Article Review 1000 wrds due in 24 hours

Vol:.(1234567890)

Community Mental Health Journal (2021) 57:1094–1110 https://doi.org/10.1007/s10597-020-00735-z

1 3

O R I G I N A L PA P E R

Randomized Controlled Trial of an Integrated Family‑Based Treatment for Adolescents Presenting to Community Mental Health Centers

Ashli J. Sheidow1 · Kristyn Zajac2 · Jason E. Chapman1 · Michael R. McCart1 · Tess K. Drazdowski1

Received: 14 May 2020 / Accepted: 23 October 2020 / Published online: 29 October 2020 © Springer Science+Business Media, LLC, part of Springer Nature 2020

Abstract Most adolescents presenting to community mental health centers have one or more comorbidities (internalizing, externalizing, and substance use problems). We evaluated an integrated family-based outpatient treatment for adolescents (OPT-A) that can be delivered in a community mental health center by a single therapist. A sample of 134 youth/families were randomized to receive OPT-A or usual services, delivered at the same public sector mental health center. Repeated, multi-informant assess- ments occurred through 18-months post-baseline. At baseline, the sample displayed low internalizing symptoms, moderate substance use, and high externalizing problems. Compared to usual services, OPT-A had effects on abstinence rates, retention, motivation, parent involvement, and satisfaction, but not on internalizing or externalizing problems. While OPT-A achieved some key improvements for youth who present to community mental health centers, and families were satisfied with treat- ment, continued work is necessary to examine treatments for comorbidity while balancing treatment feasibility and complex strategies to boost treatment effectiveness.

Keywords Comorbidity · Adolescent substance use · Adolescent mental health · Community-based treatment · Family therapy

The majority of adolescents presenting for outpatient treat- ment at public sector community mental health centers have comorbid symptoms such as internalizing, externalizing, and substance use problems, but there is a striking paucity of

experimental treatment research on integrated psychosocial treatments for these teens (Hogue et al. 2018; Hulvershorn et  al. 2015). In general, multicomponent treatments that have been found to effectively target each of these present- ing problems (e.g., cognitive behavioral therapy [CBT], behavioral family-based treatment, motivational enhance- ment) tend to be most effective for treating substance use and comorbid disorders simultaneously (Brewer et al. 2017). However, research in this area is still limited by very few well-powered randomized controlled trials; and, of these, most are very limited (e.g., substance use and depres- sion only), are sequential or parallel treatment rather than integrated (i.e., requires attendance at two or more sepa- rate treatments), and are not representative of the range of adolescents seen in public sector community mental health centers.

Over 4 million U.S. adolescents use illicit drugs each year, and over 1 million are in need of treatment for sub- stance use (SAMHSA 2018). Tens of thousands of U.S. youth receive some form of addiction treatment services each year, the vast majority of which occur in outpatient community-based clinics (SAMHSA 2018). An estimated 60% of teens with substance use problems have a comorbid

Electronic supplementary material The online version of this article (https ://doi.org/10.1007/s1059 7-020-00735 -z) contains supplementary material, which is available to authorized users.

* Kristyn Zajac [email protected]

Ashli J. Sheidow [email protected]

Jason E. Chapman [email protected]

Michael R. McCart [email protected]

Tess K. Drazdowski [email protected]

1 Oregon Social Learning Center, 10 Shelton McMurphey Blvd., Eugene, OR 97401, USA

2 University of Connecticut School of Medicine, 263 Farmington Avenue, Farmington, CT 06030, USA

1095Community Mental Health Journal (2021) 57:1094–1110

1 3

mental health diagnosis, with comorbid externalizing dis- orders being the most common (up to 50%; Armstrong and Costello 2002; Couwenbergh et  al. 2006). Internalizing disorders (i.e., mood or anxiety) are also prevalent (up to 30%; Armstrong and Costello 2002). Approximately half of adolescents presenting for outpatient treatment for either substance use or internalizing disorders exhibit both (e.g., Stephens et al. 2014; Turner et al. 2004). A number of mod- els have been proposed as explanations for this comorbid- ity among youth (e.g., the self-medication model, common factor model, bi-directional model), but no definitive model has been accepted by the field (Kaminer and Winters 2020). Regardless of the underlying reasons for the link between disorders, it is well established that youth with comorbid substance use and mental health problems present with more familial, school, and legal problems than adolescents with substance use problems alone (Grella et al. 2001). It is not surprising that clinicians in public sector community mental health centers are tasked with treating complicated youth presentations.

Furthermore, youth with comorbid disorders are at par- ticular risk for negative outcomes. For example, Copeland et al. (2007) found that youth with anxiety and substance use were 11 times more likely to engage in severe or vio- lent offending, and youth with depression and substance use were nearly 15 times more likely compared to youth without psychiatric and behavioral problems. Youth with internalizing and externalizing symptoms are also consist- ently found to be at increased risk for initiating substance use earlier (e.g., Molina et al. 2018), increased amount and severity of substance use (e.g., Warden et al. 2012), future substance-related disorders (Groenman et  al. 2017), and relapse and negative clinical outcomes (e.g., Godley et al. 2014a, b; Tomlinson et al. 2004) compared to youth without comorbid problems. Even more problematic, Evans et al. (2007) showed an interactive impact of multiple comorbid disorders rather than simple additive effects. Together, these findings highlight the need for special attention to maximiz- ing treatment outcomes for youth with comorbid disorders, especially at community-based clinics where the vast major- ity of these youth are treated (SAMHSA 2018).

Pilot research conducted specifically on treatments target- ing substance and comorbid problems among adolescents are few in number, with most being uncontrolled and demon- strating modest results and no follow-up experimental trials (e.g., Curry et al. 2003; Goldstein et al. 2014; Hides et al. 2010; Kaminer et al. 1998). As noted, well-powered experi- mental studies have typically not examined psychosocial treatments for the range of adolescent clients seen in public sector community mental health centers, although a few have specifically targeted comorbid substance use and depression. For instance, Rohde et al. (2014) evaluated sequencing of a family-based treatment for substance use and group CBT

for depression, finding substance use improved most in the group receiving substance use treatment first and depres- sion improved to a similar degree in all groups. Importantly, however, the two treatments were delivered independently by separate therapists, rather than as an integrated model. Such sequential models are poorly suited for community- based clinics that often lacking the resources and staffing to provide multiple specialized therapists for each youth. Riggs et al. (2007) large-scale fluoxetine trial for conduct disordered youth with comorbid substance use and depres- sion provided further support for outpatient CBT. CBT for substance use was provided in both pharmacotherapy con- ditions, albeit in an academic clinic rather than a commu- nity clinic setting, with equivalent gains across conditions in substance use and one depression measure. A recently published trial of an integrated family-based intervention delivered in child advocacy centers focused on subthreshold substance use and trauma-related problems in youth, with more improvement in days of substance use for the treatment group compared to usual services, but no group differences for trauma-related symptoms (Danielson et al. 2020).

Experimental research on outpatient treatments for adolescent substance use alone is more extensive than the comorbid treatment literature. Reviews support several models (e.g., Hogue et al. 2018), with family-based mod- els consistently found to be superior (Tanner-Smith et al. 2013). Within family-based approaches, inclusion of CBT and behavioral strategies, rather than a sole focus on family relationships and communication, is routine (e.g., Donohue and Azrin 2001; Henggeler et al. 2011; Liddle et al. 2001). In fact, CBT and behavioral strategies have primarily proven successful within family-based approaches, with parents used to support CBT skills and implement behavior modi- fication at home (Hogue et al. 2018). This is not surprising given the key role of family in etiology and maintenance of substance use (Rowe 2012). Further, Hogue et al.’s (2014) review concluded substance use clients with comorbid dis- orders benefited more from family treatment than other approaches.

Experimental research specific to treatments for adolescent internalizing and externalizing problems individually is more extensive as well, compared to research for substance use and comorbid disorders. Adolescent treatments for internalizing disorders have been the most extensively studied, with reviews consistently concluding that psychosocial interventions are robustly beneficial (David-Ferdon and Kaslow 2008; Higa- McMillan et al. 2016). These reviews provide strong sum- maries of the literature, identifying various behavioral and CBT approaches as evidence-based and some approaches that include a parent component as probably efficacious. Reviews have highlighted that adolescent treatments for externalizing problems have been primarily tested in justice-involved popu- lations, which are more difficult to treat, and treatment effects

1096 Community Mental Health Journal (2021) 57:1094–1110

1 3

are modest compared to treatments for internalizing behaviors (Bakker et al. 2017; McCart and Sheidow 2016). Neverthe- less, similar psychosocial approaches (i.e., CBT, behavioral, family-based) appear to be promising for adolescents with externalizing problems. Importantly, effects of psychosocial interventions are generally not as substantial when evaluated in real-world settings where dually diagnosed youth are not excluded (Weersing et al. 2006; Weersing and Weisz 2002).

Since there are few evidence-based adolescent treatments that could be deployed in public sector community mental health settings, particularly one that can address a range of comorbid problems, it is no wonder adolescent treatment programs are cited as being unprepared to serve such cases (Hawkins 2009). Thus, we endeavored to develop an effec- tive integrated model specifically for outpatient providers at public sector community mental health centers that could be used to treat a range of comorbid problems in adolescents, including internalizing, externalizing, and substance use. We incorporated prior findings establishing CBT and behavioral techniques as central to treating specific problems and used a family-based approach. Although family-based approaches have some logistical challenges (e.g., they tend to be more complex, parents and adolescents sometimes have competing goals, some parents are difficult to engage), overwhelming evi- dence supports the efficacy of such approaches in achieving and maintaining change, particularly for complex youth pres- entations (McCart and Sheidow 2016). We used an integrated instead of a sequential approach (e.g., Rohde et al. 2014). Integrated approaches are consistently recommended for outpatient treatment development, especially for adolescents with comorbid disorders (e.g., Aviram et al. 2001; McCart and Sheidow 2016), but little research on such treatments exists.

