Targeting recovery capital for adolescents through treatment for opioid use disorder: a pilot expansion of the Youth Opioid Recovery Support intervention

Introduction

Adolescents with opioid use disorder (OUD) are a critical but underserved population. The current opioid epidemic has had devastating consequences for adolescents and their families, yet this population remains understudied in addiction literature. Fatal overdoses among adolescents aged 14 to 18 in the United States more than doubled between 2019 and 2020, and in 2022 there were 5.2 reported overdose deaths per 100,000 adolescents (1). The widespread availability of illicit fentanyl is an important driver of this increase in mortality, as fentanyl is now involved in at least 75% of adolescent overdose deaths (1). Despite these alarming statistics, treatment capacity and opportunities for adolescents with OUD are notably scarce (2). Medications for OUD (MOUD), primarily buprenorphine and extended-release naltrexone (XR-NTX), are the standard of care for adolescents (3), but the uptake rates are extremely discouraging. In an examination of a large multistate Medicaid claims dataset, only 5% of adolescents diagnosed with OUD received MOUD (4). Furthermore, only one in four residential addiction treatment facilities for adolescents offers buprenorphine (5). Among the adolescents who do receive MOUD, the retention and adherence rates are remarkably low (6). One study found that only two out of 82 youth on XR-NTX received continuous treatment for six months (7). Thus, there is a clear need for the development of adolescent-focused treatment and effective strategies to increase MOUD uptake and adherence in this population.

Another important target for adolescents with OUD is building recovery capital (RC), which refers to the availability and accumulation of resources across multiple ecological levels that support an individual's recovery (8). The RC model (9) accounts for a more thorough understanding of the recovery process than traditional treatment outcomes (e.g., abstinence) by providing a wider snapshot of the contextual factors related to an individual's functioning and the relationships between such factors. Although this framework has shown promise, research utilizing the RC model has primarily focused on adult recovery (8), and there are currently no validated measures of RC for adolescents. Hennessy and colleagues (8) proposed the Recovery Capital for Adolescents Model (RCAM) as a developmentally appropriate adaptation of the RC framework that includes four dimensions: financial, human, social, and community RC (8). Key components in these RC dimensions that translate to youth may include stable living, caregiver income, and transportation access (financial); motivation, health, and school engagement (human); sober friends and environment and supportive family (social); and access to peer groups or recovery sponsors (community). Given the unique vulnerabilities and developmental needs of youth, this framework should be further explored in adolescents to support the development of appropriate treatment programs and improve outcomes in this population (8).

In developing and testing interventions for the critical population of adolescents with OUD, it is important to consider the need for targeted improvements in MOUD uptake and retention, as well as broader RC outcomes that relate to the capacity to build a more flourishing life. Youth Opioid Recovery Support (YORS) is a multicomponent intervention and service delivery model that uses low-barrier access to medication including home delivery, involvement of a parent, family member, and/or other significant-other, motivational incentives in the form of cash gift cards, and assertive outreach and therapy to improve adherence to MOUD and other outcomes. This intervention aims to address the developmental vulnerabilities of youth and has been tested iteratively in multiple trials among young adults age 18–28 since 2018. Several pilot trials have demonstrated and replicated large effect sizes on the primary outcome of MOUD adherence in addition to reductions in opioid use and other positive indicators of treatment and recovery (10–13).

Although YORS has targeted young adults so far, adolescents have many of the same developmental vulnerabilities, and, given their greater reliance on family support, this intervention may be an even better fit for adolescents. YORS has also not been tested for patients taking sublingual buprenorphine (SL-Bup), which is by far the most common medication choice among youth, despite the inherent adherence benefits of injectable formulations (14). Additionally, YORS has not yet been examined using an RC framework. In addition to their positive impact on MOUD adherence, the components of the YORS intervention could directly increase RC through easy access to treatment, transportation assistance, and integrating family support. Furthermore, we posit that MOUD adherence, the primary goal of YORS, could be a key mediator or facilitator of RC development in this population, such that greater MOUD adherence increases the likelihood that adolescents will build RC. MOUD adherence provides adolescents with a baseline level of clinical stability that enables them to work toward the growth of additional human (e.g., promoting health), financial (e.g., finding and maintaining employment), social (e.g., building relationships), and community (e.g., attending recovery support groups and engaging in school) RC.

