systems (Darlington et al., 2005; Elkington et al., 2020; Stiffman et al., 2001). Moreover, probation staff may have an incomplete knowledge of the behavioral health system or have had negative experiences with referral and cross-system collaboration (Stiffman et al., 2004), which can have ramifications for referral practices. Staff perceptions of youth’s treatment need, gender, and race/ethnicity (Farmer et al., 2003; Lopez-Williams et al., 2006; Teplin et al., 2005) have all been shown to affect probation officer (PO) screening and referral behaviors (Wasserman et al., 2008). At the family level, factors such as perceived need for treatment, stigma, family stress, and distrust of systems (Elkington et al., 2020; Gopalan et al., 2010; McKay & Bannon, 2004; Teplin et al., 2005) are significantly associated with limited service use and engagement. These findings suggest that an approach to linkage in which improved referral and interagency collaboration is combined with family engagement in the referral process and treatment planning, beginning in probation rather than the treatment setting, are critical to cross-linkage and service enrollment for YIJ.

Despite the potential of linkage programs to improve referral practices and interagency collaboration, such programs are not typically integrated into ongoing practice in probation settings. One of the few evidence-based linkage programs for youth on probation, Project Connect (Wasserman et al., 2009), is a manualized training that builds PO capacity to improve screening and referral by developing more systematized collaboration with partnering behavioral health agencies. Project Connect was successful in significantly increasing numbers of youth identified with behavioral health problems, staff referrals for treatment, and service uptake (Wasserman et al., 2009). That said, almost half of youth did not enroll in treatment despite referrals, and POs did not implement referral protocols consistently 1 year after program implementation, suggesting that a program that targets staff/organizational factors alone may not be sufficient to promote the level of uptake necessary for substantial and lasting change.

Previous research suggests that barriers to service use can be addressed by emphasizing youth/family engagement (Miller & Prinz, 2003; Nock & Kazdin, 2005). Although youth typically attain increased independence from caregivers during adolescence (Schulenberg et al., 1997), family involvement has been shown consistently to promote positive treatment outcomes, including retention (Prado et al., 2006; Staudt, 2007). Engaging the families of youth on probation at the point of identification of treatment need in the justice system may increase caregiver engagement in youth treatment, and in turn youth treatment attendance and retention. However, POs are not trained to deliver intensive family engagement strategies and adding tasks associated with family engagement to the PO’s role is not likely to be sustainable given their large caseloads and limited time (Schwalbe et al., 2014). Furthermore, the dynamic between POs and youth/families is often framed by sanction-based interactions (e.g., increased monitoring due to violations of the terms of probation), particularly for substance-using youth (Schwalbe & Maschi, 2009), and influenced by the youth/family’s prior experiences with and mistrust of child-serving systems (Elkington et al., 2020). Thus, we developed Family Connect, which aims to target both system- and family-level barriers via a linkage specialist who, after the PO has screened the youth and identified treatment need, will work exclusively to improve engagement, and actively collaborate with treatment agencies and the youth’s PO to enhance referrals and service uptake.

The goals of the current article are to describe Family Connect and to report the results of the pilot trial that examined the feasibility and acceptability, and preliminary evidence of effect of Family Connect in two probation settings in a northeastern state (Shanyinde et al., 2011). Following the intervention, a random subset of youth and caregivers who participated in Family Connect provided qualitative feedback about the feasibility and acceptability of the program, including their perceptions of, and satisfaction with, the intervention. The Behavioral Health Care Cascade has been proposed as a method to detect gaps in treatment via tracking an individual from identification, to referral, to treatment initiation (Belenko et al., 2017). Given the application of the cascade to follow movement across service systems (e.g., probation and behavioral health treatment) and to identify places for intervention (Wasserman et al., 2021), the Behavioral Health Care Cascade will be used to frame the pilot test evaluation of Family Connect. With respect to potential effect, we hypothesize that, compared with a historical control (care as usual) of youth on probation with substance use treatment needs, Family Connect will increase referrals, intake attendance, and initiation in behavioral health treatment.

