Barriers to family and CSO involvement in services exist across ecological systems. A 2024 scoping review by Tham and Solomon (26) of barriers to family involvement in behavioral services for severe mental illness identified barriers embedded in organizational culture and structure, societal attitudes, provider training and education, and relationship dynamics among patients, their families, and providers. Little is known about the barriers to family and CSO involvement in services for YA OUD specifically, although the scant research to date suggests barriers may be similar.

Limited training opportunities for providers to learn techniques to collaborate with family members, along with lack of administrative priorities for providing family services, have plagued behavioral health services (26) and may also impact family involvement in YA OUD services. In one study, YA OUD providers reported inadequate training, education, and experience engaging and collaborating with families, as well as concerns about confidentiality (25). These barriers coincide with programmatic limitations including scarce opportunities for family members to participate in services, lack of provider time for family outreach, limited staff for family groups, and restrictions on holding counseling sessions that involve family members, as well as administrative barriers related to appointment scheduling (25).

Societal attitudes and stigma are other known barriers to family members seeking services and participating in services for their loved ones (27, 28). This is especially true for family members of individuals struggling with OUD. Family members are reluctant to pursue supportive services due to experiences of stigma and shame (29). Concurrently, stigmatized clients and patients are reluctant to invite their family members to participate in services for the same reasons (30).

Even more concerning is evidence that providers may hold negative attitudes and perceptions about family members and the efficacy of family involvement in services, curbing the extent to which they pursue opportunities to collaborate with CSOs. In one study exploring attitudes among YA OUD providers, providers reported seeing less value in collaborating with families for patients over the age of 18 related to assumptions about lower reliance on the family system among this age group (31). Another study assessing experiences of family members found that not only were family members dissatisfied with their involvement in services, but also they felt dismissed and patronized by providers (29).

Finally, poor access to care is a longstanding problem for behavioral health treatment that persists for YA OUD services. Family members experience challenges navigating the treatment system, in particular understanding the available services to support their loved one (29). The emotional toll of caring for family members struggling with OUD (32) and lack of knowledge about services available to support themselves (29) further impairs access to care for CSOs.

We conducted qualitative interviews with YAs and their CSOs. YAs were recruited from two OUD treatment centers housed under one drug and alcohol treatment program in a large metropolitan area. Both treatment clinics offered opportunities for CSO participation in family education and support groups, check-ins with the treatment team, and family sessions when patients expressed desire to have family members involved and provided consent for counselors to contact them. All study protocols were approved by the governing IRB.

We used convenience sampling to recruit 25 YAs over a 13-month period (November 2022 to December 2023). At each treatment center, an in-house research assistant identified and referred to the study team patients between the ages of 18 and 30 who were currently or previously enrolled in the center's MOUD treatment program. Patients were required to be able to speak and understand English and provide informed consent. YAs who consented to participate completed a brief demographic and opioid use history survey via Qualtrics and an in-person qualitative interview during a subsequent site visit by the research team. At the end of each interview, YAs were asked for permission to contact their CSO whose role in treatment had been discussed as part of the interview. CSOs were contacted by and interviewed over a 14-month period (December 2022 to January 2024). Twenty of the 25 YA participants gave permission to contact their CSO for a follow-up interview and provided contact information. We were unable to reach 12 CSOs using the contact information provided by the YA. Because the YA's participation in the study had ended, we did not recontact YAs for new or additional contact information for their CSO. Thirteen CSOs (65%) were reached by telephone and expressed interest in participating in the study, however 3 CSOs were subsequently unable to be reached for consent and enrollment. The 10 remaining CSOs consented to participate, completed a brief demographic and substance use and treatment history questionnaire via Qualtrics, and were interviewed via video conference. All YA and CSO interviews were conducted by the same two trained research assistants (authors SD and MW). Interviews lasted ~45 min and YAs and CSOs each received a $40 Amazon gift card voucher as compensation for interview completion.

Semi-structured interview guides were used for qualitative interviews for both YAs and CSOs (see Appendix A for CSO sample). The interview guides consisted of open-ended questions and prompts designed to lead the interviewer and participant through different topics and content areas while allowing each participant space to tell their unique story. Questions explored domains relevant to OUD services and common barriers to family involvement in behavioral services: (1) past and current treatment experiences, (2) perspectives on treatment goals including MOUD treatment, (3) attitudes and beliefs about family involvement in services and recovery planning for YA OUD, (4) family relationships and dynamics including the relationship between CSOs and YAs, and (5) barriers to family participation in services. All interviews were audio recorded and transcribed verbatim by the study team. All identifiers were removed prior to coding. For this study we used a random name generator of the most common names in the United States to create gendered pseudonyms for participants to conceal identifiers (e.g., race/ethnicity).