We conducted a randomized controlled trial comparing an experimental treatment (OutPatient Treatment for Adoles- cents; OPT-A) to treatment as usual (TAU) through 18-months post-referral. To ensure high external validity (e.g., community referral streams, “real world” clients), we conducted the trial in a public sector community mental health center rather than a university clinic, with therapists who were employed by the mental health center delivering the intervention. We hypoth- esized that OPT-A would produce greater improvement in pri- mary outcomes (substance use, internalizing, externalizing) than TAU at both short- and long-term follow-up, as well as better functional outcomes (parenting, school enrollment) and therapy process (working alliance, satisfaction, motivation).

Method

Participants

Recruitment

Adolescents were recruited from a public sector community mental health center. Inclusion criteria were: (a) seeking out- patient treatment; (b) 10–17 years of age; (c) residing with a parent/adult caregiver; and (d) per clinic staff, determined to require treatment for current comorbid substance use and internalizing (mood or anxiety) disorders. We emphasized these comorbid disorders because most adolescents in com- munity mental health centers present with externalizing problems, and this was an effort to recruit a broader sample that is typically understudied and more difficult to identify in community mental health settings. Exclusion criteria were minimal to maintain a “real-world” sample: pervasive developmental disorder, psychotic disorder, or severe/pro- found intellectual disability. To maximize the study’s exter- nal validity and mirror customary admittance practices for community mental health centers, the clinic’s assessment of treatment need was used for inclusion criteria (d). Of 152 youth assessed by the research team and determined to be eligible, 18 families declined and 134 were randomized (88% recruitment rate; Fig. 1).

Demographics

Participants were 11.9–17.9 years of age (M = 16.0, SD 1.1) and representative of the community: 76.4% White, 15.7% African American, 5.7% more than one race, 2.1% not reported; 2.1% were Hispanic. Over half were male (62.1%). A slight majority resided in single-parent homes (52.9%), and 67.2% of the youth’s mothers were the primary parent. Families were economically diverse (income: 19% < $10 k, 18% $10  k–$20  k, 33% $20  k–$40  k, 30% > $40  k; 50% received government assistance). Parents generally worked full time (49.3%), although some worked part-time (5.2%), were homemakers or students (12.7%), retired (7.5%), or unemployed (25.3%). Parental education was: 16.3% < HS diploma, 26.7% HS graduate, 34.1% some college, 20.0% college graduate or more. There were no between-group dif- ferences in demographics.

Procedures

Once clinic staff referred youth, researchers arranged an in- person meeting at a time and place convenient for the family to obtain parental consent, adolescent assent, and confiden- tial baseline assessments. Average length of time between

1097Community Mental Health Journal (2021) 57:1094–1110

1 3

clinic referral and study consent was 12 days. Full interviews were at 3-, 6-, 12-, and 18-months post-baseline; at 1- and 2-months, select instruments (see below) were administered. Families received $60 for baseline, $40 for each subsequent full assessment, and $60 total for monthly phone calls to maintain contact information and collect brief data at 1- and 2-months post-baseline.

Urn randomization was used to balance: (1) sex; (2) age (10.0–12.9; 13.0–15.9; 16.0–17.9); (3) internalizing

severity (Revised Children’s Anxiety and Depression Scale [RCADS] borderline clinical threshold T-score < 65 vs 65+); and (4) substance use “severity” (urine screen posi- tive for < 3 different drugs vs 3+). Age was stratified due to the expectation that most would fall in the oldest cat- egory and fewest in the youngest category. An investigator conducted randomization to keep interviewers blind. IRBs at the Medical University of South Carolina and the state Department of Mental Health approved procedures.

Fig. 1 CONSORT participant flow diagram. *The n for specific measures differed due to adolescents being in detention, residential treatment settings, etc. Also, some parents were occasionally unavail-

able for assessments even if their youth were available. Intent-to-treat analyses were utilized

1098 Community Mental Health Journal (2021) 57:1094–1110

1 3

Assessments

Clinical Outcomes

Data were from baseline and 3-, 6-, 12-, and 18-months post-baseline. Substance use was assessed using lab-based 7-panel urine toxicology screens (marijuana, cocaine, ben- zodiazepine, amphetamine, methamphetamine, opiate, synthetic marijuana). Interviewers took steps to prevent tampering (see Henggeler et  al. 2011), and adulterants were lab-tested. Substance use was considered present if any results were positive. Results were excluded if parents reported youth medication that would result in a positive screen.

Internalizing symptoms were measured using the 47-item youth-reported RCADS (Chorpita et al. 2000), an adapta- tion of the Spence Children’s Anxiety Scale (SCAS; Spence 1998) developed to closely correspond to DSM-IV anxiety disorders and major depression. The measure is composed of 7 subscales (separation anxiety disorder, social phobia, generalized anxiety disorder, obsessive compulsive disorder, panic disorder, total anxiety, major depressive disorder) and a Total scale. Total scale T-scores were used in the current analyses (Cronbach’s alpha = 0.94). Previous studies have reported adequate internal consistency, test–retest stability (Chorpita et al. 2000), and favorable convergent, discrimi- nant, and factorial validity for both anxiety and depressive disorders (Chorpita et  al. 2005). The RCADS measures internalizing symptoms similarly to well established youth measures of depression and anxiety (de Ross et al. 2002).

Externalizing problems were measured by the par- ent Child Behavior Checklist (CBCL; Achenbach 1991) T-scores (Cronbach’s alpha = 0.93). The CBCL has good inter-rater reliability, test–retest reliability, criterion-related validity, and convergent and discriminant validity.

Functional Outcomes

School enrollment was reported by parents at 6-month inter- vals. This variable was coded “1” if youth were enrolled or had completed their education and “0” if the youth dropped out or was expelled or in placement. Parent and youth ratings of parenting were collected at baseline and at 3-, 6-, 12-, and 18-months post-baseline. The 42-item Alabama Parent- ing Questionnaire (APQ; Shelton et al. 1996; Zlomke et al. 2014) assesses: Corporal Punishment, Inconsistent Disci- pline, Poor Monitoring/Supervision, Positive Involvement, Positive Practices/Reinforcement (Cronbach’s alpha = 0.72 for parent, 0.74 for youth). Two 4-item subscales were added to capture additional parenting practices specific to adoles- cents. A rules/expectations subscale assessed enforcement of household rules and behavioral expectations. An Indirect Supervision subscale, similar to the Oregon Social Learning

Center Monitoring scale (Brown et  al. 1991), measured supervision at parties/friends’ houses and parents’ knowl- edge of the youth’s whereabouts.

Therapy Process

Except for motivation, the therapy process variables were measured at 1-, 2-, and 3-months post-baseline. Parent sat- isfaction with treatment was assessed with an abbreviated consumer satisfaction questionnaire (CSQ; Attkisson and Greenfield 2004). Three items account for > 85% of variance and represent general satisfaction, willingness to recommend the service, and willingness to seek the same services again if needed (Larsen et al. 1979). Previous studies reported high internal consistency (Attkisson and Greenfield 2004). Cronbach’s alpha in the current sample was 0.88. Parental involvement in treatment was measured using three items from the Vanderbilt Satisfaction Scale (VSS; Brannan et al. 1996), which assesses the parent’s understanding of and involvement in treatment, as well as satisfaction with level of involvement. Cronbach’s alpha for the current sample was 0.92. Working alliance was measured using parent- and youth-reports on the 12-item Working Alliance Inventory (WAI; Horvath and Greenberg 1989), which measures emo- tional bond and agreement on therapeutic goals and tasks. It has adequate psychometric properties (Martin et al. 2000), and both youth and parent reports predict treatment out- come (Gaudiano and Miller 2006; Hawley and Weisz 2005; Kazdin and Whitley 2006). Cronbach’s alpha for the current sample was 0.87 for parent and 0.84 for adolescent versions. Due to similarities between the goals and tasks subscales, these were averaged (Agreement on Goals & Tasks), while Emotional Bond was analyzed separately.

Treatment motivation was assessed at baseline and at 1-, 2-, and 3-months post-baseline, using a 4-item, 4-point Client Motivation scale informed by other youth treatment studies (Friedman et al. 1995). High scores indicated more motivation to change and more faith in treatment. Motiva- tion to continue treatment was assessed at follow-ups. A few items focused exclusively on substance use. All parents rated “how willing are you to help your child cut down on use?” at the highest level; thus, this item was dropped. For the youth version, this item was analyzed separately as Motiva- tion to Cut Down. Two items (“how willing are you/is your child to receive outpatient services?” and “how confident are you that outpatient treatment will help you/your child?”) were averaged for each informant as Treatment Motivation/ Confidence.

Conditions

The two conditions were OPT-A and treatment as usual (TAU). Both conditions were provided by master’s level

1099Community Mental Health Journal (2021) 57:1094–1110

1 3

clinicians employed in the same Child, Adolescent, and Family Services (CAFS) program of a public sector commu- nity mental health center. This county mental health center (i.e., each county in the state has a community mental health center) is operated by a state Department of Mental Health. All TAU and OPT-A treatment was paid by Medicaid, pri- vate insurance, or state funding. Two therapists only deliv- ered OPT-A, and TAU cases were assigned to therapists not trained in OPT-A. Therapists agreed to maintain separation of the conditions, and OPT-A therapists refrained from dis- cussing cases or OPT-A with colleagues; further, the nature of OPT-A (i.e., extensive training and monitoring required; see below) makes likelihood of contamination low.

OPT‑A

The treatment manual is available from the authors. Briefly, OPT-A is a novel approach to clinic-based treatment that draws on the Multisystemic Therapy (MST; Henggeler et al. 2009) framework for problem conceptualization and treat- ment planning (e.g., systemic conceptualization, provider accountability for overcoming barriers, iterative hypothesis testing). This framework allows comprehensive conceptual- ization and treatment of multiple problems simultaneously rather than sequentially and guides therapists to integrate specific strategies from evidence-based interventions (e.g., CBT, behavior modification, parenting and communication skills) based on presenting problems of each youth. The OPT-A supervisor leads treatment planning using the MST principles and analytical process (see Henggeler et al. 2009), resulting in individualized regimens such that an identical set of interventions is not delivered to each case. Treatment length is criterion-based and dependent on goals being met. OPT-A sessions occur roughly weekly and average length was 9.7 months (SD 4.5).