This report describes our adaptation of YORS for an adolescent OUD population by focusing the individual and family therapy components on adolescent needs and incorporating strategies for SL-Bup adherence monitoring and motivational incentives. Aims of this study included: 1) Testing the preliminary acceptability and feasibility of the YORS intervention for adolescents with OUD seeking treatment with XR-Bup, SL-Bup, or XR-NTX and 2) Measure MOUD adherence outcomes; and 3) Explore changes in RC related variables as early signals of the potential impact of YORS on building RC. This study is limited by the lack of a validated measure of RC and instead explores functional indicators of RC that were available in the data. We hoped to establish the feasibility of the intervention in this population and document a preliminary signal supporting the inclusion of more systematic RC measures in future studies, contributing to the new and emerging area of research on adolescent RC.

Methods

This study was a 16-week uncontrolled feasibility pilot trial to test the YORS intervention for adolescents. This pilot feasibility study was not pre-registered. We recruited two types of participants for this study: 1) Adolescents with OUD, and 2) Treatment Significant Other (TSO) participants nominated by the adolescent patient participant. Adolescent participants were recruited from three community SUD treatment programs during an acute inpatient or outpatient treatment episode between April 2024 and January 2025 under a single provider organization in central Maryland, USA. Key inclusion criteria included: 1) English-speaking, 2) age 13–17 years (inclusive), 3) had OUD, 4) anticipating initiating sublingual buprenorphine (SL-Bup), extended-release buprenorphine (XR-Bup), or extended-release naltrexone (XR-NTX) OR had already initiated SL-BUP within the past two weeks, and 5) willing to designate at least one person as their TSO. The study excluded patient participants with a medical or psychiatric instability or living situation that, in the opinion of the PI, would preclude participation in the trial. Key inclusion criteria for TSO participants were: 1) English-speaking, 2) 18 years of age or older, and 3) parent, legal guardian, or other support person designated by the adolescent participant to be involved as a TSO. TSOs were only excluded if they were known to be sharing drugs with the adolescent participant. The WCG IRB approved the study procedures in March 2024 (20193366), and all participants provided informed consent, including adolescent assent and parental consent for adolescent participation. Research staff contacted a parent or legal guardian with permission from the adolescent patient. Adolescents nominated TSOs for participation and signed a release of information before the research team proceeded with communication about the study, treatment, or urine drug screen results. Participants were compensated for providing study data. Safety was also closely monitored given that participants in this trial were minors. Research staff assessed adolescent participants for adverse events (AEs) biweekly using a standardized checklist and reported severe adverse events (SAEs; hospitalization, overdose, etc) to the NIDA project officer immediately. SAEs were classified as medical or psychiatric events that were life threatening or that resulted in hospitalization, death, disability, or required intervention to prevent permanent impairment. AEs were classified as medical, psychiatric, or social events that caused distress to the participant without meeting SAE severity criteria. This trial, including any reported AEs or SAEs, was reviewed every six months by a Data Safety Monitoring Board.

All adolescent participants were seeking usual care for OUD regardless of the research, and usual care was not provided by the study. Usual care consisted of completion of an acute inpatient stay (if the participant was recruited from inpatient treatment) and facilitated referral to continuing care (for example, long-term residential or intensive outpatient). These outpatient services are heterogeneous, but typically consist of group and individual counseling, appointments with a medical provider, and medication. The degree to which participants engaged in usual care varied widely, with the YORS intervention layered on top of other services. In this trial, adolescents received the YORS intervention for 16 weeks in addition to usual care. We adapted YORS for adolescent patient participants by incorporating strategies that promote adherence to SL-Bup in addition to XR-MOUD, and by catering the therapeutic components toward typical adolescent needs as compared to young adults (e.g., scheduling sessions around school schedule; more significant involvement of TSO). The intervention is primarily delivered outpatient by a mental health therapist and a care coordinator/research assistant. Weekly supervision with the principal investigator was used to promote fidelity to the model (target “dose” of components described above) without formal fidelity checklists or tracking forms, consistent with the pilot nature of this project. Medical personnel are also available to facilitate MOUD but have less direct interaction with study participants. Participants recruited from an acute inpatient program were oriented to the intervention and connected to their study therapist prior to discharge. The core components of the YORS intervention are:

(1) Low-barrier access to MOUD. Participants were encouraged but not required to enroll in standard locally available psychosocial treatment programs concurrently. There was no requirement for in-person, telehealth or any other participation in sessions to obtain XR-Bup, XR-NTX, or SL-Bup, which was prescribed to a local pharmacy of patient choice in supplies from 1 to 4 weeks at a time. All participants were offered transportation via rideshare services, free of charge, to a treatment facility to receive an MOUD injection or to a pharmacy to pick up a prescription for SL-Bup. Research staff also monitored and facilitated medication refills with providers and pharmacy staff, transported medication between treatment facilities if needed, sent patient reminders, and coordinated medication visits. Participants were offered flexibility in scheduling, including same-day appointments or visits outside of regular business hours.