We used mixed deductive-inductive qualitative content analysis procedures (33) with team-based codebook refinement (34). The coder cohort consisted of 6 co-authors. Four identified as White, and two as Multirace. Four identified as cisgender female, one as cisgender male, and one as non-binary. Three had master's degrees and one had a doctoral degree. The lead qualitative interviewer (author SD) created deductive codebooks based on the questions asked and responses commonly collected during interviews. The study team reviewed the proposed codebooks and engaged in an iterative process to refine codes and definitions for clarity and precision before beginning to code.

Coding started with group consensus coding, whereby each member of the coding team independently coded a study transcript and participated in a live coding group to discuss codes and reach agreement. Next, we double-coded study transcripts and continued live coding groups to reach agreement. During live meetings, coders completed a detailed segment-by-segment review of the entire transcript as well as the application of codes in each segment. We continued iterative refinement of existing codes and definitions while also using inductive qualitative content analysis procedures to add new codes and definitions to the codebooks to reflect unanticipated emerging themes. After intercoder reliability was established, coding continued independently with weekly coding meetings to maintain reliability. All coding was conducted using ATLAS.ti software version 24 (35). Matrix analysis was used as a complementary strategy to synthesize and interpret coding results using a grid that listed interview participants along the x-axis and concepts along the y-axis to show which concepts emerged in each interview as well as the frequency of each concept across the study sample (36). To analyze CSO data in the context of YA data, each column represented a CSO-YA dyad rather than an individual participant. With these analytic procedures, we learned about CSO past and current experiences accessing and participating in services for their loved ones in the context of YA experiences and opinions on related topics.

Coding reliability was focused on internal reliability and the quality, authenticity, and truthfulness of our findings (42). Therefore, we favored techniques to enhance coding trustworthiness rather than quantitative measures of inter-coder agreement, consistent with current recommendations (37). The coding team mitigated potential coding bias during thematic analysis to enhance trustworthiness and credibility in four main ways. First, coders received training on qualitative data analysis. Second, as described above, 20% of transcripts were coded by two or more coders to enhance reliability. Third, two members of the authorship team who were not involved in coding and thematic analysis independently reviewed all themes and supporting quotations. Finally, we followed the Standards for Reporting Qualitative Research [SRQR; (38)].

The coding and authorship teams engaged in team-reflexive discussions to understand each person's position within the research, experiences and beliefs about the content matter, experiences and orientation toward qualitative research, and favored data analysis procedures throughout the design, execution, and reporting of this qualitative study (39).

Participants were 10 YA-CSO dyads. YAs were on average 27 years old (SD = 4.1, range = 20–36; 70% age 20 to 30). Self-identified race was 50% Black or African American (n = 5), 30% White (n = 3), 20% Multirace or another race (n = 2). Twenty percent (n = 2) identified as Latino/a/x, Spanish, or Hispanic. Participants were predominantly male (60%; n = 6). CSOs were five mothers, two grandmothers, two partners, and one aunt. Self-identified race and ethnicity was 50% White (n = 5), 30% Black or African American (n = 3), and 20% Latino/a/x, Spanish, or Hispanic. Participants were predominantly female (80%; n = 8). Three CSOs (30%) reported a history of substance use disorder and of those 2 reported receiving substance use disorder treatment. Demographic data were missing for 2 CSOs.

We identified five main themes related to study aims. Each theme is described below along with exemplar statements drawn from interviews and organized in rows by dyad.

Among the 10 CSO-YA dyads, 8 dyads shared stories centered on the resiliency of their relationship, describing their connection as one that has withstood conflict and challenges associated with substance use and relapse (e.g., stealing; Seth and Barbara, fighting; Megan and Amy). Many YAs and CSOs described feeling more connected with one another having navigated and overcome difficult experiences together (Table 1).

Eight YAs described ways in which family relationships, and particularly the relationship with their CSOs, was a source of motivation for pursing abstinence and recovery goals. These YAs all had CSOs who held compassion for how difficult addiction can be and believed that supporting their loved one was crucial to their treatment and recovery (Table 2).

Every CSO believed family involvement in services for YAs is essential to OUD treatment. When asked, 8 out of 10 CSOs endorsed that families have a responsibility to help and support their loved one in treatment and recovery. Four CSOs elaborated on what it might be like for someone in treatment without family support, speaking to potential YA isolation and hopelessness (Table 3).