METHOD

FAMILY CONNECT INTERVENTION

Grounded in community-based participatory research (CBPR) principles, Family Connect was developed in an iterative fashion with key stakeholders (youth, caregivers, probation, and treatment staff) who had firsthand knowledge of the target population and the probation context. This approach ensured that Family Connect incorporated relevant cultural (to both probation families and the organizational climate) and process factors and were acceptable and feasible given structural constraints specific to probation departments. This intervention development process using stakeholder input to develop targeted programs has been used previously to achieve health equity (Wallerstein & Duran, 2010), particularly for high-risk populations who are often overlooked (Holliday et al., 2020; Parsai et al., 2011). Youth and caregiver stakeholders involved in intervention development were not included in the pilot trial sample.

Family Connect was developed as a flexible intervention, to be delivered to youth on probation and their caregivers who resided in the community. Intervention length was variable, depending on youth and family needs; however, certain practices were retained across all families. The intervention was delivered by a linkage specialist who, possessing knowledge of both the justice and behavioral health systems, worked with the family to engage the youth into treatment. The intervention drew from Motivational Interviewing and Motivational Enhancement Therapy approaches and comprised five tasks that a linkage specialist accomplished with a family to assist the youth to initiate treatment. Table 1 describes the five tasks, goals of the tasks, and specific activities completed by the linkage specialists. The tasks did not need to be completed in order, nor did a task have to be completed before moving on to a different task; one exception to this was that the first task, which focused on engaging the family, had to be started first. This was to ensure maximum flexibility for the linkage specialist to work with and address the family’s needs. The linkage specialist met with the family in their home, probation agency or other agreed upon space where confidentiality could be ensured, and kept in contact with the caregiver and youth via phone and texts. Meetings occurred weekly or as often as needed until the youth engaged in treatment. The linkage specialist also collaborated with the youth’s PO and provider to facilitate communication and information sharing and to bring the family, behavioral health, and justice systems into closer alignment around the youth’s treatment.

TABLE 1:

Family Connect Intervention Tasks, Task-Related Goals, and Activities

Task	Goal	Linkage specialist activities
Task 1: Develop a foundation for a collaborative working relationship.	To engage the family though active listening and validation of the family’s experiences and current needs as perceived by the family.	Understand if the youth and caregiver believe youth has a substance use problem that requires treatment. Understand caregiver’s and youth’s attitudes about substance use and treatment. Understand family’s goals and motivations.
Task 2: Clarify the need for treatment and promote an opportunity for change.	To promote the idea that treatment is necessary for the youth, and that treatment could provide an opportunity for change	Build on knowledge learned in Task 1 for perceived need for youth’s treatment Move from listening to intervening, by identifying and setting goals.
Task 3: Explore caregiver-youth relationship and potential rupture.	To promote re-engagement of either caregiver and/or youth into the relationship, thereby facilitating entry into treatment.	Generate understanding of relationship tension. Get youth and caregiver to commit to improving communication for the purposes of linkage work. Support and reinforce the caregiver for what they do well and promote importance of the caregiver’s role in the treatment.
Task 4: Address concrete barriers to accessing and enrolling in treatment.	To remove concrete barriers that were in the way of achieving enrollment into treatment	Review available treatment providers; support family in choosing a provider. Identify and address any concrete obstacles to care (e.g., transport, insurance) Chose a provider, eliminate remaining concrete obstacles to accessing care, and attend an intake appointment.
Task 5: Facilitate treatment enrollment and retention.		Reinforce positive and reframe any negative treatment experiences. Locate new provider if necessary. Successfully terminate with family after youth and family attend treatment regularly in manner determined by the provider and the family (e.g., bi-weekly; weekly) for one month.

PARTICIPANTS AND RECRUITMENT

Pilot participants were recruited from county probation departments in two counties in a northeastern state between December 2015 and May 2017; treatment programs all provided services to youth on probation. Pilot procedures were approved by the New York State Psychiatric Institute Institutional Review Board.