(2) Motivational incentives for medication adherence. Participants were provided with cash gift cards as incentives for taking their MOUD. They were offered a reinforcement schedule that totaled up to $200 for XR adherence or $208 for SL-BUP adherence over the 16-week trial. For patients on XR-MOUD: dose 1 = $20, dose 2 = $25, dose 3 = $30, dose 4 = $35, dose 5 = $40, and the bonus for getting all scheduled doses = $50. Patients on SL-BUP received $2 per day, Monday through Friday, and a $3 weekly bonus for taking at least 4 of 5 scheduled doses. No incentives for SL-BUP were offered on weekends/holidays due to staff limitations. If a participant switched from SL-Bup to XR-MOUD during the intervention period, they started with same dose 1 dollar amount. Participants were only incentivized when MOUD was verified by medical records or direct observation by research staff, clinical staff, or a trained TSO. SL-Bup observation by research staff was required to be eligible for the motivational incentives and was conducted through scheduled video calls, but monitoring was not required to continue MOUD treatment.

(3) Assertive outreach and therapy. All participants were assigned a YORS therapist for one-on-one therapy, family sessions, and frequent outreach by text and phone. The YORS therapist and research staff, essentially acting as care coordinators, contacted patients weekly and TSOs at least biweekly to check in on progress, coordinate care, and send appointment reminders. Therapists aimed to have a minimum of 30 min sessions, but appointments were flexible to accommodate the patient's schedules and needs, with maintaining engagement as the top priority. Study staff tracked sessions and outreach with progress notes and study logs. At the end of the intervention, the YORS therapist helped participants transition to regular continuing care.

(4) Family/TSO involvement. Participants were asked to designate a family member, significant other, or friend to participate in their treatment. TSOs who agreed to participate were connected with the YORS therapist. At the start of the intervention, the therapist collaboratively created a written treatment agreement with the patient and TSO to set goals and expectations. The therapist aimed to hold at least one 30 min family session per month to revisit the treatment agreement, evaluate progress on goals, and problem solve as needed. The therapist also offered the TSO education and coaching on OUD, treatment options, and coping skills via phone, text, and sessions. Group text chats included patient, TSO, therapist and care coordinators. The TSO was also offered additional support resources, including a peer support specialist and weekly family support groups.

Outcome variables and metrics are detailed in Table 1. The primary outcome of this study was the number of days of MOUD adherence during the 16-week intervention period and was assessed using the timeline follow back (TLFB; 15), an interview-based assessment administered biweekly. Self-reported adherence (yes/no) to daily SL-Bup was determined using the TLFB with each day of self-reported adherence contributing to this outcome. Self-reported adherence to XR-MOUD was determined using the TLFB with each day of self-reported injection indicating 28 days of adherence. Corroborating data sources were used to flag discordance in the self-reported data in the following way: for each study week that a study participant reported taking MOUD on at least one day, research staff gathered information from at least one additional data source, when available, to support or contradict that week's self-reported MOUD adherence (no biological data was collected to confirm ingestion of MOUD). Only self-report data was retained in the analysis of the primary outcome. Secondary MOUD information sources are recorded descriptively. Additional data sources included direct observation by study staff, medical records with injection administration or MOUD prescription, or the Prescription Drug Monitoring Program (PDMP), which shows pharmacy dispensations of SL-Bup.

Table 1. Operational definitions and results for primary and secondary outcome variables.

OUD, opioid use disorder; MOUD, medication for opioid use disorder; UDS, urine drug screen; YORS, youth opioid recovery support; TSO, treatment significant other. Cells with a “–” indicate that this was not assessed.

