Engaging young adults experiencing early psychosis in specialized mental health services is critical for promoting mental health recovery (
1,
2), but even in specialty settings at least 30% drop out within the first year (
3). Early treatment dropout can contribute to poor recovery outcomes, including rehospitalizations, homelessness, interpersonal violence, and suicide (
4–
6). The best ways to engage young adults remain unclear (
3,
7–
11).
There is a growing literature on “pathways to care” (one’s pathway to the initial evaluation for psychosis), but little research on “pathways through care,” or how young adults make the decision to engage in care beyond initial contact and so move toward recovery (
12,
13). We need to understand the earliest patient-cited barriers to treatment (
14), especially for young adults from racial-ethnic minority groups, whose pathways to care are often negative and complex (
15). [A discussion of pathways to or through care for patients from racial-ethnic minority groups is available in an
online supplement to this article.] More longitudinal, qualitative research is needed on how to engage young adults from minority groups and their key supporters in care (
8,
10). A key supporter is someone identified by young people as crucial in supporting their mental health recovery.
This prospective, longitudinal, and ethnographic study was the first that we know of to engage young adults from predominantly racial-ethnic minority groups who were diagnosed as having a psychotic disorder and their self-identified key supporters in research at multiple time points across a critical period for engagement—the 12 weeks during and after one’s initial hospitalization—to examine real-time decision making between first hospital admission for early psychosis and treatment engagement or dropout (
10,
16). [A discussion of qualitative studies in research on first-episode psychosis and a discussion of the “critical period.” is available in the
online supplement.] We sought to generate hypotheses about key treatment decision-making factors to improve mental health service engagement.
Methods
This study used ethnographic methods, which are a promising tool for mental health services [see the
online supplement for a discussion of ethnography and mental health services research]. Ethnography generates hypotheses rather than tests hypotheses (
17–
19). Ethnographers spend a lot of time with participants to develop holistic insights (
20–
22). Ethnographers also prioritize data that appear repeatedly across multiple time points and sources to increase the trustworthiness of findings (
23,
24).
Participants were recruited from 2014 to 2016 from a psychiatric emergency hospital serving patients with any insurance (including no insurance) in a seven-county area of North Texas. The hospital served 23,000 patients per year and an estimated two youths with early psychosis per week in 2014. The facility could hold patients up to 72 hours involuntarily. Some participants were transferred to other inpatient care facilities for longer-term stays, such as a state hospital. None of the participants in our study qualified for a fledgling local early intervention service that only accepted indigent patients who were not eligible for Medicaid or Medicare and were living below 200% of the federal poverty level.
Our sample was purposive (
18). Eligibility criteria for young adults included being age 18 to 30 with no history of a psychiatric hospitalization for a psychotic disorder earlier than the 3 prior months, no history of antipsychotic medication usage for more than three months, and a diagnosis of a nonaffective psychotic disorder. Diagnoses and age were confirmed by medical record and with the treatment team. The team prioritized recruiting young adults from racial-ethnic minority backgrounds. Key supporters were age 18 or older.
Treatment staff identified potential participants. When they were stable, we told them about the study, obtained written informed consent (as overseen by the North Texas Institutional Review Board at Medical City Dallas and the Southern Methodist University Institutional Review Boards), requested names and phone numbers of potential key supporters, administered a demographic survey, and engaged them in an initial audio-recorded interview. Once the young adult was released from the hospital, we engaged him or her in at least three follow-up visits [see the
online supplement for a description of the study timeline]. Two team members observed participants’ everyday routines, activities, and conversations, which were then recorded in field notes (
25,
26). We also engaged participants in 45- to 75-minute, open-ended, person-centered interviews (
27). Young adults received up to $180 in compensation across the interviews; key supporters received up to $80. Professional transcriptions were obtained quickly after each interview for ongoing analysis. In weekly lab meetings, we reviewed interview transcripts and field notes to develop person-specific follow-up probes for subsequent interviews. This “member-checking” technique has been shown to increase validity (
18,
28,
29). We also documented research team memos about iterative patterns we observed across visits and across participants.