Family Connect

Youth and caregivers were recruited for participation as dyads. Youth on probation were eligible if they (a) were between 10 and 17 years of age; (b) had a score of ≥4 (“caution” cutoff) on the Massachusetts Youth Screening Instrument–Version 2 (MAYSI-2; Grisso et al., 2012) drug and alcohol use subscale, indicating “probable clinical significance,” or as indicated on the Youth Assessment and Screening Instrument (YASI) substance abuse domain, or based other documented evidence of substance use problems (e.g., school/parent complaint, current complaint/charge, positive urine toxicology); and (c) had a caregiver willing to participate in the study. All substance use measures (i.e., YASI and MAYSI-2) were administered by probation officers as part of intake procedures. Caregiver participants were eligible if they had a youth willing to participate in the study and were the legal guardian of the youth. POs were asked to describe the study to all youth and caregivers on their caseloads that met study criteria and then provided the contact information of interested families to research assistants. Research assistants then followed-up with the families to establish study eligibility, further explain participation and to take consent/assent before study participation. POs were instructed to not consider this a referral to services so to allow families the ability to refuse research participation without concern for accessing services through probation to which they would normally be entitled. Youth and caregivers were compensated for study participation. Following consent and enrollment into the project, the linkage specialist contacted families and scheduled the first meeting to begin working toward the goal of youth treatment initiation.

Linkage Specialists

The Linkage Specialists were a licensed mental health counselor (LMHCs) and a licensed clinical social worker (LCSW). The Linkage Specialists received training in Family Connect by the first author, a licensed clinical psychologist, over three full days. The Linkage Specialists then received weekly supervision with the first author for the duration of participant enrollment in the study, during which a review of session activities, tasks addressed and completed were discussed along with any facilitators and barriers to the role.

Historical Comparison

The participants in the historical control comparison group were included if they (a) were between 10 and 17 years of age, (b) completed a probation intake during the 6 months before the Family Connect implementation, and (c) were identified as in need of substance use treatment. Substance use treatment need was based on responses to the YASI substance abuse domain as administered by the PO, or other documented evidence of substance use problems (e.g., school/parent complaint; current complaint/charge; positive urine toxicology). This information was gathered during the intake process.

MEASURES

Behavioral Health Need

As described above, youth substance use problems were identified through the YASI, or MAYSI-2 for youth in Family Connect group, and/or based other documented evidence. In addition, youth’s need for mental health treatment was similarly defined based on the YASI or MAYS-2.

Pilot Trial Outcomes

Outcomes for the pilot included whether the youth (a) received a referral to treatment (referral), (b) attended an intake appointment (intake), and (c) attended a first appointment following intake (initiation; intake + 1 session) for any behavioral health services (mental health, substance use treatment, or both). Data on youth treatment need, demographics, referral, intake, and attendance as documented by POs were derived from records held in the probation system’s management information system for both Family Connect and the historical control groups; linkage specialists also confirmed attendance with the youth providers for those in the Family Connect group. De-identified data were exported from the management information system at the end of the study in two ways: data were de-identified by probation staff and exported into password protected data files and data extraction forms were also used to ensure systematic review of probation records for treatment attendance. Linkage specialists, one at each location, completed weekly logs documenting type (e.g., phone, text, in-person) and number of contacts with families and providers, appointments made and attended with dates, including information about the treatment agency.

Demographic Information

Demographic variables for both groups included gender, age, race, ethnicity, probation case type, and county. During data extraction, race was categorized as White versus non-White to promote confidentiality, and ethnicity was categorized as Hispanic versus non-Hispanic. Because of the small sample size, a combined race/ethnicity variable was also created, categorized as White non-Hispanic versus non-White and/or Hispanic. Case type was defined as adjudicated juvenile delinquent versus persons in need of supervision (PINS); PINS cases are those referred to probation via school or parent referral and are considered voluntary.

Feasibility and Acceptability of Family Connect

Feasibility of the intervention was measured by the percentage of assented/consented participants who completed at least one meeting with the linkage specialist. Acceptability of the intervention was assessed via interviews that explored youth and caregiver satisfaction with the intervention. Following completion of their participation in Family Connect, half of the families enrolled in Family Connect were randomly selected (n = 9) to complete a brief in-depth interview with a research assistant that explored experiences working with the linkage specialist (e.g., “In thinking about how you and the linkage specialist worked together, what was it like?”), facilitators and barriers to participating in the program (e.g., “Why did you decide to participate in this program”; “Do you think it was helpful that the suggestion to participate came from the probation officer?”), as well as how to improve the program (e.g., “What would you add or change about the program to make it more helpful?”). All invited participants agreed to be interviewed, and youth and caregivers were interviewed separately.