Each session included multiple intervention steps based on prioritized needs. OPT-A includes traditional CBT tech- niques (e.g., mood and thought monitoring, identifying and modifying cognitive distortions, relaxation techniques, collaborative session agenda setting, use of homework to reinforce in-session learning; Beck 1995; Friedberg and McClure 2002) tailored to each youth’s presenting problems. In OPT-A, these are usually taught both to adolescents and parents to encourage parents to practice CBT skills with adolescents between sessions and, ultimately, improve gen- eralizability. OPT-A also provides families with extensive behavior modification targeting substance use and exter- nalizing behaviors (e.g., individualized token-economy behavior plans including well-defined rules, rewards, and consequences). For substance use, adolescents submit urine drug screens in session, and parents are trained to administer frequent random screens to sustain monitoring and rewards, similar to procedures described by Henggeler et al. (2011).

Most families receive family communication and parenting training. Therapists have access to modest funds (< $100/ adolescent) to purchase initial rewards for behavior plans. Families can choose to continue medication for ADHD and/or mental health symptoms, and OPT-A therapists are trained and supervised to coordinate medication monitoring with the prescribing physician (e.g., 45% were prescribed medication at the 6-month assessment). A comprehensive treatment engagement protocol is implemented, based on the work of McKay (2000) on service provision for outpatient clinics. Therapists leverage technology (e.g., some sessions conducted by phone, treatment materials sent via email), and extensive homework protocols are implemented.

OPT-A was conducted by two master’s level clinicians employed by the CAFS program. These therapists expressed interest in serving as study therapists and were chosen by their clinic director. Therapists were trained and supervised in OPT-A by the investigators. Each therapist carried case- loads of up to 12 active clients, and only treated clients receiving OPT-A as part of this study. They received 65 h of initial training on OPT-A, followed by intensive supervi- sion on 2 pilot cases prior to treating study cases. Half-day booster trainings were provided quarterly. Individual weekly supervision occurred using web-conferencing. Adherence was ensured using a system similar to MST’s intensive quality assurance procedures (Henggeler et al. 2009; Smith- Boydston et al. 2014), except that supervision was individual rather than group. Highly structured session notes (focused on specific session tasks/agendas, intervention outcomes, and the MST methods for conceptualization, tracking, and planning) were reviewed by supervisors in advance of super- vision. Supervisors also reviewed session audiotapes and materials (e.g., homework, behavior plans, functional analy- ses). Because OPT-A was in Stage II of development, it was premature to develop a parent- or client-reported adherence instrument, typically used at the point when treatment devel- opers are no longer directly supervising treatment. Instead, the ongoing MST methods for achieving adherence (i.e., weekly review of session notes, tapes, and materials fol- lowed by supervision and intensive training by developers) were utilized throughout the trial. No concerns arose regard- ing OPT-A being delivered as intended.

To confirm adherence, observational coding was com- pleted on each youth’s taped sessions (744 session tapes were collected and coded). We used the Therapy Proce- dures Checklist (TPC; Kolko et al. 2009; Weersing et al. 2002), an assessment that includes 62 items representing the most common techniques used by community-based thera- pists in treating youth. TPC subscales represent Behavio- ral, Cognitive, Psychodynamic, and Family Therapy tech- niques. Results from this observational coding indicated that nearly all (97.3%) families received parent manage- ment and other behavioral techniques and most received

1100 Community Mental Health Journal (2021) 57:1094–1110

1 3

family therapy (83.3%) and cognitive (73.0%) techniques. Few (27.6%) received what would be considered psychody- namic techniques. When looking at the percent of sessions in which techniques occurred (averaged within category), the most frequently occurring techniques were family ther- apy (M = 48.8%), behavioral (M = 36.9%), and cognitive (M = 34.3%), while psychodynamic techniques were infre- quent (M = 6.1%). Thus, observational coding, in addition to the intensive oversight methods used in this Stage II trial, provided assurance of therapist adherence to the primary techniques of OPT-A.

Control

Treatment as usual (TAU) was provided in the same CAFS clinic. Therapists in the CAFS provide a wide range of assessment and psychotherapeutic services to children, adolescents, and families presenting with a diverse array of mental health and behavioral concerns. A waitlist control was ethically untenable given the need for treatment, while parallel, sequential, or coordinated treatments for each prob- lem would require attending 2+ treatments. Thus, TAU by master’s level therapists at the same clinic where OPT-A was delivered was selected as an “active placebo” for this Stage II trial. TAU number of sessions varied, although they were typically scheduled weekly and included individual and family sessions. In this study, TAU averaged 7.1 months (SD 5.8).

Data Analysis Strategy

Intention-to-treat analyses were performed using two-level mixed-effects regression models (Raudenbush and Bryk 2002) implemented in HLM (v7.01; Raudenbush et  al. 2013). The substance use outcome was dichotomous, mod- eled using a Bernoulli sampling distribution (logit link), with predicted probabilities computed as exp(β)/[1 + exp(β)]. The other outcomes were modeled as continuous. To test for short- and long-term change across the 18-month follow-up period, the modeling strategy was informed by a descrip- tive, longitudinal procedure. Specifically, following the rec- ommendation of Hedeker and Gibbons (2006), “spaghetti plots” were used to illustrate the observed trajectory for each participant on each outcome. This revealed consider- able variation in the patterns of change over time, and as a result, the usual strategy of testing linear or curvilinear change was poorly suited. Theoretical considerations (see “Discussion”) also support the selected approach. Thus, the models were specified with a series of level-1 dummy-coded time contrasts to differentiate measurements at month-3, month-6, month-12, and month-18 from those at month-0 (i.e., baseline). This conforms to a mixed-effects formula- tion of a repeated measures ANOVA model that, instead

of testing a specific pattern of change over time, identifies whether there was significant change and when the change occurred (Hedeker and Gibbons 2006). Intervention condi- tion (0 = TAU, 1 = OPT-A) was entered as a main effect at youth/parent-level (i.e., level-2), along with cross-level inter- actions between condition and the month indicators. With this formulation, the results reflect (a) TAU’s change from baseline to each later month and (b) the difference in change between OPT-A and TAU. Control variables for youth age and sex (0 = male, 1 = female) were entered as main effects and interactions with the month indicators. To facilitate interpretation, the controls were grand mean centered so that the resulting estimates reflect youth of average age, weighted for the sample proportion of females-to-males.

Results

Baseline Clinical Problems

As expected in community mental health centers, external- izing problems were common, and both groups’ baseline CBCL T-scores were above the clinical threshold (OPT- A: mean = 73.1, SD 10.4; TAU: mean = 74.1, SD 11.1), with 89% exceeding the borderline clinical threshold. In an attempt to ensure a sample that had comorbid problems specifically, we had the mental health center identify partici- pants with an emphasis on substance use and internalizing problems. Although all cases were identified by the clinic as having comorbid substance use and internalizing prob- lems at baseline, the degree of active substance use at base- line was moderate based on research-collected toxicology screens (OPT-A = 41%, TAU = 35%). Study referrals were also identified as having internalizing problems by the clinic, yet baseline research assessments (RCADS) indicated low levels of internalizing symptoms. Specifically, both study groups fell below the borderline clinical-level T-score of 65 (OPT-A: mean = 54.4, SD 15.5; TAU: mean = 47.6, SD 13.2), with only 17% reaching borderline level. This also impacted the randomization procedure; that is, with the vast majority of youth unexpectedly falling below 65, internaliz- ing status was not a meaningful factor for the urn randomiza- tion procedure (which balanced groups based on T ≥ 65). In sum, research measures suggested substance use and inter- nalizing problems were seen in only 6 cases who met both criteria and only 54 who met either, although clinic staff indicated both problems existed across all cases (in addition to potential externalizing problems).

Treatment Effects on Clinical Outcomes

Results for the clinical outcomes are provided in Table 1.

1101Community Mental Health Journal (2021) 57:1094–1110

1 3

Substance Use

At baseline, the OPT-A and TAU groups did not differ sig- nificantly on substance use (41% vs 35%). From baseline to month 3, the groups differed in the amount they changed over time, with OPT-A decreasing to 29% and TAU increas- ing to 56% (OR 0.25). From baseline to month 6 and month 12, the groups did not differ in the change in substance use. From baseline to month 18, substance use increased in both

groups but the levels of substance use increased significantly less in the OPT-A group compared to the TAU group (52% vs 81%; OR 0.19).

