The YFE program is a telephone-based, health coach intervention tailored to meet the needs of adolescents and their families who need additional help (eg, access to food and housing) or adolescent patients who could benefit from case management. The program goals include improving adolescent patients’ psychiatric appointment attendance and medication adherence while reducing emergency department visits and assisting with crisis management (eg, suicidal ideation). The YFE program was implemented in the winter of 2022 at a large, academic learning health system in the Southeastern United States. All adolescent participants in the YFE program receive mental health services through the academic learning health system regularly meeting with a counselor, therapist, or psychiatrist, which can occur in either outpatient or school-based settings. YFE referrals are placed by patients’ clinicians via the electronic health record, with YFE staff enrolling interested patients and families. Once in the YFE program, patients independently meet with health coaches biweekly, while parents have monthly meetings, which occur for the duration of the 6-month program. One of the 2 health coaches is bilingual, and services are in either Spanish or English, depending upon the preferences of patients and families.

All participants were referred by YFE program staff, with the project team conducting recruitment outreach to explain the purpose and scope of the project and to schedule interviews. Patients and parents were interviewed separately. From July 2023 to October 2023, we conducted semistructured telephone interviews with adolescent patients (n=9, 56%) and parents (n=11, 92%) who participated in the YFE program, as well as video interviews (Microsoft Teams) with clinicians who placed program referrals (n=6, 100%).

This project was approved and deemed quality improvement by the Wake Forest University School of Medicine Institutional Review Board (IRB00088503). We followed the Consolidated Criteria for Reporting Qualitative Research (COREQ; Multimedia Appendix 1) [21]. All participants provided verbal consent prior to being interviewed. To protect the privacy of research participants, all study-related data were deidentified and given a corresponding identifier. No compensation was offered to research participants.

Of those referred, semistructured interviews were completed with 9 (56%) out of 16 patients, 11 (92%) out of 12 parents, and 6 (100%) out of 6 clinicians. While we were successful in interviewing most of the individuals referred, scheduling conflicts and an inability to connect by telephone hindered the recruitment of some potential participants, particularly among adolescent patients. Interviews lasted 15-30 minutes, with all patient and clinician interviews being conducted in English, while 7 (64%) out of 11 parent interviews were in Spanish. Semistructured interview guides were collaboratively developed by the team (Multimedia Appendices 2-5), which included researchers trained in qualitative methods and who had expertise in adolescent mental health. Interview guides explored participants’ perceptions of the YFE program, experiences working with health coaches, suggestions for program changes, and program goals. Demographic characteristics were obtained from the electronic health record and via preinterview questionnaires to provide descriptive information about participants (Table 1). Two researchers with graduate-level training in qualitative methodologies, 1 bilingual female and 1 male, performed all data collection and analysis (AR and JK).

Semistructured interviews (N=26) were audio recorded, transcribed verbatim, deidentified, and analyzed using ATLAS.ti software (ATLAS.ti Scientific Software Development GmbH). Inductive coding methodologies were used to capture emergent themes in the data, with thematic analysis strategies used to systematically organize data [22]. The project team collaboratively developed a codebook, which included 8 parent codes and 20 subcodes, that AR or JK then used to independently code the same 3 transcripts (>10%). Any discrepancies were reconciled, and the codebook was refined as necessary, with the remaining transcripts being independently coded by either AR or JK. Questions that arose during the analysis were presented to the wider team for review. Our sample size was sufficient to reach thematic saturation. Participants did not assist with data analysis, interpretation of findings, or transcription verification processes, as these were handled by the study team.

Interviews were conducted with 26 participants: 9 (56%) out of 16 patients, 11 (92%) out of 12 parents, and 6 (100%) out of 6 clinicians. All patient and clinician interviews were conducted in English, while most parent interviews (7/11, 64%) were conducted in Spanish. Patients were demographically diverse across sex (female: 5/9, 56%; male: 4/9, 44%), race (Black or African American: 3/9, 33%; White: 3/9, 33%; and American Indian or Alaskan Native: 1/9, 11%), and ethnicity (Hispanic: 4/9, 44% and non-Hispanic: 4/9, 44%), with an age range of between 11 and 18 (mean 13.4, SD 2.4) years. Participating parents were mostly female (9/11, 82%) and Hispanic (8/11, 73%), with an age range between 31 and 48 (mean 39.5, SD 6.0) years. All clinician participants were non-Hispanic and female; however, they were diverse across race (White: 3/6, 50%; Asian: 2/6, 33%; and American Indian or Alaskan Native: 1/6, 17%); clinical role (psychiatrists: 3/6, 50%; therapists: 3/6, 50%); and years of clinical practice, which ranged from 10 to 21 (mean 18.8, SD 5.12) years. Further demographic data can be found in Table 1.

This study contributes to the literature on telehealth programs, specifically related to mental health, by qualitatively examining key themes that impact patient experiences. Results from this study may help providers interested in programs that use health coaches or digital health programs to tailor inventions that increase patient care. Our key findings are as follows: (1) participants and clinicians valued the convenience of digital appointments; (2) health coaches provide tailored support to meet the needs of patients and parents; (3) coaching equips participants with key mental health skills such as positive thinking, improved communication, and proactive problem-solving; and (4) bilingual and bicultural health coaches can enhance accessibility and effectiveness of mental health services for Spanish-speaking patients. Our findings highlighting the convenience and value of digital mental health interventions are consistent with previous research reporting higher levels of satisfaction among patients [8-10]. Participants cited similar reasons for their preference for web-based services that other literature have highlighted, such as improved accessibility and increased flexibility [8,9].