Secondary outcomes included days of opioid use, opioid abstinence, feasibility, acceptability, and change in recovery capital indicators. Days of opioid use and opioid abstinence were determined by self-report data using the TLFB. UDS data were examined as a validity measure of the self-report data to identify weeks of the intervention period with contradictory results. These data were collected using commercially available cups with standard cutoffs. Feasibility was operationalized by the percentage of target enrollment numbers (N = 8 adolescents) and delivery of the components of the YORS intervention. Acceptability was measured by adolescent and TSO responses on a treatment satisfaction survey administered by research assistants. Participants rated their satisfaction with each component on a scale from 1 to 10, compared YORS to past treatment (better, worse, never been in treatment before), and reported satisfaction with the level of outreach (too much, too little, the right amount). Pre/post self-reported indicators of recovery capital were measured at week 1 and week 14 of the intervention. Indicators included housing status, employment, enrollment in school, support group attendance, legal status, and ratings (1–10) of family involvement in the adolescent's recovery. These are not validated measures of RC, rather they are exploratory functional indicators that align with the key components of the dimensions in the RC model. Data from this trial is not made publicly available to protect confidentiality but may be shared upon request.

Data analyses

This study calculated descriptive statistics on all variables. Inferential statistics were not used in the pre/post comparisons for change in RC due to the small sample size. Missing data on MOUD adherence was conservatively imputed as non-adherent. The average rate of missing MOUD adherence data was 24% across the sample. For proportion of opioid use days, we calculated a percentage of days of self-reported use divided by the available self-report data.

Results

N = 11 adolescent patient participants were recruited for the 16-week uncontrolled YORS trial. Regarding participant flow, there were 40 adolescents with OUD who were enrolled in inpatient or outpatient treatment between April 2024 and January 2025 (recruitment period) who were pre-screened for interest and eligibility. Twelve adolescents who met eligibility criteria completed informed consent, resulting in 11 enrollments and one screen fail. Participant baseline characteristics are presented in Table 2. Overall, the sample was primarily African American and/or Hispanic, heterosexual, and male, with a mean age 15.8 years (SD = 0.6), who were smoking fentanyl and used public insurance. All participants selected a parent to participate as the TSO (7 mothers, 4 fathers). Three SAEs (psychiatric hospitalizations) and seven AEs (two medical, two psychiatric, and three social events) occurred during the trial. These events were classified as not related to the study intervention, as supported by PI and DSMB review. The findings for each outcome variable are displayed in Table 1 and individual case vignettes are described in Table 3.

Table 2. Demographic characteristics of adolescent patient sample (N = 11).

Table 3. Adolescent participant case vignettes.

The order of patients reflects the graphs in Figure 1. MOUD, medication for opioid use disorder; SL-Bup, sublingual buprenorphine; NA, narcotics anonymous; OUD, opioid use disorder; TAU, treatment as usual; TSO, treatment significant other; YORS, youth opioid recovery support; XR-Bup, extended-release buprenorphine; XR-NTX, extended-release naltrexone.

MOUD adherence

Participants averaged 78.7 days (SD = 32.9), or a 70.3% rate, of self-reported adherence to MOUD across the 112 days of the YORS intervention (See Figure 1A). There was no discordance flagged across the self-reported data using available corroborating sources, but these sources did not biologically confirm ingestion of MOUD (see methods section for additional detail). There was a 24% rate of missingness for days of MOUD, with missing days conservatively imputed assuming that MOUD was not received. We also calculated the proportion of days on MOUD out of the days on which a participant had self-reported data available, which yielded a 92.4% rate of adherence across the intervention period. At the time of enrollment, seven participants (63.6%) initiated treatment with SL-Bup and four participants (36.4%) initiated treatment with XR-MOUD. Four participants (36.4%) switched to a different MOUD during the intervention period. At the end of the intervention, four participants were on SL-Bup (36.4%) and seven participants were on XR-MOUD (63.6%).

Proportion of days of opioid use and opioid abstinence

Across 11 participants, the average proportion of self-reported days of opioid use across available self-reported data was 4.8% (M = 5.4 days, SD = 15.3 days; see Figure 1B). Two participants reported at least one day of use, while nine participants (81.8%) reported complete opioid abstinence. UDS data indicated that one participant self-reported two weeks of opioid abstinence that were contradicted by a positive UDS. This discordance occurred within the first month of the intervention and did not change the primary outcome reported, which is entirely reliant on self-reported data on the TLFB regardless of UDS results. The rate of missing self-report opioid use data due to loss to follow up was 20% at the end of the 16-week intervention.