Internalizing Symptoms

At baseline, OPT-A had significantly higher levels of internalizing (54.5 vs 47.5), but the baseline scores for both groups fell below even the borderline clinical range,

Table 1 Mixed-effect regression model results for treatment effects on clinical outcomes

Models controlled for Sex and Age at each time point. Estimates for months 3, 6, 12, and 18 reflect changes from month 0 (i.e., baseline). The t-ratio test statistic was computed as β/SE a OR 1.31 [0.56, 3.04] b OR 0.25 [0.08, 0.75] c OR 0.49 [0.14, 1.64] d OR 0.26 [0.07, 1.00] e OR 0.19 [0.04, 0.96]

Outcome β SE df p 95% CI

Substance use  Month 0: TAU − 0.62 0.32 130 .057 [− 1.25, 0.01]

OPT-A vs TAU 0.27a 0.43 130 .540 [− 0.57, 1.11]  Month 3: TAU 0.87 0.38 256 .024 [0.13, 1.61]

OPT-A vs TAU − 1.39b 0.56 256 .013 [− 2.49, − 0.29]  Month 6: TAU 0.86 0.45 256 .055 [− 0.02, 1.74]

OPT-A vs TAU − 0.72c 0.62 256 .242 [− 1.94, 0.50]  Month 12: TAU 0.65 0.49 256 .184 [− 0.31, 1.61]

OPT-A vs TAU − 1.35d 0.69 256 .053 [− 2.70, 0.002]  Month 18: TAU 2.08 0.66 256 .002 [0.79, 3.37]

OPT-A vs TAU − 1.65e 0.82 256 .046 [− 3.26, − 0.04] Internalizing symptoms  Month 0: TAU 47.52 1.60 130 < .001 [44.38, 50.66]

OPT-A vs TAU 6.96 2.36 130 .004 [2.33, 11.59]  Month 3: TAU − 5.41 1.36 270 < .001 [− 8.08, − 2.74]

OPT-A vs TAU − 3.33 2.35 270 .157 [− 7.94, 1.28]  Month 6: TAU − 4.96 1.35 270 < .001 [− 7.61, − 2.31]

OPT-A vs TAU − 4.80 2.07 270 .021 [− 8.86, − 0.74]  Month 12: TAU − 6.41 1.81 270 < .001 [− 10.00, − 2.86]

OPT-A vs TAU − 5.08 2.94 270 .085 [− 10.84, 0.68]  Month 18: TAU − 5.64 2.20 270 .011 [− 9.95, − 1.33]

OPT-A vs TAU − 3.30 3.08 270 .284 [− 9.34, 2.74] Externalizing problems  Month 0: TAU 74.16 1.35 130 < .001 [71.51, 76.81]

OPT-A vs TAU − 0.97 1.82 130 .594 [− 4.54, 2.60]  Month 3: TAU − 3.27 1.26 334 .010 [− 5.74, − 0.80]

OPT-A vs TAU − 1.12 1.67 334 .505 [− 4.39, 2.15]  Month 6: TAU − 5.18 1.81 334 .005 [− 8.73, − 1.63]

OPT-A vs TAU − 3.72 2.36 334 .115 [− 8.35, 0.91]  Month 12: TAU − 6.94 1.66 334 < .001 [− 10.19, − 3.69]

OPT-A vs TAU − 0.09 2.29 334 .970 [− 4.58, 4.40]  Month 18: TAU − 7.15 1.85 334 < .001 [− 10.78, − 3.52]

OPT-A vs TAU − 0.77 2.37 334 .747 [− 5.42, 3.88]

1102 Community Mental Health Journal (2021) 57:1094–1110

1 3

indicating that neither group had elevated internalizing problems. From baseline to each later assessment, internal- izing symptoms decreased significantly for TAU, but with one exception, the groups did not differ significantly in their change. From baseline to the 6-month follow-up, internal- izing symptoms decreased significantly more in the OPT-A group (to 44.7) than in the TAU group (to 42.6).

Externalizing Problems

At baseline, externalizing problems did not differ signifi- cantly between the OPT-A and TAU groups (74.2 vs 73.2). From baseline to each later assessment, externalizing decreased significantly for TAU, and there was no significant difference between groups in their change over time.

Treatment Effect on Functional Outcomes

This question was evaluated using the same model formula- tion described for clinical outcomes. Table 2 includes the results limited to variables with a significant treatment effect from parent or youth reports, and complete results are avail- able in online supplemental materials.

Parenting

At baseline, and from baseline to later assessments, OPT-A and TAU did not differ on youth- or parent-reported corporal punishment, inconsistent discipline, positive involvement, or positive practices. Similarly, there were no between group differences in parent reports of poor monitoring/supervision either at baseline or over time. For youth reports, from base- line to month 3, OPT-A decreased significantly more (from 1.44 to 1.16) relative to TAU (from 1.51 to 1.72), which increased. For rules/expectations, there were no differences for youth reports. For parent reports, from baseline to month 3 and month 6, OPT-A increased significantly more (from 2.98 to 3.16 at month 3 and 3.27 at month 6) than did TAU (from 3.15 to 3.00 at month 3 and month 6). For indirect supervision, the groups did not differ at baseline or in their change from baseline to month 6, 12, or 18; however, from baseline to month 3, OPT-A increased more (from 2.60 to 3.02 for parents; from 2.06 to 2.11 for youth) than did TAU (from 2.85 to 2.90 for parents; from 2.19 to 1.76 for youth).

School Enrollment

At baseline, and from baseline to later assessments, OPT-A and TAU did not differ on the log-odds of school enrollment.

Treatment Effects on Therapy Process Outcomes

The next models differed in two ways: First, outcomes reflect therapy process, so assessment occurred during treatment (months 1, 2, and 3). Second, Satisfaction with Treatment, Parental Involvement in Treatment, and Working Alliance were not applicable at baseline. Results are presented in Table 3.

Parental Satisfaction with Treatment

At month 1, OPT-A had significantly higher satisfaction relative to TAU (3.82 vs 3.64), and the groups did not dif- fer in change over time. That is, OPT-A levels were higher at month 1 and this difference in satisfaction was relatively stable over time.

Parental Involvement in Treatment

Similar to parental satisfaction, at month 1, OPT-A had sig- nificantly higher involvement in treatment relative to TAU (3.80 vs 3.29), and the groups did not differ in change over time.

Working Alliance

At month 1, and from month 1 to month 2 and month 3, OPT-A and TAU did not differ significantly on parent- or youth-reported working alliance. There was no initial dif- ference and, like the previous process variables, working alliance scores were relatively stable over time.

Treatment Motivation

At baseline, OPT-A and TAU did not differ on parent (2.16 vs 2.13) or youth (1.80 vs 1.71) reports of motivation/con- fidence. From baseline to month 1, 2, and 3, the change for TAU was not significant. From baseline to month 1, OPT-A and TAU did not differ significantly. From baseline to month 2, OPT-A increased more (to 2.26 for parents; 2.11 for youth) than did TAU (to 1.99 for parents; to 1.63 for youth), and from baseline to month 3, OPT-A had less of a decrease in treatment motivation (to 2.22 for parents) than did TAU (to 1.90 for parents). For motivation to cut down substance use, the groups did not differ at baseline or in change from baseline to later assessments.

Discussion

This experimental trial evaluated a family-based integrated treatment for comorbidity designed to be implemented in a public sector community mental health setting. Nearly

1103Community Mental Health Journal (2021) 57:1094–1110

1 3

Table 2 Mixed-effect regression model results for treatment effects on functional outcomes

Outcome β SE df p 95% CI

Parent reports  Poor monitoring/supervision   Month 0: TAU 1.07 0.07 130 < .001 [0.93, 1.21]

OPT-A vs TAU − 0.08 0.10 130 .442 [− 0.28, 0.12]   Month 3: TAU − 0.08 0.09 334 .328 [− 0.26, 0.10]

OPT-A vs TAU − 0.09 0.11 334 .419 [− 0.31, 0.13]   Month 6: TAU 0.04 0.10 334 .680 [− 0.16, 0.24]

OPT-A vs TAU − 0.07 0.13 334 .569 [− 0.32, 0.18]   Month 12: TAU 0.17 0.13 334 .187 [− 0.08, 0.42]

OPT-A vs TAU 0.05 0.16 334 .775 [− 0.26, 0.36]   Month 18: TAU 0.09 0.13 334 .466 [− 0.16, 0.34]

OPT-A vs TAU 0.18 0.16 334 .237 [− 0.13, 0.49]  Indirect supervision   Month 0: TAU 2.85 0.11 130 < .001 [2.63, 3.07]

OPT-A vs TAU − 0.25 0.16 130 .115 [− 0.56, 0.06]   Month 3: TAU 0.05 0.12 334 .699 [− 0.19, 0.29]

OPT-A vs TAU 0.37 0.16 334 .025 [0.06, 0.68]   Month 6: TAU 0.14 0.13 334 .273 [− 0.11, 0.39]

OPT-A vs TAU 0.16 0.18 334 .372 [− 0.19, 0.51]   Month 12: TAU − 0.28 0.15 334 .061 [− 0.57, 0.01]

OPT-A vs TAU 0.35 0.20 334 .084 [− 0.04, 0.74]   Month 18: TAU − 0.19 0.14 334 .185 [− 0.46, 0.08]

OPT-A vs TAU 0.09 0.20 334 .666 [− 0.30, 0.48]  Rules/expectations   Month 0: TAU 3.15 0.06 130 < .001 [3.03, 3.27]

OPT-A vs TAU − 0.17 0.09 130 .054 [− 0.35, 0.01]   Month 3: TAU − 0.12 0.09 334 .179 [− 0.30, 0.06]

OPT-A vs TAU 0.30 0.12 334 .010 [0.06, 0.54]   Month 6: TAU − 0.15 0.07 334 .042 [− 0.29, − 0.01]

OPT-A vs TAU 0.44 0.10 334 < .001 [0.24, 0.64]   Month 12: TAU 0.05 0.09 334 .602 [− 0.13, 0.23]

OPT-A vs TAU 0.08 0.13 334 .537 [− 0.17, 0.33]   Month 18: TAU − 0.05 0.07 334 .496 [− 0.19, 0.09]

OPT-A vs TAU 0.16 0.11 334 .148 [− 0.06, 0.38] Youth reports  Poor monitoring/supervision   Month 0: TAU 1.51 0.08 130 < .001 [1.35, 1.67]

OPT-A vs TAU − 0.07 0.12 130 .549 [− 0.31, 0.17]   Month 3: TAU 0.21 0.08 269 .013 [0.05, 0.37]

OPT-A vs TAU − 0.49 0.12 269 < .001 [− 0.73, − 0.25]   Month 6: TAU 0.04 0.09 269 .692 [− 0.14, 0.22]

OPT-A vs TAU − 0.12 0.14 269 .404 [− 0.39, 0.15]   Month 12: TAU 0.13 0.14 269 .387 [− 0.14, 0.40]

OPT-A vs TAU − 0.14 0.18 269 .444 [− 0.49, 0.21]   Month 18: TAU 0.21 0.18 269 .222 [− 0.14, 0.56]

OPT-A vs TAU 0.02 0.22 269 .938 [− 0.41, 0.45]  Indirect supervision   Month 0: TAU 2.19 0.13 130 < .001 [1.94, 2.44]

OPT-A vs TAU − 0.13 0.19 130 .494 [− 0.50, 0.24]   Month 3: TAU − 0.43 0.15 269 .004 [− 0.72, − 0.14]

OPT-A vs TAU 0.48 0.20 269 .019 [0.09, 0.87]

1104 Community Mental Health Journal (2021) 57:1094–1110

1 3

all evidence-based psychosocial treatments focus on treat- ing single disorders, yet most adolescents seek treatment at non-specialty clinics and present with multiple comor- bidities. Given the realities of the adolescent outpatient population, the experimental treatment was designed to be implemented by a single therapist for various comorbidity combinations. To replicate typical procedures for commu- nity mental health settings, we relied on clinic staff assess- ments of clinical problems rather than using research assess- ments to determine eligibility, which is more often the case in highly controlled clinical trials. Allowing clinic staff to determine eligibility for receiving OPT-A increased exter- nal validity, because, in real-world clinical settings, such assessments direct treatment plans and interventions. How- ever, this approach resulted in a sample that included ado- lescents who did not meet clinical cut-offs on more rigorous research assessments, which has interesting implications for the study’s results.