Data Analysis

Youth enrolled in Family Connect (n = 18) were compared with a historical control group (n = 95) on each step of the behavioral health cascade—referral, intake, and initiation (intake + 1 session) for any behavioral health services (mental health, substance use treatment, or both). First, we examined demographic differences (gender, age, race, ethnicity, probation case type, and county) between Family Connect and control group youth, testing for significant differences with chi-square and t tests, for binary or continuous variables as appropriate. Next, we used logistic regression models to examine differences for each step of the cascade, comparing Family Connect to the control group. We first fit a model with just the main effect of treatment group. We then fit a model with demographic covariates: gender, age, and race/ethnicity combined variable, and case type.

All qualitative transcripts were completely de-identified and read by two team members (a coder and a second rater). Codes addressing (a) opinions about the acceptability of Family Connect and (b) suggestions for improvements were identified. Consistent with a thematic coding methodology, the coder created memos of emergent subthemes (Fereday & Muir-Cochrane, 2006). All memos were reviewed by the second rater and discussed by both the coder and second rater to achieve consensus. Subthemes were combined and summarized across participants.

RESULTS

Family Connect Pilot Sample Characteristics

As shown in Table 2, Family Connect youth (N = 18) were compared with control group (N = 95) on gender, age, race, ethnicity, case type, and county. The groups were not significantly different by gender, age, ethnicity, or county. Race was significantly different between the groups, χ²(1, N = 111) = 4.78, p = .03, with a greater proportion of non-White youth in Family Connect, compared with the control. Specifically, 72.2% of the Family Connect sample were minority youth, compared with 44.1% in the control. Case type was also significantly different, χ²(1, N =113) = 3.74, p = .05; there was a greater proportion of PINS youth in Family Connect, compared with the control group. As a condition for inclusion in the study, all youth in both groups had a substance use problem and in addition, 83.3% (n = 15) youth in the Family Connect group and 89.5% (n = 85) youth in the control group had a mental health problem.

TABLE 2:

Sample Characteristics

Characteristic	Family connect (n = 18) % / meana (SD)	Control (n = 95) % / meana (SD)	χ2/t testb
Youth
Gender			1.331, p = .249
Female	44.4%	30.5%,
Male	55.6%	69.5%
Age	14.49a (1.11)	14.88a (1.15)	0.21b, p = .83
Race		(n = 93)	4.78, p = .03
White	27.8%	55.9%
Hispanic	40%	26.0%
Non-White	72.2%	44.1%
Hispanic	46.1%	23.2%
Ethnicity		(n = 86)	4.28, p = .04
Non-Hispanic	55.6%	73.3%
Hispanic	44.4%	26.7%
Case type			3.74, p = .05
PINS	72.2%	47.4%
JD	27.8%	52.6%
County			0.912, p = .34
County A	77.8%	66.3%
County B	22.2%	33.7%
Caregivers
Gender
Female	83.3%
Male	16.7%
Age	45.56a (9.87)
Race
White	44.4%
Non-White	55.6%

Note. PINS = persons in need of supervision; JD = juvenile delinquent.

^aindicates mean value as opposed to percent;

^bindicates t-test statistic as oposed to χ².

Linkage Specialist Activities

Table 3 presents linkage specialist activities with Family Connect participants, which included in-person visits, phone, texts, and emails with the caregiver, youth, or both. The linkage specialist met all enrolled youth and their families at least once and worked with families for a median of 5.6 months (SD = 3.2, range = 1.1–13.0) and during that time met face-to-face with the youth and caregiver together an average of 5.1 times (SD = 5.8, range = 0–22). Face-to-face meetings with both the youth and their caregiver were more frequent than meeting with either caregiver (M = 1.5, SD = 2.2, range = 0–7) or youth (M = 1.1, SD = 2.1, range = 0–9) alone. The linkage specialist also communicated via phone/text/email between face-to-face sessions an average of 29.4 times (SD = 29.7; range = 4–110), primarily contacting the caregiver.