Feasibility

The enrollment target for this pilot trial was N = 8 adolescent participants, which was achieved with a total of 11 enrollments and 10 TSOs (90.9% of expected). Eight (72.7%) family-patient dyads engaged in at least one family session and the average number of family sessions across the sample was 1.2 sessions (SD = 0.9). All family members were encouraged to attend a virtual weekly family support group and 2 TSOs attended (18.2%), with one attending 8 groups and one attending 4 groups. Eight (72.7%) of the adolescent patient participants completed at least one individual mental health session with the YORS therapist, defined as an in-person or telehealth session documented in electronic health records, and the average number of individual sessions was 3.3 sessions (SD = 3.9). Across 112 days of the intervention, YORS staff averaged 47.9 days (SD = 20.3) of outreach per patient-TSO dyad. The average amount of monetary incentives dispensed across the adolescent participants was $93.64 (SD = 66.7), which equates to a 45.9% dispensation rate when compared to the total possible amount for each participant.

Acceptability

Treatment satisfaction surveys were completed by seven adolescent participants (63.6%) and five TSOs (50%). Missing data was a result of loss to follow up. Mean adolescent satisfaction with family involvement was 8.9 out of 10 (SD = 1.4) whereas TSOs reported 9 out of 10 (SD = 1) satisfaction with their loved one's recovery. All participants and TSOs with prior treatment episodes rated YORS as better than prior treatment experiences. All adolescents reported that the outreach level was optimal and most TSOs agreed (4 of 5) with one TSO stating that the level of outreach was too much. Adolescents rated the various components of YORS according to their helpfulness to their recovery on a 10-point scale: having a support person involved in their recovery (M = 10, SD = 0), low-barrier access to MOUD (M = 9.7, SD = 0.8), cash incentives for MOUD (M = 9.4, SD = 1.1), taking a relapse prevention medication (M = 8.4, SD = 1.1), individual sessions with therapist (M = 8.3, SD = 2.0), frequent calls and texts from the treatment team (M = 8.3, SD = 1.6), and family sessions (M = 7.3, SD = 2.2). The TSOs also ranked these components: having a support person involved (M = 10, SD = 0), low-barrier access to medication (M = 10, SD = 0), frequent calls and texts from the treatment team (M = 9.6, SD = 0.9), taking a relapse prevention medication (M = 9.4, SD = 1.3), family sessions (M = 9.3, SD = 1.5), cash incentives for MOUD (M = 9, SD = 2.2), and individual sessions with therapist (M = 8.8, SD = 2.2).

Change in other recovery capital indicators

The rate of missing data for RC variables due to loss to follow up was 36% in the adolescent sample and 50% for TSO participants. Of the seven participants with complete demographic data, two (28.6%) reported homelessness within the past year at baseline and only one participant (14.3%) reported homelessness that occurred during the YORS intervention period. Five of the seven participants (71.4%) reported being either in school or employed at baseline and all seven participants (100%) reported being in school or employed at their week 14 follow-up. At intake, none of the adolescent participants (0%) reported that they attended self-help groups (i.e., Narcotics Anonymous). At the week 14 assessment, four participants (57.1%) reported that they had attended at least one group, with two participants (28.6%) still actively attending. Additionally, at baseline, five of the seven participants (71.4%) reported at least one lifetime arrest and two participants (28.6%) reported having current open charges. At the week 14 assessment, one patient (14.3%) reported being arrested during the intervention period but no patient (0%) reported being charged with a new offense. In terms of social RC, patients and TSOs were asked to report how involved the TSO was in their loved one's recovery on a scale from 1 to 10. Adolescent participants reported an average of 9.1 (SD = 1) and TSO participants reported a 9.6 average (SD = 1).

Discussion

The results of this pilot study support the acceptability and feasibility of the YORS intervention for adolescents and encourage further investigation in a controlled trial. Participants self-reported adherence to MOUD on 70.3% of days (on average) across 16 weeks of the intervention period. Six of the 11 participants reported at least 90% adherence (Figure 1), which is well over the 80% standard threshold for good medication adherence (16). These results are encouraging, given the poor larger-scale findings of adolescent MOUD retention (6). We did not conduct statistical analysis to compare our results with those of previous YORS trials in young adults, but the current adolescent sample reported MOUD adherence at or above those of young adults from previous trials, on average (10–13). Although these cross-study comparisons are not equivalent and we cannot determine efficacy of the intervention in this small uncontrolled feasibility pilot, we speculate that adolescents may be more receptive to the strategies and support services provided by the intervention due to their younger age and normative reliance on family and authority figures. These early findings are promising and suggest that strategies based on growing recovery capital (RC) should be investigated in youth OUD treatment on a larger scale.