The high rate of externalizing problems in this sample was not surprising, given conduct problems are the most fre- quent presenting problem for community mental health cent- ers (Foster et al. 2001). However, the degree of active sub- stance use at baseline was moderate and lower than expected, considering that youth in the study were all identified by clinic staff as needing treatment for substance use. Simi- larly, internalizing symptoms were unexpectedly low in the sample, which is perplexing. Among adolescents, lifetime and past-year prevalence rates are 32% and 25% for anxiety disorders, respectively and 11% and between 7.5 and 13%, respectively, for major depression (Avenevoli et al. 2015;

Kessler et al. 2012; Merikangas et al. 2010; Twenge et al. 2019), and youth in this sample were explicitly identified by clinic staff as needing treatment for internalizing problems. As noted above, this study used clinic staff assessments for study inclusion to prioritize external validity, which is key when conducting community mental health center research. However, other types of research may instead over-prioritize internal validity (e.g., research assessors conduct diagnostic interview screens). Thus, study results will be interpreted in light of these methodological considerations.

Based on the sample recruited, we can draw some conclu- sions about OPT-A’s efficacy for substance use and comor- bid symptoms among typical adolescents presenting to a community mental health center. First, adolescents in both the TAU and OPT-A condition showed reductions in exter- nalizing behaviors, yet scores were still clinically elevated at the end of study participation. This finding mirrors an unfortunate gap in the adolescent treatment field: very few clinic-based treatments have reached evidence-based sta- tus for adolescent conduct problems (McCart and Sheidow 2016). Clearly, more work is needed in developing effective clinic-based interventions for adolescent externalizing prob- lems. Second, internalizing symptoms showed greater reduc- tions in the OPT-A condition compared to TAU, as OPT-A essentially reduced internalizing symptoms to the level of TAU. However, all of these levels (both pre and post) were within the non-clinical range and, thus, these results have muted clinical significance. Therefore, the study procedures (i.e., clinic staff identifying adolescents in need of treatment for internalizing problems) and the resulting sample (i.e.,

Table 2 (continued) Outcome β SE df p 95% CI

  Month 6: TAU − 0.01 0.18 269 .990 [− 0.35, 0.35] OPT-A vs TAU 0.14 0.22 269 .541 [− 0.29, 0.57]

  Month 12: TAU − 0.23 0.23 269 .316 [− 0.68, 0.22] OPT-A vs TAU 0.05 0.29 269 .872 [− 0.52, 0.62]

  Month 18: TAU − 0.03 0.28 269 .922 [− 0.58, 0.52] OPT-A vs TAU − 0.31 0.34 269 .351 [− 0.98, 0.36]

 Rules/expectations   Month 0: TAU 2.29 0.11 130 < .001 [2.07, 2.51]

OPT-A vs TAU 0.04 0.14 130 .774 [− 0.23, 0.31]   Month 3: TAU 0.18 0.11 269 .111 [− 0.04, 0.40]

OPT-A vs TAU 0.01 0.18 269 .941 [− 0.34, 0.36]   Month 6: TAU 0.14 0.17 269 .393 [− 0.19, 0.47]

OPT-A vs TAU 0.27 0.21 269 .194 [− 0.14, 0.68]   Month 12: TAU 0.09 0.15 269 .542 [− 0.20, 0.38]

OPT-A vs TAU 0.33 0.21 269 .109 [− 0.08, 0.74]   Month 18: TAU 0.25 0.19 269 .188 [− 0.12, 0.62]

OPT-A vs TAU 0.08 0.23 269 .733 [− 0.37, 0.53]

Models controlled for Sex and Age at each time point. Estimates for months 3, 6, 12, and 18 reflect changes from month 0 (i.e., baseline). The t-ratio test statistic was computed as β/SE

1105Community Mental Health Journal (2021) 57:1094–1110

1 3

Table 3 Mixed-effect regression model results for treatment effects on therapy process outcomes

Outcome β SE Df P 95% CI

Parental satisfaction with treatment  Month 1: TAU 3.64 0.08 117 < .001 [3.48, 3.80]

OPT-A vs TAU 0.18 0.09 117 .039 [0.003, 0.36]  Month 2: TAU − 0.12 0.07 166 .078 [− 0.26, 0.02]

OPT-A vs TAU 0.11 0.08 166 .174 [− 0.05, 0.27]  Month 3: TAU − 0.12 0.09 166 .185 [− 0.30, 0.06]

OPT-A vs TAU 0.17 0.10 166 .098 [− 0.03, 0.37] Parental involvement in treatment  Month 1: TAU 3.29 0.13 117 < .001 [3.04, 3.54]

OPT-A vs TAU 0.51 0.14 117 < .001 [0.24, 0.78]  Month 2: TAU − 0.04 0.12 165 .729 [− 0.28, 0.20]

OPT-A vs TAU 0.10 0.13 165 .441 [− 0.15, 0.35]  Month 3: TAU − 0.02 0.10 165 .855 [− 0.22, 0.18]

OPT-A vs TAU 0.08 0.12 165 .500 [− 0.16, 0.32] Parental working alliance: emotional bond  Month 1: TAU 2.54 0.11 112 < .001 [2.32, 2.76]

OPT-A vs TAU 0.17 0.12 112 .165 [− 0.07, 0.41]  Month 2: TAU 0.01 0.13 141 .953 [− 0.24, 0.26]

OPT-A vs TAU − 0.03 0.15 141 .846 [− 0.32, 0.26]  Month 3: TAU − 0.03 0.16 141 .875 [− 0.34, 0.28]

OPT-A vs TAU 0.05 0.17 141 .760 [− 0.28, 0.38] Parental working alliance: agreement on tasks/goals  Month 1: TAU 2.55 0.10 112 < .001 [2.35, 2.75]

OPT-A vs TAU 0.15 0.11 112 .192 [− 0.07, 0.37]  Month 2: TAU − 0.14 0.12 141 .246 [− 0.38, 0.10]

OPT-A vs TAU 0.09 0.14 141 .517 [− 0.18, 0.36]  Month 3: TAU − 0.17 0.15 141 .241 [− 0.46, 0.12]

OPT-A vs TAU 0.20 0.15 141 .199 [− 0.09, 0.49] Youth working alliance: emotional bond  Month 1: TAU 2.23 0.11 106 < .001 [2.01, 2.45]

OPT-A vs TAU 0.06 0.15 106 .692 [− 0.23, 0.35]  Month 2: TAU − 0.07 0.12 138 .570 [− 0.31, 0.17]

OPT-A vs TAU 0.07 0.15 138 .630 [− 0.22, 0.36]  Month 3: TAU − 0.01 0.11 138 .940 [− 0.23, 0.21]

OPT-A vs TAU − 0.01 0.16 138 .974 [− 0.32, 0.30] Youth working alliance: agreement on tasks/goals  Month 1: TAU 2.03 0.10 106 < .001 [1.83, 2.23]

OPT-A vs TAU 0.09 0.14 106 .515 [− 0.18, 0.36]  Month 2: TAU 0.04 0.08 138 .573 [− 0.12, 0.20]

OPT-A vs TAU − 0.06 0.11 138 .606 [− 0.28, 0.16]  Month 3: TAU 0.01 0.11 138 .923 [− 0.21, 0.23]

OPT-A vs TAU − 0.07 0.14 138 .609 [− 0.34, 0.20] Parental treatment motivation  Month 0: TAU 2.13 0.09 130 < .001 [1.95, 2.31]

OPT-A vs TAU 0.03 0.12 130 .809 [− 0.21, 0.27]  Month 1: TAU − 0.06 0.09 294 .500 [− 0.24, 0.12]

OPT-A vs TAU 0.06 0.12 294 .633 [− 0.18, 0.30]  Month 2: TAU − 0.14 0.09 294 .141 [− 0.32, 0.04]

OPT-A vs TAU 0.24 0.12 294 .048 [0.004, 0.48]  Month 3: TAU − 0.23 0.12 294 .052 [− 0.47, 0.01]

1106 Community Mental Health Journal (2021) 57:1094–1110

1 3

adolescents with low levels of internalizing problems) lim- ited our ability to make conclusions about OPT-A’s efficacy for internalizing symptoms. Third, for substance use, based on toxicology screens, drug abstinence in TAU shifted from 65% at baseline to 44% at 3-months, 49% at 12-months, and only 19% at 18-months. OPT-A generated better abstinence rates, shifting from 59% (baseline) to 71% (month-3) to 74% (month-12) to 48% (month-18). Aside from comparing absti- nence rates between these two randomly assigned groups, these rates can be benchmarked against normative samples. For example, a 7-point decrease in abstinence is expected over time in the normal adolescent population not in treat- ment (comparable to the 11-point decrease in abstinence for OPT-A), suggesting OPT-A youth generated a normative trajectory while TAU had a 46-point decrease in abstinence (Miech et al. 2016).