Participants elaborated more about why they preferred the increased accessibility and flexibility digital mental health services offer. Parents in our study oftentimes balance competing demands such as having one or more jobs, so the flexibility of health coaches to virtually meet after standard business hours made it possible for parents to participate. Of the 11 parents, 36% (n=4) reported having a steady living situation at the time of the interview but worried about maintaining their situation. Therefore, similar programs might benefit from connecting with school officials and leaders, especially at schools with lower SES. Doing so could help families who most need the resources and support that programs like YFE offer. By partnering with school counselors and actively doing outreach in underserved communities, similar programs might better identify and support adolescents and families who may benefit from additional support.

Parents and patients come to mental health programs with a range of needs that may require different levels of support. Therefore, it is important to remain flexible and open when assessing parent and patient needs and determining mode of communication, program duration, and meeting interval frequency. Although video or in-person sessions may not work for every participant, other parents and adolescent patients might benefit from expanding their web-based meetings with health coaches. Embracing video platforms could facilitate the connection between participants and health coaches and increase rapport and engagement. This digital face-to-face interaction could be especially impactful for participants with transportation challenges or difficulty in establishing rapport over the phone. Additionally, participants might benefit from flexibility in program duration and the frequency of meeting intervals to better address parent and patient needs. Some families might necessitate longer-term support beyond the standard 6-month timeframe, while others might benefit from shorter-term support to meet their needs. Also, some families could benefit from meeting more frequently with health coaches beyond the standard monthly meeting for parents and biweekly meetings for adolescent patients.

In this study, we saw an opportunity for our program to further impact populations that may not have had access to mental health services before. Of our parent participants, 73% (8/11) identified as Hispanic and 64% (7/11) were Spanish speakers. Most of the Hispanic and Spanish-speaking parents in our program remarked on navigating the United States as immigrants, with 1 parent describing feeling “all alone” in this country. These sentiments are consistent with studies that have shown how stressors experienced by immigrants, like financial hardship and postmigration difficulties, are associated with mental illness and distress [23,24]. Most of Spanish-speaking parents in the YFE program reported feeling high levels of support from their health coach who spoke Spanish and had the cultural competency to understand their experiences as either Hispanic or an immigrant in the United States. These experiences underscore the pivotal role that bilingual health coaches play in providing culturally sensitive and effective support, which enhances the accessibility and effectiveness of mental health services for Spanish-speaking families. This finding is consistent with other research looking at immigrant populations in countries like Norway and Australia who benefited from mental health providers who shared patients’ cultural background and same language [25-27]. Those studies found that mental health coaches who shared a similar cultural background to patients better equipped them to understand cultural nuances, values, and beliefs which helped foster trust and rapport with patients and led to better communication and understanding [25-27]. Although the YFE program was not designed specifically for Spanish-speaking families, we found that our program was particularly impactful for this population since we had a Spanish-speaking health coach who shared the cultural background of Hispanic families participating in the program.

Both parents and patients cited the strong rapport they developed with their health coach, with 1 parent calling her health coach her “friend” and even her “angel.” Although there is literature that documents the high levels of support and alliance patients can establish with clinicians through digital methods, we were surprised to hear the highly positive language used to describe participants’ relationships with their health coach, especially by parents in our program [11,12]. Therefore, other programs could benefit by increasing behavioral health support for parents as well as adolescent patients. Cultivating more space for parents to honestly share their experiences may prove fruitful for other programs. Increased program content around behavioral health and emotional support could increase program effectiveness, especially for vulnerable populations. Although most of the current literature about digital health coaches specifically works with adult populations [13,14,17,28], this study demonstrates that interventions that work with adolescent patients and parents can facilitate a supportive environment where both feel heard and valued, ultimately enhancing their mental well-being and the parent-child relationship.

By implementing these suggestions for change, similar web-based mental health programs for adolescents and their patients can enhance accessibility, flexibility, and effectiveness, which could possibly improve outcomes and advance the well-being of young patients facing diverse mental health challenges.

The main limitation of our evaluation is selection bias. Our interview sample was composed of parents and patients actively involved in the YFE program and referred to the evaluation team by health coaches. These participants were more likely to have the time and motivation to participate in interviews to share their program experiences. Data collection occurred over the span of 3 months. The YFE program had been in operation for 18 months prior to interview recruitment commencing, with 62 patients having dropped out prior to our evaluation, primarily due to health coaches being unable to reach them for follow-up. Given that interview participation was limited to patients active in the program, we did not collect the experiences of patients and parents who had withdrawn. This may contribute to bias and yield more positive experiences of the YFE program. In spite of these limitations, our qualitative evaluation provided descriptive details of participants’ experiences with the YFE program. Further, since we collected data from patients, parents, and providers, we were able to triangulate our findings to identify the themes that were consistent across all participants, which informed our suggestions.

Web-based health coach programs may play a key role in providing wraparound care to adolescent patients receiving mental health services and their families. Participants routinely cited satisfaction with the YFE program, mentioning specific benefits that could potentially be leveraged by future interventions, such as the convenience of web-based appointments, the importance of tailored support, improved parent-child communication, and the value of speaking with bilingual health coaches in their native languages. Furthermore, the significance of health coach–parent concordance, both pertaining to language and culture competency, was frequently noted as facilitating stronger relationships and may be important to promote both improved outcomes and sustained participant engagement in digital mental health programs. Future research should explore the impact of culture and language on participants’ relationships with mental health professionals, particularly in digital spaces, as these may be essential to providing safe, responsive, and culturally informed care.