TABLE 3:

Linkage Specialist Activities

Activity	Median	SD (range)
Length of time working with families	171.6 days (5.6 months)	96.0 days (32–394)
Caregiver & youth	5.1	5.8 (0–22)
Face-to-face visits	Mean	SD (range)
Caregiver	1.5	2.2 (0–7)
Youth	1.1	2.1 (0–9)
Call/text/email	Mean	SD (range)
Caregiver & youth	0.6	0.8 (0–3)
Caregiver	29.4	29.7 (4–110)
Youth	3.4	7.0 (0–30)

Referral, Intake Attendance, and Treatment Initiation

Figure 1 shows the percent of youth in Family Connect and those in the historical control who were referred, attended an intake, and initiated behavioral health treatment (intake + 1 session). Our study was not powered to detect significance. However, examining bivariate relationships between behavioral health cascade outcomes and study group (Table 4), we examined the preliminary evidence of effect of the intervention. Specifically, we found a trend toward significance such that youth in Family Connect (94%, n = 17) were 6 times as likely to be referred to behavioral health treatment, compared with youth in the historical control (74%, n = 70) (OR = 6.07; 95% CI [0.77, 48.01]; p = .09). Among those referred, 88% (n = 15) of youth in Family Connect compared with 71% (n = 50) of those in the control group attended an intake appointment (OR = 3.00; 95% CI [0.63, 14.33]; p = .17) and just over half of the Family Connect youth (53%; n = 9) attended a treatment session compared with 61% of the historical control (n = 43; OR = 0.81; 95% CI [0.27, 2.42]; p = .70). Although between-group differences for intake between the Family Connect and historical control group were not significant, the odds ratio suggests that, if powered appropriately, we may have found a statistically significant difference.

Figure 1: Rates of BH Referral, Intake, and Initiation for BH Services Both Control (Pre-Study) and Treatment Groups.

^aOne youth in the Family Connect group completed intake for both substance use and mental health treatment but only initiated mental health treatment. ^bDenominator for Intake and Initiation defined as total referred to BH treatment.

TABLE 4:

Predicting Screening and Intake in Any BH Treatment

	Referral				Intake
Predictors	OR	[95% CI]	AOR	[95% CI]	OR	[95% CI]	AOR	[95%CI]
Family connect	6.07	[0.77, 48.0]	4.34	[0.52, 36.25]	3.00	[0.62, 14.33]	4.10	[0.78, 21.72]
Female	—		0.99	[0.34, 2.85]	—		2.14	[0.65, 7.04]
Age	—		1.10	[0.72, 1.68]	—		1.15	[0.72, 1.82]
Non-White/Hispanica	—		1.72	[0.61, 4.84]	—		0.51	[0.16, 1.64]
PINSb	—		3.27	[1.09, 9.82]	—		0.34	[0.10, 1.14]

Note. OR = odds ratio; AOR = adjusted odds ratio; CI = confidence interval; PINS = persons in need of supervision.

^aWhite non-Hispanic referent category.

^bJuvenile delinquent (JD) referent category.

Figure 1 also shows the percent of youth who were referred for, completed an intake, or initiated treatment for substance use treatment only, mental health treatment only, or both mental health and substance use treatment. In Family Connect, the majority of youth (88%) were referred to substance use and mental health services as compared with either substance use only or mental health only. As such, the majority of those who completed an intake (80%) and initiated treatment (78%) did so for substance use and mental health treatment. Whereas in the control group, more youth were referred to (43%), completed an intake (46%), and initiated (47%) mental health treatment only as compared with either substance use only or both mental health and substance use treatment.

After adjusting for gender, age, race/ethnicity, and case type (PINS versus JD), we found no significant differences between youth in Family Connect and youth in the historical control group with respect to referral or intake (see Table 2). Nonetheless, the effects are in the expected direction, suggesting with sufficient power in a larger randomized controlled trial Family Connect may show efficacy. In the referral model, only case type was significantly associated with referral, with PINS youth more likely to be referred compared with JD youth (OR = 3.27; 95% CI [1.09, 9.82]), and none of the covariates were significant predictors for intake.