Other clinical trials used research-confirmed diagno- ses for study entry, but timepoint analyses in the present study allow some degree of comparison. The Rohde et al. (2014) substance use and comorbid depression trial achieved 23–44% substance use disorder remission (across 3 condi- tions) at week 20 (this trial achieved 44–71% abstinence at month 3 across conditions) and 17–33% at week 72 (this trial achieved 19–48% abstinence at month 18). In the large-scale Cannabis Youth Trial (CYT), 597 adolescents received evi- dence-based interventions (Dennis et al. 2004). About 20% of the CYT sample self-reported no use in the month prior

to baseline and at 12-months self-reported abstinence across six treatment conditions was 17–34% (this trial achieved 49–74% at month 12). OPT-A abstinence rates appear com- parable to or better than those found in other adolescent trials; however, as noted above, participants in other trials met criteria for substance use disorder whereas many in our sample had negative drug screens at baseline, although they had been identified as needing treatment for substance use.

Results also have implications related to community mental health treatment satisfaction and retention. In past studies, direct parent involvement in such treatment was linked to the strongest effects on youth outcomes and reten- tion (Hawley and Weisz 2005). While these factors do not necessarily lead to better clinical outcomes, the current study found higher rates of parental satisfaction and involvement in OPT-A versus TAU. Further, 96% of the adolescents in OPT-A continued in treatment until at least 3 months, a com- mon benchmark for retention success in outpatient treatment (Hser et al. 2001; Rigter et al. 2013), compared to only 65% in TAU.

OPT-A successfully targeted some parenting variables. Youth’s perception of direct and indirect parental supervi- sion improved early in OPT-A compared to TAU, but this difference was not sustained later. Parents’ indirect supervi- sion (e.g., monitoring when youth is not home), as well as effectively using rules and expectations, followed the same pattern. Thus, OPT-A was successful in improving parenting

Table 3 (continued)

Outcome β SE Df P 95% CI

OPT-A vs TAU 0.29 0.14 294 .037 [0.02, 0.56] Youth treatment motivation  Month 0: TAU 1.71 0.09 130 < .001 [1.53, 1.89]

OPT-A vs TAU 0.09 0.12 130 .444 [− 0.15, 0.33]  Month 1: TAU 0.01 0.10 276 .986 [− 0.19, 0.20]

OPT-A vs TAU 0.06 0.13 276 .633 [− 0.19, 0.31]  Month 2: TAU − 0.08 0.11 276 .445 [− 0.30, 0.14]

OPT-A vs TAU 0.39 0.14 276 .005 [0.12, 0.66]  Month 3: TAU 0.04 0.11 276 .741 [− 0.18, 0.26]

OPT-A vs TAU 0.17 0.16 276 .290 [− 0.14, 0.48] Youth motivation to cut down ASU  Month 0: TAU 2.11 0.12 130 < .001 [1.87, 2.35]

OPT-A vs TAU 0.09 0.16 130 .560 [− 0.22, 0.40]  Month 1: TAU − 0.02 0.11 276 .886 [− 0.24, 0.20]

OPT-A vs TAU 0.08 0.15 276 .593 [− 0.21, 0.37]  Month 2: TAU − 0.07 0.13 276 .615 [− 0.32, 0.18]

OPT-A vs TAU 0.01 0.17 276 .968 [− 0.32, 0.34]  Month 3: TAU − 0.17 0.15 276 .253 [− 0.46, 0.12]

OPT-A vs TAU 0.32 0.19 276 .084 [− 0.05, 0.69]

Models controlled for Sex and Age at each time point. Estimates for months 3, 6, 12, and 18 reflect changes from month 0 (i.e., baseline). The t-ratio test statistic was computed as β/SE

1107Community Mental Health Journal (2021) 57:1094–1110

1 3

more rapidly but did not necessarily have a long-term advan- tage over TAU.

While we are unable to make definitive conclusions about the effectiveness of OPT-A for adolescents with substance use and comorbid diagnoses, it is possible that OPT-A would work differently and produce better outcomes for adolescents with more severe problems who meet diagnostic criteria for comorbid problems. In fact, in a pilot study of OPT-A using university-based clinicians to confirm diagnoses of refer- rals, the sample all had (by definition) clinically significant substance use and internalizing problems at baseline, and promising treatment effects were found on substance use disorders and internalizing problems (presentation citation removed for blinding). Thus, the moderate results of the current study should not halt research on such integrated treatments. That said, results of this study point to a much larger population of adolescents who are seeking treatment at community mental health centers and who are identified by clinic personnel as having problems, but who do not meet clinical thresholds using rigorous research-administered cri- teria. These adolescents would rarely be included in clinical trials that have stringent inclusion criteria, but they likely represent the population most commonly treated by com- munity-based mental health providers. We suggest future research to bridge the gap between the important work of developing and testing treatments for the subset of youth with diagnostic levels of substance use and mental health disorders, and providing community mental health clinicians with tools for the more complete range of adolescents who they treat in “real life.”

Overall, this study advances knowledge for adolescent treatment at community mental health centers, particularly for substance use outcomes. Still, we would like to see the magnitude of these outcomes be even better. Thus, instead of using these moderate, albeit statistically and clinically significant, outcomes to argue for further research on OPT- A, we would prefer to focus attention on re-thinking an approach that is more powerful for achieving each clinical outcome. We simultaneously want to encourage the field to focus on interventions that are more feasible for the majority of public sector mental health centers that, unfortunately, have very limited resources. That is, when considering new treatments, we must ensure new approaches do not impede ongoing efforts to narrow the science-practice gap, the public health importance of which has been described in several reviews and is a national priority (Compton et al. 2005; Hogue et al. 2018). In this context, it is important to note that we found OPT-A to require intensive training and quality assurance. While specializing in only one interven- tion—thus requiring multiple clinicians—does not appear to be optimal based on prior research (Hides et al. 2011; Rohde et al. 2014), perhaps a simpler intervention that is broadly applicable to both externalizing and substance use behaviors

would be optimal. For example, family-based Contingency Management (Godley et al. 2014a, b; Henggeler et al. 2011; Stanger et al. 2015) could target both and has been found to be transportable to a range of community providers (e.g., Sheidow et al. 2020).

Results should be considered in light of certain limita- tions. Because the study was designed to maximize exter- nal validity (i.e., relying on clinic staff to determine which adolescents were in need of treatment for substance use and internalizing problems), many participants likely did not meet full diagnostic criteria, and this may have contributed to a lack of treatment effects (Shelef et al. 2005). The issues caused by this limitation, however, were offset somewhat by high external validity, as the methods mirrored typical practices in community mental health settings and resulted in a sample that is likely more representative of families seeking treatment in such settings. Also, racial and ethnicity variability was limited by geographic location and did not allow for examination of differences by these characteristics. Due to logistical realities of conducting a research study at a real-world community mental health center, neither the clinic nor therapists were randomly chosen, which could have introduced some bias. Finally, logistical limitations prevented collection of audio-recordings for TAU therapists or specific characteristics of these therapists; thus, we have limited information about the nature of TAU.

There were two potential limitations related to statistical analyses. First, for OPT-A and TAU, length of treatment was dissimilar. In the reported analyses, treatment length was not controlled; however, supplementary models were conducted and revealed that conclusions did not change when treat- ment length was controlled. Second, to test for change over time, each assessment occasion was compared to baseline. A simpler alternative would have been to test a pattern of change over time (e.g., linear, quadratic); however, this was not supported by the model building steps and also would have eliminated direct comparison to key adolescent treat- ment studies.

Despite limitations, this study had important strengths. It is a large experimental trial with a very lengthy follow- up. Biological indices were used to measure substance use, avoiding potential bias of self-report. Multiple respondents limited source variance problems. External validity was improved with a variety of factors, including relatively few exclusion criteria and use of community-based mental health clinicians delivering treatment with real-world fund- ing mechanisms. Unlike many other substance use studies, the sample was not limited to justice-involved youth (Hogue et al. 2014). Finally, we reported data on key variables of interest beyond primary outcomes.

In summary, we attempted to address an understudied population that is the most common adolescent population presenting to community mental health centers. OPT-A

1108 Community Mental Health Journal (2021) 57:1094–1110

1 3

achieved improvements on some clinical and functional out- comes, and their families were satisfied with the treatment; however, our experiences in this trial suggest that OPT-A may present some barriers for community mental health centers. Like other evidence-based substance use interven- tions for adolescents, OPT-A’s effects were moderate and the treatment required extensive training and support. However, the experiences we report of delivering a highly individual- ized treatment in a real-world clinic-based setting provide unique insight for next steps in adolescent community men- tal health treatment research.

Acknowledgements The authors extend their appreciation to partici- pating youth and families, to the South Carolina Department of Mental Health and Lexington County Mental Health, to Dr. Scott Henggeler for his early mentorship in this work, and to Bothwell Graham, Cath- erine Glen, Lauren Molen, Jennifer Wilson, Alfred Thomas, Jr., and Ligia Navas-Murphy for assistance in implementing the study.

Author Contributions AJS, MRM, and JEC contributed to the study conception and design. Material preparation, data collection and analy- sis were performed by AJS, KZ, MRM, and JEC. The first draft of the manuscript was written by AJS, KZ, MRM, and JEC. TKD substan- tively contributed to writing and reviewing the manuscript, and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Funding This work was supported by the National Institutes of Health (NIH) [Grant Numbers R01DA025616, R21DA017118, K23DA015658, and K23DA034879]. The content is solely the respon- sibility of the authors and does not necessarily represent the official views of the NIH.

Data Availability Available from the corresponding author upon request.

Compliance with Ethical Standards

Conflict of interest The authors declare that they have no conflict of interest.

Ethical Approval IRBs at Medical University of South Carolina and the state Department of Mental Health approved procedures.

Informed Consent Informed consent was obtained from all individual participants included in the study.

References

Achenbach, T. M. (1991). Integrative guide for the 1991 CBCL/4-18, TSR, and TRF profiles. Burlington: Department of Psychiatry, University of Vermont.