Interview transcripts were uploaded into Dedoose and coded [see the
online supplement for a description of the coding techniques]. Once coded, text was easily extracted, aligned, and compared across transcripts, visits, and participants (
28,
30). Consistent with constructivist grounded theory methods, the team reviewed the coding analysis, field notes, and lab meeting research memos to look for iterative patterns across the data and generate hypotheses about key factors influencing treatment decision making (
18). We then created a table showing relevant factors in order of their frequency.
Results
Most young adults (more than 100) who were approached refused to participate. Thirty-seven young adults engaged in at least one interview. Nineteen were lost to follow-up before the 12-week period ended. [Demographic information for all recruited participants is presented in the online supplement.]
Eighteen young adults completed the study (
Table 1) by participating in three or more interviews during the first 12 weeks and are the focus of this report. Most young adults (N=16, 89%) engaged in four visits and interviews. Among the 18 young adults who completed the study, the mean±SD age was 21.9±3.1 years. Eight were male, and 10 were female. Seven participants were black, eight were Latino, two were white, and one was Asian. Most were at least high school graduates (N=15, 83%). One-half were second-generation immigrants, and three were first-generation immigrants from South Asia, West Africa, and Central America, respectively. At the 12-week point, eight continued to attend mental health appointments, and 10 (55%) had dropped out of care.
Nineteen key supporters of 15 of the young adults who completed the study also participated in at least one 45- to 60-minute interview (
Table 2). Two-thirds of the key supporters were women (N=12) and mothers (N=8, 42%). Five were siblings. Three key supporters identified by young adults were not interviewed because of language barriers or persistent scheduling conflicts.
We ranked emergent factors that young people and their key supporters discussed as being important for treatment decision making (
Table 3). A narrative description of these ranked factors with illustrative quotes is presented below. Each quote is identified by a letter or letters referring to an individual young person (i.e., A, B, C, etc.) or the key supporter (KS) of that individual (e.g., A.KS) and by the week in which the interview was conducted.
Rank 1: “Getting Back to Normal”
A desire to focus on “getting back to normal” by returning to work or school was mentioned by nearly all young adults and key supporters (97%). Participants interpreted this as directly conflicting with standard treatment advice to avoid stress. One explained, “What am I supposed to live off of? So that stressed me out too, because they tell you, ‘you can’t drive, you can’t work, you can’t do anything’” (L8). Another said, “I want to go back to college . . . schooling will always help because my brain is in a vulnerable state. If I don’t work it at least a little bit, I probably wouldn’t be as functional to society as I should be” (P3).
Rank 2: Insufficient Mental Health Care
The next most frequently mentioned concern was that the mental health care on offer was insufficient (78%). “I feel like the doctor is just giving medicine. . . . He would just look at your profile but he wouldn’t discuss [it] with you” (G16). Another said, “I would need someone that’s actually gonna listen to what I’m saying, not what a piece of paper says” (H8). Some wanted to talk to others with similar experiences outside of a clinical setting. “[I]t would help to meet other[s] . . . going through the same thing” (G16). Key supporters wanted more information, “because the unknown is worse than anything. . . . You don’t know what’s triggering it, what’s causing it, why it’s happening” (E.KS.12).
Rank 3: Concerns With Police Involvement
Many young adults and key supporters expressed distress over police involvement in the pathway to care (70%). Key supporters often called police in a desperate moment. “And then, police, he helped me take her because she don’t want to go... I want to take care of my daughter, but this is not good conditions” (G.KS.12). Another said, “I used the hospital for shelter because he was kind of violent and I had no place to send him, not for treatment” (F.KS.12). Another said, “We got scared. I called the police” (K.KS.16). Interactions with the police were often negative. “The cops . . . knocked my front door in” (A3). When the interactions went poorly, the involvement of police in the initial pathway to care (even if initially sought out) reduced people’s interest in using mental health care in the future.
Rank 4: Feeling Worse and Desire for Help With Relationship Repair
Two different challenges tied for the fourth-ranking concern that affected treatment engagement. More than two-thirds of participants said that mental health treatment made the young adult feel worse. One described her medication experiences in multiple ways over time: “it makes me feel like a robot;” “it makes me feel like I’m not happy as much;” and “I’m more down” (B1). Another said, “It was hard because of the medicine . . . it made me feel, like, ‘I don’t want to do anything’” (M3).