Family Acceptability

Overall, youth and caregivers found Family Connect to be acceptable. Qualitative analyses revealed two primary emergent themes from youth and caregiver interview data that spoke to the importance and acceptability of the linkage specialist role: (a) the importance of the linkage specialist in improving youth–caregiver communication and alignment about treatment, and (b) the importance of the linkage specialist in providing education about treatment in a collaborative manner. First, both youth and caregivers reported that they found the presence of the linkage specialist promoted youth/caregiver communication, and that the first conjoint meetings were particularly important. For example, when asked whether meeting together, rather than separately, for the first meeting was preferable, one youth stated, “It was helpful [to meet conjointly], because [the linkage specialist] was on my side. My mom needed another adult to be there, so she’d understand.” Similarly, a caregiver reported, “I think it’s good for there to be some group meetings, but you know I want [youth] not to think I [was] just standing around the corner, so I would go outside or whatever, so she could talk, too. But it’s good for everyone to have one or maybe a couple group meetings.” In these responses, youth and caregivers highlighted the importance of blending individual with conjoint meetings between the linkage specialist, youth, and their caregivers (versus only conjoint or only individual meetings).

Second, youth and caregivers agreed that the way the linkage specialist communicated about treatment services was helpful. Specifically, the communications from the linkage specialist were clear, and information about substance use and treatment services were presented in a collaborative manner, as nonthreatening and supportive of youth. For example, underscoring the utility of supporting youth’s knowledge about substance use and treatment to increase treatment motivation and uptake, one youth reported,

She told me like a lot of facts like what drugs would do to you. Like why I’m surprised? She gave me a lot of facts and information [about treatment], I had no idea…that basically it’s a place to get how you feel out more frequently…They’re there to help you. She just made it seem like it was about positive vibes.

Similarly, emphasizing the utility of using a collaborative method one caregiver described how she appreciated the linkage specialist’s approach to communication and reported, “She wanted to talk to me and him. Just asked us questions and she would offer services and ask us ‘do we think this would help,’ ‘do we think that would help.’”

Finally, youth and caregivers made multiple suggestions to improve future iterations of Family Connect. Youth, but not caregivers, expressed an interest in more frequent contact with the linkage specialist. When asked how the program could be improved, one youth stated, “Seeing the linkage specialist more than once every week. More than four times in a month—that would have been helpful.” Youth and caregivers also reported that it would be helpful to incorporate additional types of referrals in future iterations of Family Connect, although they differed on the types of additional referrals they would have liked; caregivers emphasized the importance of anger management referrals, while youth reported wanting family therapy.

DISCUSSION

This family linkage program for youth on probation, Family Connect, shows promise, feasibility, and family acceptability in achieving cross-systems linkage as youth move from the justice system to access treatment in the behavioral health system. Specifically, although our study was not powered to achieve significance, results suggest support for Family Connect as an approach to increase cross-systems referral and behavioral health treatment initiation. Moreover, findings suggested support for the feasibility of the intervention and indicated that justice-involved youth and their caregivers found the intervention to be acceptable.

Informed by input from key stakeholder groups, Family Connect was implemented by a linkage specialist who, while working with the youth and his or her caregiver, facilitated the youth’s linkage to behavioral health treatment through the completion of five key tasks designed to promote engagement and reduce perceived and logistical barriers to treatment. All but one family enrolled in the study met with the linkage specialist at least once, most working with the linkage specialist for just under 6 months.

In concert with justice reform over the past decade and research indicating that incarcerating youth does not necessarily result in superior outcomes compared with probation with community-based intervention (Lee et al., 2012; Loughran et al., 2009), today more youth are placed on probation than in the past (Skeem et al., 2014). While provision of services occurs “onsite” in locked facilities, the majority of probation agencies refer “offsite” into the community for necessary services, after identification of treatment need (Scott et al., 2019). Thus, movement between two systems is required, which frequently results in youth falling between the two systems, never accessing the treatment they need (Elkington et al., 2020; Wasserman et al., 2021). The Behavioral Health Care Cascade is a useful framework that can drive the development of linkage interventions by identifying points of intervention (i.e., where youth fall off the cascade) and inform the evaluation of the intervention (Belenko et al., 2017). To our knowledge, this is the first published use of the Behavioral Health Care Cascade to evaluate a linkage intervention. Using the Behavioral Health Care Cascade, the current study found that Family Connect demonstrated promise in both referring and linking (i.e., intake) youth to treatment.