The secondary findings in this study showed promise as well. Across 11 participants, only two participants reported any opioid use, with the remaining nine reporting complete opioid abstinence. Given that there is generally a high likelihood for opioid relapse among adolescents (17), these findings are encouraging and support further investigation of the YORS intervention in this population. Previous YORS studies on young adults found opioid relapse rates ranging from 53%–61% among those receiving the intervention (10–12), but we did not perform any statistical tests comparing these data. Additionally, the results support the feasibility and acceptability of the YORS intervention for adolescents on XR-MOUD and SL-Bup. We exceeded our target adolescent enrollment goal of N = 8 with a total of N = 11 participants, and we reached 90.9% of our enrollment goal to enroll one TSO for each adolescent participant. Individual components of the intervention, including therapy, family involvement, and assertive outreach, were successfully delivered across the sample (Table 1). Distribution of motivational incentives (approximately 50% of available incentives) was on par with what is commonly reported in the literature (18). It should be noted, however, that individual and family therapy session attendance was low. This was likely due to the overarching principle of YORS that encourages flexible accommodation of the level of care that patients will engage in, including informal check-ins, while the team continues to reengage them in regular therapy. The YORS intervention for adolescents remains a work-in-progress, and future research should incorporate additional strategies to engage patients and families in therapy services. Perhaps engagement in the YORS intervention model is suited better for patients with particular characteristics (e.g., those with higher baseline RC). However, we were not able to test these interactions in this small sample. But, despite the low attendance rate at scheduled longer-duration formal sessions, there were high levels of additional interaction with patients through shorter-duration, impromptu staff outreach from both therapists and care coordinators, and strong ratings of TSO involvement by both adolescent and TSO participants. This discrepancy illustrates how the assertive and comprehensive YORS model may supplement scheduled session-based care and may even suggest that the assertive outreach framework could be more effective in improving adolescent treatment outcomes (i.e., brief encounters, phone calls, text messages, and persistent outreach). It is beyond the scope of this article to elaborate on therapeutic assertive outreach techniques utilized in the YORS model, but these strategies are described in previously published work (12).

Reponses on treatment satisfaction surveys suggested that adolescent and TSO participants were very satisfied with the YORS intervention. Patient-TSO dyads who had experienced treatment before rated the YORS intervention as better compared to prior treatment. Adolescents expressed high satisfaction with their TSO's involvement in their recovery and TSO participants were highly satisfied with their loved one's recovery since starting the intervention. Participants consistently provided high ratings of the individual components of the YORS intervention and expressed being content with the level of outreach by YORS staff. The response rate on these surveys was low, likely due to wavering motivation for engagement among young patients and logistical barriers such as changes in phone access. Despite this limitation, these data support the acceptability of the YORS intervention across adolescents and their families.

Our results also provided preliminary signals that YORS may promote the development of RC. Pre/post comparisons of RC indicators revealed that the number of participants who reported homelessness at the end of the intervention was less than the number of participants who reported homelessness within the previous year at baseline. Additionally, there was an increase in participants enrolled in school or employed and in self-help group attendance. Only one participant reported an arrest during the intervention period, which was encouraging given that 71.4% reported at least one lifetime arrest at baseline. Regarding social RC, adolescent and TSO participants rated the level of TSO involvement in their loved one's recovery a value over nine on a scale from 1 to 10. Given that these measures are not yet validated, and given the high rate of missing data, we cannot conclude that these early findings indicate a meaningful change in RC. However, they encourage further exploration of the YORS intervention to determine its impact on adolescent RC.