Armstrong, T., & Costello, E. (2002). Community studies on ado- lescent substance use, abuse, or dependence and psychiatric comorbidity. Journal of Consulting & Clinical Psychology, 70, 1224–1239.

Attkisson, C. C., & Greenfield, T. K. (2004). The client satisfaction questionnaire-8. In M. Maruish (Ed.), The use of psychological

testing for treatment planning and outcome assessment (3rd ed., Vol. 3). Mahwah, NJ: Lawrence Erlbaum Associates.

Avenevoli, S., Swendsen, J., He, J., Burstein, M., & Merikangas, K. R. (2015). Major depression in the National Comorbidity Sur- vey—adolescent supplement: Prevalence, correlates, and treat- ment. Journal of the American Academy of Child & Adolescent Psychiatry, 54, 37–44.

Aviram, R. B., Rhum, M., & Levin, F. R. (2001). Psychotherapy of adults with comorbid attention-deficit/hyperactivity disorder and psychoactive substance use disorder. Journal of Psycho- therapy Practice and Research, 10(3), 179–186.

Bakker, M. J., Greven, C. U., Buitelaar, J. K., & Glennon, J. C. (2017). Practitioner Review: Psychological treatments for chil- dren and adolescents with conduct disorder problems–a system- atic review and meta-analysis. Journal of Child Psychology and Psychiatry, 58(1), 4–18.

Beck, J. S. (1995). Cognitive therapy: Basics and beyond. New York: Guilford Press.

Brannan, A. M., Sonnichsen, S. E., & Hef linger, C. A. (1996). Measuring satisfaction with children’s mental health services: Validity and reliability of the satisfaction scales. Evaluation and Program Planning, 19(2), 131–141.

Brewer, S., Godley, M. D., & Hulvershorn, L. A. (2017). Treating mental health and substance use disorders in adolescents: What is on the menu? Current Psychiatry Reports, 19(1), 5.

Brown, G., Dishion, T., & Kavanagh, K. (1991). Monitoring (Tech. Rep. No. 102). Eugene, OR: Oregon Social Learning Center.

Chorpita, B. F., Yim, L., Moffitt, C., Umemoto, L. A., & Francis, S. E. (2000). Assessment of symptoms of DSM-IV anxiety and depression in children: A revised child anxiety and depression scale. Behaviour Research and Therapy, 38(8), 835–855.

Chorpita, B. F., Moffitt, C. E., & Gray, J. (2005). Psychometric prop- erties of the Revised Child Anxiety and Depression Scale in a clinical sample. Behaviour Research and Therapy, 43, 309–322.

Compton, W. M., Stein, J. B., Robertson, E. B., Pintello, D., Pringle, B., & Volkow, N. D. (2005). Charting a course for health ser- vices research at the national institute on drug abuse. Journal of Substance Abuse Treatment, 29(3), 167–172.

Copeland, W. E., Miller-Johnson, S., Keeler, G., Angold, A., & Cos- tello, E. J. (2007). Childhood psychiatric disorders and young adult crime: A prospective, population-based study. American Journal of Psychiatry, 164(11), 1668–1675.

Couwenbergh, C., van den Brink, W., Zwart, K., Vreugdenhil, C., van Wijngaarden-Cremers, P., & van der Gaag, R. J. (2006). Comorbid psychopathology in adolescents and young adults treated for substance use disorders. European Child and Ado- lescent Psychiatry, 15(6), 319–328.

Curry, J. F., Wells, K. C., Lochman, J. E., Craighead, W. E., & Nagy, P. D. (2003). Cognitive-behavioral intervention for depressed, substance-abusing adolescents: Development and pilot testing. Journal of the American Academy of Child and Adolescent Psy- chiatry, 42(6), 656–665.

Danielson, C. K., Adams, Z., McCart, M. R., Chapman, J. E., Sheidow, A. J., Walker, J., et  al. (2020). Safety and efficacy of exposure-based risk reduction through family therapy for co-occurring substance use problems and posttraumatic stress disorder symptoms among adolescents: A randomized clinical trial. JAMA Psychiatry, 77, 574–586.

David-Ferdon, C., & Kaslow, N. J. (2008). Evidence-based psycho- social treatments for child and adolescent depression. Journal of Clinical Child and Adolescent Psychology, 37, 62–104.

Dennis, M. L., Godley, S. H., Diamond, G., Tims, F. M., Babor, T., Donaldson, J., et  al. (2004). The Cannabis Youth Treatment (CYT) study: Main findings from two randomized trials. Jour- nal of Substance Abuse Treatment, 27(3), 197–213.

1109Community Mental Health Journal (2021) 57:1094–1110

1 3

de Ross, R. L., Gullone, E., Chorpita, B. F. (2002). The Revised Child Anxiety and Depression Scale: A Psychometric Investigation with Australian Youth. Behaviour Change, 19(2), 90–101

Donohue, B., & Azrin, N. H. (2001). Family behavior therapy. In E. F. Wagner & H. B. Waldron (Eds.), Innovations in adolescent sub- stance abuse interventions (pp. 204–227). New York: Pergamon.

Evans, A. S., Spirito, A., Celio, M., Dyl, J., & Hunt, J. (2007). The relationship of substance use to trauma and conduct disorder in adolescent psychiatric population. Journal of Child and Adoles- cent Substance Abuse, 17(1), 29–49.

Foster, E. M., Kelsch, C. C., Kamradt, B., Sosna, T., & Yang, Z. (2001). Expenditures and sustainability in systems of care. Journal of Emotional and Behavioral Disorders, 9(1), 53–62.

Friedberg, R. D., & McClure, J. M. (2002). Clinical practice of cogni- tive therapy with children and adolescents: The nuts and bolts. New York: Guilford Press.

Friedman, A. S., Terras, A., & Kreisher, C. (1995). Family and client characteristics as predictors of treatment outcome for adolescent drug abusers. Journal of Substance Abuse, 7, 345–356.

Gaudiano, B. A., & Miller, I. W. (2006). Patients’ expectancies, the alliance in pharmacotherapy, and treatment outcomes in bipolar disorder. Journal of Consulting and Clinical Psychology, 74, 671–676.

Godley, M. D., Godley, S. H., Dennis, M. L., Funk, R. R., Passetti, L. L., & Petry, N. M. (2014a). A randomized trial of assertive continuing care and contingency management for adolescents with substance use disorders. Journal of Consulting and Clinical Psychology, 82(1), 40–51.

Godley, S. H., Hunter, B. D., Fernández-Artamendi, S., Smith, J. E., Meyers, R. J., & Godley, M. D. (2014b). A comparison of treat- ment outcomes for adolescent community reinforcement approach participants with and without co-occurring problems. Journal of Substance Abuse Treatment, 46(4), 463–471.

Goldstein, B. I., Goldstein, T. R., Collinger, K. A., Axelson, D. A., Bukstein, O. G., Birmaher, B., & Miklowitz, D. J. (2014). Treat- ment development and feasibility study of family-focused treat- ment for adolescents with bipolar disorder and comorbid sub- stance use disorders. Journal of Psychiatric Practice, 20(3), 237–248.

Grella, C. E., Hser, Y. I., Joshi, V., & Rounds-Bryant, J. (2001). Drug treatment outcomes for adolescents with comorbid mental and substance use disorders. The Journal of Nervous and Mental Dis- ease, 189(6), 384–392.

Groenman, A. P., Janssen, T. W., & Oosterlaan, J. (2017). Childhood psychiatric disorders as risk factor for subsequent substance abuse: A meta-analysis. Journal of the American Academy of Child & Adolescent Psychiatry, 56(7), 556–569.

Hawkins, E. H. (2009). A tale of two systems: Co-occurring mental health and substance abuse disorders treatment for adolescents. Annual Review of Psychology, 60, 197–227.

Hawley, K. M., & Weisz, J. R. (2005). Youth versus parent working alliance in usual clinical care: Distinctive associations with reten- tion, satisfaction, and treatment outcomes. Journal of Clinical Child and Adolescent Psychology, 34(1), 117–128.

Hedeker, D., & Gibbons, R. D. (2006). Longitudinal data analysis. Hoboken, NJ: Wiley.

Henggeler, S. W., Schoenwald, S. K., Borduin, C. M., Rowland, M. D., & Cunningham, P. B. (2009). MST for antisocial behavior in children and adolescents (2nd ed.). New York: Guilford Press.

Henggeler, S. W., Cunningham, P. B., Rowland, M. D., Schoenwald, S. K., Swenson, C. C., Sheidow, A. J., et al. (2011). Contingency management for adolescent substance abuse: A practitioner’s guide. New York: Guilford Press.

Hides, L., Carroll, S., Catania, L., Cotton, S. M., Baker, A., Scaffidi, A., & Lubman, D. I. (2010). Outcomes of an inte- grated cognitive behaviour therapy (CBT) treatment program

for co-occurring depression and substance misuse in young peo- ple. Journal of Affective Disorders, 121(1), 169–174.

Hides, L. M., Elkins, K. S., Scaffidi, A., Cotton, S. M., Carroll, S., & Lubman, D. I. (2011). Does the addition of integrated CBT and MI improve the outcomes of standard care for young people with comorbid depression & substance misuse? The Medical Journal of Australia, 195, S31–S37.

Higa-McMillan, C. K., Francis, S. E., Rith-Najarian, L., & Chorpita, B. F. (2016). Evidence base update: 50 years of research on treatment for child and adolescent anxiety. Journal of Clinical Child and Adolescent Psychology, 45(2), 91–113.

Hogue, A., Henderson, C. E., Ozechowski, T. J., & Robbins, M. S. (2014). Evidence base on outpatient behavioral treatments for adolescent substance use: Updates and recommendations 2007–2013. Journal of Clinical Child and Adolescent Psychol- ogy, 43(5), 695–720.

Hogue, A., Henderson, C. E., Becker, S. J., & Knight, D. K. (2018). Evidence base on outpatient behavioral treatments for adoles- cent substance use, 2014–2017: Outcomes, treatment delivery, and promising horizons. Journal of Clinical Child and Adoles- cent Psychology, 47, 499–526.