A recent study (Wasserman et al., 2021) in which the Behavioral Health Care Cascade was used to identify barriers to treatment initiation by youth on probation found PO referral to services to represent a significant drop-off along the care cascade: Among those identified as in need, approximately four-fifths were not referred to treatment. Although the current study was not powered to detect differences between groups, we found a trend toward significance for more youth in Family Connect to be referred to care as compared with youth in the control group. Prior research suggests that referral by probation officers to treatment is determined by organizational support, the degree to which POs perceive providing referrals as relevant to their job role, PO knowledge of the treatment system, and PO perceptions of youth’s treatment needs (Knight et al., 2019; Lopez-Williams et al., 2006; Stiffman et al., 2004; Teplin et al., 2005). With knowledge of behavioral health and the behavioral health system, and an explicit focus on working with families to achieve treatment engagement, data suggest that the linkage specialist has the ability to augment PO referral and linkage practice through focused and accurate referrals and family engagement.

Intake marks the point in the behavioral health cascade in which youth pass from the juvenile justice system into the behavioral health system. Although the current study was not powered to identify significant differences in initiation between the treatment and control conditions, the direction and magnitude of the effects suggest that youth in Family Connect may have been more likely to attend an intake appointment. Elkington and colleagues (2020) found that for youth on probation and their families, distrust of providers and the behavioral health system, which is associated with stigma and past negative experiences, serves as a meaningful barrier to care. However, as the qualitative findings presented here suggest, by developing rapport and presenting treatment options in a supportive manner to youth and families, the linkage specialist was able to reduce barriers and promote families accessing treatment.

In this pilot study, we found that approximately half of the youth in Family Connect initiated treatment (i.e., attended one session following intake); slightly less than the youth in the historical control. Given that more youth in the control group were JDs, the rate of treatment retention among the controls may have been a function of court mandated treatment, which has been shown to be strongly associated with treatment retention (Wasserman et al., 2021). This finding suggests that in future trials of Family Connect adjustments should be made specifically to ensure that PINS youth remain in treatment following intake. For example, the linkage specialist could help by addressing communication and coordination issues between probation departments and the treatment system, serving as a liaison between these systems and the family. Early therapeutic alliance is important for maximizing treatment outcomes and retaining youth in care (Van Benthem et al., 2020). Therefore, the linkage specialist could facilitate the establishment of alliance between the clinician and the youth by explaining to the youth and caregiver what early treatment will look like, so as to set expectations (e.g., explaining that the first session will likely focus on paperwork, but that additional sessions will be more tailored to the youth’s needs). Moreover, this underscores the importance of having the linkage specialist, who already successfully developed rapport with the family, reinforcing (e.g., through praise) the youth/caregiver for engaging in a session following intake.

Rates of comorbid mental health problems across both study groups were high, consistent with prior studies of justice-involved youth (Teplin et al., 2021; Wasserman et al., 2010). As such, in Family Connect most youth were referred to both mental health and substance use services based on clinical need. Interestingly, for some youth referred to both substance use and mental health services, intake and treatment initiation were only successful at mental health services. This may be due to the stigma and shame for parents associated with youth substance use and associated treatment, particularly for parents of youth involved in the justice system who are concerned about their children becoming further system involved (Elkington et al., 2020). In contrast, more youth in the historical control received a referral for mental health treatment only, and thus more youth completed an intake and initiated mental health treatment only, despite all youth demonstrating a substance use problem. Such referrals may reflect POs limited knowledge of behavioral health as well as the treatment system (Scott et al., 2019), which results in an inaccurate treatment referral.