Every CSO described personal and relational benefits from participating in their loved one's services and supporting their recovery journey. CSOs shared about experiencing reassurance and relief from receiving updates on treatment progress by participating in family sessions and check-ins with the treatment team (n = 3; e.g., Barbara). Others reflected on achieving more open family communication (n = 3; e.g., Alana and Caitlin) and about personal growth like better understanding addiction, shame, and stigma (n = 3; e.g., Alana). Two CSOs disclosed their own lived experience with addiction and described the mutual benefits to their loved one's recovery and their own (e.g., Janis) (Table 4).

CSO involvement in services and a family approach in general can be conceptualized as a continuum of family involvement interventions with an array of complementary treatment goals (13). CSOs were involved in supporting YAs entering treatment (n = 8) and also further along the continuum (n = 9), participating in individual YA treatment sessions and also family therapy sessions. Facilitating treatment included expressing the importance, urgency, and necessity of attending treatment as well as specific facilitative actions such as bringing the YA to treatment and allowing natural consequences (e.g., withholding money from a YA not attending treatment) (Table 5).

In addition to facilitating treatment, some CSOs participated with YAs in treatment. While only one dyad participated in regular (weekly) family therapy sessions in which relationship dynamics were addressed in addition to individual treatment goals, nine family members participated in the YA's treatment in meaningful ways. Importantly, YAs valued the ways that CSO participation included communication with their therapists or other treatment providers, as this open and bidirectional communication provided accountability, positive reinforcement for treatment adherence, and advocacy when encountering roadblocks. Just as CSOs played important and unique roles in facilitating YA treatment entry, ongoing CSO participation supported ongoing YA treatment attendance and treatment outcomes (Table 6).

All 10 YAs reported being in alignment with their CSOs on treatment goals. Treatment goals most commonly included self-sufficiency and long-term recovery (n = 5, e.g., Helen and Christine). Other goals related to childcare and custody (n = 1), medication as treatment for OUD (n = 1), and housing stability (n = 1) (Table 7).

Despite all CSOs being supportive of treatment, seven mentioned significant challenges or barriers to being involved with their YA's treatment. Barriers included logistical constraints such as having a too-busy daily schedule (n = 4; e.g., Jackie), the treatment center being far away (n = 1; e.g., Christine), and their own physical health issues that made traveling for visits difficult (n = 1; e.g., Karen). CSOs also mentioned emotional challenges such as feeling burnt out in their caregiving role (n = 3; e.g., Jackie), believing that their YA was rejecting the treatment being offered (n = 1), and feeling disappointed and angry in their loved one (n = 1). One CSO mentioned their own lack of knowledge about addiction and the treatment system as a barrier to their involvement (Table 8).

YAs understood their CSOs' feelings and opinions about MOUD treatment, which were either positive or mixed though generally positive. We previously reported on MOUD goal alignment as perceived by YAs themselves, finding that most YAs believed their CSO felt similarly about MOUD as they did (23). Here we provide evidence that YAs perceptions were accurate, even when CSOs held more complex mixed views about MOUD. Nine YAs accurately described their CSO's attitudes toward MOUD; one CSO was not asked about this. For four dyads, CSOs confirmed positive opinions about MOUD. The most cited positive benefit of MOUD by both YAs and CSOs was relieving cravings and withdrawal symptoms to facilitate reducing and stopping opioid use. The remaining five CSOs confirmed mixed feelings about MOUD. These CSOs recognized the positive benefits of MOUD in general and for their YA specifically while holding concerns and worries. YAs were able to identify CSO concerns about MOUD, which were often rooted in unknowns about long-term use, addictive properties of MOUD, and importance to recovery (Table 9).

The strength of this qualitative study lies foremost in the rigorous data collection and analysis procedures. With the vulnerabilities of the participant population and the sensitive nature of the interview topics in mind, interviewers were afforded flexibility to build rapport while the structure of the interview guide ensured all important topics were covered. Data collection was followed by innovative analysis procedures that blended traditional thematic content analysis with matrix analysis, allowing for analysis of key themes and the frequency with which themes and subthemes emerged across family members and dyads.

A main study limitation is possible self-selection bias influencing emerging themes. More than half of the CSOs who were contacted to participate in the current study were unreachable or declined to participate. Is it possible, if not likely, that CSOs who agreed to participate in the study were fundamentally different with respect to their level of involvement in their Yas' treatment and beliefs about the importance of family involvement in treatment.

Future directions include developing and disseminating (1) training resources designed to enhance provider knowledge, skill, and confidence collaborating with family members and inviting their participation in treatment; and (2) family-facing resources on navigating the substance use disorder treatment system. Future research should also prioritize learning more from YA OUD providers about their attitudes, beliefs, and experiences offering services to CSOs.