Strengths of this study include the innovative YORS model, the diverse sample, and its focus on the critical and understudied population of adolescents with OUD. The YORS model combines multiple known strategies into a comprehensive, patient-centered intervention seeking to enhance MOUD adherence. This innovative design uses developmentally specific approaches that target adherence determinants at multiple levels including patient, family, and provider. Though small, the sample in this study consisted of a diverse group of participants, including male and female adolescents age 15–17 identifying as Hispanic, Black, White, and mixed-race individuals. Furthermore, participants were recruited from both inpatient and outpatient treatment programs and sought treatment with multiple types of MOUD, including SL-Bup, XR-Bup, and XR-NTX. This diverse sample promotes the applicability of the findings, and this research contributes to the very limited literature base on adolescents on MOUD. Expanding our knowledge of adolescent addiction treatment and recovery is critical, given that adolescents are in a key developmental time-period that can significantly impact their life trajectory.

This study also has several important limitations. First, the small sample size limits the generalizability of these findings, and it precluded us from analyzing the pre/post comparisons of RC indicators with inferential statistics. We cannot make conclusions about the efficacy of the YORS intervention in adolescents based on this pilot feasibility study. Additionally, the rate of missing data among the secondary outcomes further limited the impact of these findings. For example, 20% of data was missing for opioid use outcomes. Although it is impossible to know whether participants used opioids when data was not collected, we suspect that at least some of these days may have been days on which participants used opioids. This suggests that the days of opioid use outcome (calculated as a percentage of available data) may be an underestimate of actual opioid use days. The non-randomized design also did not allow for between-group comparisons or conclusions to be drawn regarding the efficacy of the intervention. Although support for outcomes (adherence and opioid use) was sought through additional corroborating data sources, we relied primarily on self-report, which had low response rates for some outcomes. Another limitation is that this trial was not pre-registered. Additionally, this trial was restricted to English-speaking patients and family members. Given the large share of Hispanic people in our patient population, excluding monolingual Spanish-speaking families further limits the generalizability of the study. The lack of Hispanic or Spanish-speaking research staff also introduced noticeable language and cultural barriers that complicated the delivery of the therapeutic and family involvement components of the intervention to this population (e.g., communicating through a professional interpreter service). Patients were also excluded if they met criteria for a psychiatric or medical comorbidity or living situation (i.e., homelessness) that would interfere with their ability to participate, further limiting the generalizability of the findings. Another limitation of this trial was the fixed motivational incentives schedule utilized for SL-Bup adherence, as existing research demonstrates that escalating schedules are more effective in motivating behavior (19). Additionally, the results of the satisfaction surveys could be skewed because participants were provided compensation and may have felt pressure to respond positively. However, this risk was mitigated by having research staff who were not study clinicians administer the survey. Finally, this study did not include a validated measure of RC, as there is not currently a validated measure for adolescents used in addiction research (20).

This study highlights the need for several lines of future research. First, the components of the YORS intervention for adolescents should be formalized and tested in a larger sample for a longer duration of time and should also consider cultural adaptations for Spanish-speaking families. These findings should be used to inform an adequately powered controlled trial in order to determine the efficacy of the intervention among this new population. A larger RCT design that could explore predictors of treatment response to the YORS intervention would be instructive. For example, the MOUD adherence data in this sample shows a clear bi-modal distribution in which half of the adolescents demonstrated excellent adherence (>90%) with the other half struggling (ranging from 25%–53%). Future investigations could seek to understand what characteristics make YORS a good fit for adolescents and their families and contribute to the variability in engagement of the core components. This study also emphasizes the need to develop tools of measuring RC in adolescents and young adults, a need that has been highlighted by experts in the field (21). For example, validated measures of family support and involvement would allow investigators to better research ways to improve this critical facet of RC for young people. These tools would allow for further exploration of the impact of YORS on RC to better examine if the intervention, and promoting MOUD adherence, fosters RC growth. Furthermore, such tools would help clinicians to learn more about their adolescent patients to better inform their care.

In summary, this pilot study confirms the preliminary feasibility of the modified YORS intervention for adolescents and their families, and supports further investigation of this model. These findings widen the potential impact of YORS and serve as a preliminary signal that these strategies may promote the growth of RC. This work helps to lay a foundation for future research examining RC growth in adolescents with OUD, and the role of RC in improving treatment outcomes in this critical population. This study additionally highlights the need to improve our understanding and develop appropriate measures of youth RC, and to expand efforts to promote treatment retention in the adolescent population.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher's note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Jo-Hanna H Ivers, Trinity Biomedical Sciences Institute, Trinity College Dublin, Ireland

Reviewed by: Vincenzo Maria Romeo, Specialization school in Psychoanalytic and Groupanalytical Psychotherapy, Italy

Emily A. Hennessy, Harvard Medical School, United States