Horvath, A. O., & Greenberg, L. S. (1989). Development and valida- tion of the working alliance inventory. Journal of Counseling Psychology, 36(2), 223–233.

Hser, Y. I., Grella, C. E., Hubbard, R. L., Hsieh, S. C., Fletcher, B. W., Brown, B. S., & Anglin, M. D. (2001). An evaluation of drug treatments for adolescents in 4 US cities. Archives of General Psychiatry, 58(7), 689–695.

Hulvershorn, L. A., Quinn, P. D., & Scott, E. L. (2015). Treatment of adolescent substance use disorders and co-occurring internal- izing disorders: A critical review and proposed model. Current Drug Abuse Reviews, 8(1), 41–49.

Kaminer, Y., Burleson, J. A., Blitz, C., Sussman, J., & Rounsaville, B. J. (1998). Psychotherapies for adolescent substance abusers: A pilot study. The Journal of Nervous and Mental Disease, 186(11), 684–690.

Kaminer, Y., & Winters, K. C. (2020). Part IV: Co-occurring disor- ders in youth. In Y. Kaminer & K. C. Winters (Eds.), Clinical manual of youth addictive disorders (2nd ed.). Washington DC: American Psychiatric Publishing.

Kazdin, A. E., & Whitley, M. K. (2006). Comorbidity, case complex- ity, and effects of evidence-based treatment for children referred for disruptive behavior. Journal of Consulting & Clinical Psy- chology, 74, 455–467.

Kessler, R. C., Avenevoli, S., McLaughlin, K. A., Green, J. G., Lakoma, M. D., Petukhova, M., et al. (2012). Lifetime comor- bidity of DSM-IV disorders in the NCS-R adolescent supple- ment (NCS-A). Psychological Medicine, 42(9), 1997.

Kolko, D. J., Cohen, J. A., Mannarino, A. P., Baumann, B. L., & Knudsen, K. (2009). Community treatment of child sexual abuse: A survey of practitioners in the national child traumatic stress network. Administration and Policy in Mental Health, 36, 37–49.

Larsen, D. L., Attkisson, C. C., Hargreaves, W. A., & Nguyen, T. D. (1979). Assessment of client satisfaction: Development of a gen- eral scale. Evaluation and Program Planning, 2, 197–207.

Liddle, H. A., Dakof, G. A., Diamond, G., Parker, K., Barrett, K., & Tejeda, M. (2001). Multidimensional family therapy for adoles- cent drug abuse: Results of a randomized clinical trial. American Journal of Drug and Alcohol Abuse, 27, 652–687.

Martin, T. A., Donders, J., & Thompson, E. (2000). Potential of and problems with new measures of psychometric intelligence after traumatic brain injury. Rehabilitation Psychology, 45(4), 402–408.

McCart, M. R., & Sheidow, A. J. (2016). Evidence-based psychosocial treatments for adolescents with disruptive behavior. Journal of Clinical Child and Adolescent Psychology, 45(5), 529–563.

1110 Community Mental Health Journal (2021) 57:1094–1110

1 3

McKay, M. M. (2000). What we can do to increase involvement of urban children and families in mental health services and preven- tion programs. Emotional and Behavioral Disorders, 1, 11–12.

Merikangas, K. R., He, J., Burstein, M., Swanson, S. A., Avenevoli, S., Cui, L., et al. (2010). Lifetime prevalence of mental disorders in U.S. adolescents: Results from the national comorbidity study. Journal of the American Academy of Child & Adolescent Psy- chiatry, 49, 980–989.

Miech, R. A., Johnston, L. D., O’Malley, P. M., Bachman, J. G., & Schlenberg, J. E. (2016). Monitoring the Future national survey results on drug use 1975–2015: Volume I, secondary school stu- dents. Ann Arbor, MI: The University of Michigan Institute for Social Research.

Molina, B., Howard, A., Swanson, J., Stehli, A., Mitchell, J., Kennedy, T., et al. (2018). Substance use through adolescence into early adulthood after childhood-diagnosed ADHD: Findings from the MTA longitudinal study. Journal of Child Psychology and Psy- chiatry, 59(6), 692–702.

Raudenbush, S. W., & Bryk, A. S. (2002). Hierarchical linear mod- els: Applications and data analysis methods (2nd ed.). Thousand Oaks, CA: Sage Publications.

Raudenbush, S. W., Bryk, A. S., & Congdon, R. (2013). HLM 7: Hier- archical linear & nonlinear modeling (version 7.00) [Computer software & manual]. Lincolnwood, IL: SSI.

Riggs, P. D., Mikulich-Gilbertson, S. K., Davies, R. D., Lohman, M., Klein, C., Stover, S. K. (2007). A Randomized Controlled Trial of Fluoxetine and Cognitive Behavioral Therapy in Adolescents With Major Depression, Behavior Problems, and Substance Use Dis- orders. Archives of Pediatrics & Adolescent Medicine, 161(11), 1026

Rigter, H., Henderson, C. E., Pelc, I., Tossmann, P., Phan, O., Hen- driks, V., et al. (2013). Multidimensional family therapy lowers the rate of cannabis dependence in adolescents: A RCT in Western European outpatient settings. Drug and Alcohol Dependence, 130, 85–93.

Rohde, P., Waldron, H. B., Turner, C. W., Brody, J., & Jorgensen, J. (2014). Sequenced versus coordinated treatment for adolescents with comorbid depressive and substance use disorders. Journal of Consulting and Clinical Psychology, 82(2), 342–348.

Rowe, C. L. (2012). Family therapy for drug abuse: Review and updates 2003–2010. Journal of Marital and Family Therapy, 38(1), 59–81.

Sheidow, A. J., McCart, M. R., Chapman, J. E., & Drazdowski, T. K. (2020). Capacity of juvenile probation officers in low-resourced, rural settings to deliver an evidence-based substance use interven- tion to adolescents. Psychology of Addictive Behaviors, 34(1), 76–88.

Shelef, K., Diamond, G. M., Diamond, G. S., & Liddle, H. A. (2005). Adolescent and parent alliance and treatment outcome in MDFT. Journal of Consulting and Clinical Psychology, 73(4), 689–698.

Shelton, K. K., Frick, P. J., & Wootton, J. (1996). Assessment of par- enting practices in families of elementary school-age children. Journal of Clinical Child Psychology, 25(3), 317–329.

Smith-Boydston, J. M., Holtzman, R. J., & Roberts, M. C. (2014). Transportability of multisystemic therapy to community settings: Can a program sustain outcomes without MST services oversight? Child & Youth Care Forum, 43(5), 593–605.

Spence, S. H. (1998). A measure of anxiety symptoms among children. Behavior Research and Therapy, 36, 545–566.

Stanger, C., Ryan, S. R., Scherer, E. A., Norton, G. E., & Budney, A. J. (2015). Clinic- and home-based contingency management plus parent training for adolescent cannabis use. Journal of the Ameri- can Academy of Child and Adolescent Psychiatry, 54, 445–453.

Stephens, J. R., Heffner, J. L., Adler, C. M., Blom, T. J., Anthenelli, R. M., Fleck, D. E., et al. (2014). Risk and protective factors associ- ated with substance use disorders in adolescents with first-episode mania. Journal of the American Academy of Child and Adolescent Psychiatry, 53(7), 771–779.

Substance Abuse and Mental Health Services Administration (SAMHSA). (2018). Key substance use and mental health indi- cators in the United States: Results from the 2017 national survey on drug use and health (HHS No. SMA 18–5068, NSDUH Series H-53). Rockville, MD: SAMHSA.

Tanner-Smith, E. E., Wilson, S. J., & Lipsey, M. W. (2013). The comparative effectiveness of outpatient treatment for adolescent substance abuse: A meta-analysis. Journal of Substance Abuse Treatment, 44, 145–158.

Tomlinson, K. L., Brown, S. A., & Abrantes, A. (2004). Psychiatric comorbidity and substance use treatment outcomes of adolescents. Psychology of Addictive Behaviors, 18, 160–169.

Turner, W. C., Muck, R. D., Muck, R. J., Stephens, R. L., & Sukumar, B. (2004). Co-occurring disorders in the adolescent mental health and substance abuse treatment systems. Journal of Psychoactive Drugs, 36(4), 455–462.

Twenge, J. M., Cooper, A. B., Joiner, T. E., Duffy, M. E., & Binau, S. G. (2019). Age, period, and cohort trends in mood disorder indica- tors and suicide-related outcomes in a nationally representative dataset, 2005–2017. Journal of Abnormal Psychology, 128(3), 185–200.

Warden, D., Riggs, P. D., Min, S.-J., Mikulich-Gilbertson, S. K., Tamm, L., Trello-Rishel, K., & Winhusen, T. (2012). Major depression and treatment response in adolescents with ADHD and substance use disorder. Drug and Alcohol Dependence, 120(1–3), 214–219.

Weersing, V. R., Iyengar, S., Kolko, D. J., Birmaher, B., & Brent, D. A. (2006). Effectiveness of CBT for adolescent depression: A bench- marking investigation. Behavior Therapy, 37, 36–48.

Weersing, V. R., & Weisz, J. R. (2002). Community clinic treatment of depressed youth: Benchmarking usual care against CBT clinical trials. Journal of Consulting & Clinical Psychology, 70, 299–310.

Weersing, V. R., Weisz, J. R., & Donenberg, G. R. (2002). Develop- ment of the therapy procedures checklist: A therapist-report meas- ure of technique use in child and adolescent treatment. Journal of Clinical Psychology, 31, 168–180.

Zlomke, K. R., Lamport, D., Bauman, S., Garland, B., & Talbot, B. (2014). Parenting adolescents: Examining the factor structure of the Alabama Parenting Questionnaire for adolescents. Journal of Child and Family Studies, 23(8), 1484–1490.

Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Community Mental Health Journal is a copyright of Springer, 2021. All Rights Reserved.