There were several limitations in this pilot study. First, as a pilot study the Family Connect treatment group had a small sample size (n = 18), which precluded our ability to determine statistically significant differences between study groups. This small sample size also limited our ability to analyze and determine patterns related to additional variables, including race, gender, and case type. Given the overrepresentation of youth from minority groups in the justice system, and the fact that YIJ of color are less likely than their White peers to receive treatment (Dalton et al., 2009; Maschi et al., 2008; White, 2019), future research should specifically examine whether Family Connect is effective at reducing racial/ethnic disparities in service referral and uptake. Using a retrospective control group, our data may be biased due to historical effects and nonrandomization; moreover, youth substance use in the Family Connect group was also identified by the MAYSI-2 in addition to the YASI. In addition, we compared linkage specialist-confirmed intake and attendance data for youth within Family Connect administrative records data for youth in the historical control comparison. Family Connect participants were identified by POs, which may lead to biases based on POs subjective assessment of the youth and families. Indeed, this might account for the overrepresentation of PINS cases in the Family Connect sample. We did not ascertain whether the PO also made a treatment referral for the families enrolled in the current study nor do we have data on PO caseload during the Family Connect study period to ascertain how many other “eligible” youth should have been recruited and handed off to Family Connect as well as referred to behavioral health services but were not. In a larger efficacy trial, we would be better able to compare concurrent treatment versus standard of care arms with randomly assigned participants. Nonetheless, despite these limitations, findings from the current study have several important implications for the field.

For many youth in the justice system, and youth from minority racial/ethnic groups in particular, involvement in the justice system may be the first time a behavioral health problem is identified (Lopez-Williams et al., 2006; Maschi et al., 2008; White, 2019). When youth behavioral health needs are not identified or are not referred to treatment, the justice system represents a missed opportunity to get these youth the services they need. Involving families in treatment is critical for increasing service utilization and success rates (Hogue et al., 2021). However, family engagement is low, with barriers including system distrust, denial/minimization of youth problems, relational barriers, and family disruption (Elkington et al., 2020). Family Connect was developed to target the unique needs of youth on probation and their families by incorporating input from families in the formative work and by using a CBPR approach. Pilot findings indicate that the intervention can facilitate youth’s linkage to care, in the face of these obstacles. Examination of the program in a larger study is warranted to determine the efficacy of Family Connect as well as further explore the mechanisms through which the program achieves improvements in referral, linkage, and retention in treatment.

Biographies

Katherine S. Elkington, PhD, is an associate professor of medical psychology (in psychiatry) at Columbia University and a research scientist and the New York State Psychiatric Institute. She serves as the director of the Center for Behavioral Health and Youth Justice. She has over 15 years of research experience in justice settings, with expertise in understanding correlates of mental illness, substance use/disorders, and HIV risk behaviors and using these data to develop and implement prevention programming, and in understanding multilevel barriers to treatment access for justice-involved individuals. Most recently, her work has focused on the development and evaluation of implementation interventions to increase access to behavioral health and health services, of which a centerpiece is building cross-system relationships between justice and community treatment systems.

Gail Robson, MPH, is a research coordinator at Columbia University, in the Division of Child and Adolescent Psychiatry, Department of Psychiatry. She received her MPH from Columbia University Mailman School of Public Health in Population and Family Health in 2019 and has worked on research projects focused on mental health, substance use, and social determinants of health.

Corianna E. Sichel, PhD, is a postdoctoral research scientist at Columbia University, in the Division of Child and Adolescent Psychiatry, Department of Psychiatry. Her research focuses on understanding and addressing the etiology and behavioral health sequelae of traumatic experiences in vulnerable populations of young people, including youth involved in the juvenile and criminal legal systems.

Jacqueline Lee, RN, is a registered nurse in New York. She is completing her doctor of nursing practice in psychiatric mental health at Columbia University School of Nursing.

Gail A. Wasserman, PhD, is a professor of medical psychology (in psychiatry) at Columbia University in the Division of Child and Adolescent Psychiatry, Department of Psychiatry, and the founding director of the Center for the Promotion of Mental Health in Juvenile Justice (now the Center for Behavioral Health and Youth Justice). She has been investigating the psychopathology of child and adolescent conduct problems for over 30 years. Her contributions to the field include the systematic documentation of the range, severity, and correlates of substance use and mental health problems in youth in contact with the juvenile justice system, and the development of multilevel interventions to identify behavioral health need among youth involved in the justice system, and successfully link them to appropriate care.