More than two-thirds also expressed concern about relationship tensions that arose as a result of the mental health breakdown and the subsequent pathway to care. As one young person said, “I was very hostile towards my mom and dad, so I didn’t want to talk to them. . .I didn’t know what was happening. I felt lost and confused” (K16). People wanted more help working through relationship tensions.
Rank 5: Living Independently, Paying for Care, Distrusting Diagnosis, and Concerns About Social Use of Substances
Most young adults (78%) were concerned with their future ability to live independently. They felt that their initial hospitalization threatened their ability to live independently. This was less concerning to key supporters (47%).
Paying for care was also a major concern that limited interest in treatment engagement for young adults (83%). Typically, they had no insurance coverage beyond a parent’s medical plan until age 26 years. After receiving the costly hospital bill (around $10,000 for one week, which one compared with the cost of a year at community college), they dreaded ongoing costs. “I would like to know how long I have to take the medication. . . . Because between the three medications? They’re not cheap. It’s $150 for a month since I don’t have insurance” (E3). Concerns about paying for care were mentioned less often by key supporters (42%).
Nearly two-thirds of the young people (67%) and more than one-half of their key supporters (58%) trusted neither the highly stigmatized diagnosis of a psychotic disorder nor the person who assigned that label after a brief interaction. As one young adult explained, “He said that it is psychosis, but I know what I am. [Psychosis is] a disconnection from reality. I’m not disconnected from reality!” (A6). Another said, “Don’t you need to see my brain first to judge it and say there’s something wrong with it?” (B1). Key supporters preferred alternative explanations, like romantic difficulties, pressure to perform at work or school, and substance (especially marijuana) use. For example, one key supporter said, “I don’t feel like it was schizophrenia. I feel like it was isolation” (D.KS.20). Perhaps as a result, some key supporters discouraged their loved ones from continuing treatment. As one said, “He doesn’t need it” (F.KS.12). Another said, “She is normal. She doesn’t have anything at all that is affecting her” (G.KS.16).
Two-thirds of the young adults also fretted about medical advice to abstain from alcohol and marijuana while taking psychiatric medications, which they saw as an impediment to their social lives. “Whenever they tell me to stop taking meds, I can start drinking,” one said (E0). Many key supporters also shared concerns about the social pressure to use substances faced by the young people (58%). Some participants resisted medications and services to continue substance use.
Rank 6: Feeling Disempowered by the Initial Hospitalization
The hospitalization process left many young adults (78%) feeling disempowered. “It was kind of heartbreaking,” one told us (Q0). Another argued, “You can’t treat people like they’re animals and expect them to get better” (O3). Another said, “They can do anything they want . . . if you say anything, you’re crazy” (M3). Many participants did not want to return to mental health care, which they associated with a hospital. “I would prefer just to go back to my sister’s and not have any strings attached to this” (I0). Several key supporters (37%) also felt disempowered by the hospitalization process. “They disconnect completely the patient from the family” (J.KS.8).
Rank 7: Transportation Challenges
Two-thirds of young people reported being told not to drive for at least 90 days by mental health professionals, and mass transit was not widely available, making transportation to medical appointments, school, and work a challenge. As one explained, “Hitting my sister or my niece up for a ride somewhere makes me want to pull my hair out because they get frustrated” (I12). Key supporters did not mention this as often (42%).
Discussion
Engaging racially and ethnically diverse young adults early in mental health care that is meaningful for them and addresses their needs is critical (
1,
2,
4,
10). This study sought to understand what might motivate young adults, primarily from racial-ethnic minority groups, to engage in a pathway through care instead of dropping out of care after an initial hospitalization for a psychotic disorder. Our findings suggest that pathways to and through care are messy and complex. Using data from interviews and field notes, we generated hypotheses about interactive factors that likely affect treatment decision making, at least for racial-ethnic minority, and likely for many, young adults. These hypotheses also suggest ways that early psychosis services can better engage young adults and key supporters in care.
Studies have shown that for most young adults who have been given a diagnosis of early psychosis (perhaps especially racial-ethnic minority young adults [
31]), initial mental health contact often occurs in a crisis situation involving police intervention and emergency treatment (
16,
32–
34). This can be traumatic and is likely to engender mistrust of services (
31,
35). We also found that many youths in our study followed toxic pathways to care that included police involvement, disempowering interactions with mental health care providers and settings during emergency hospitalizations, little support for relationship repair, and a severe diagnosis that was often poorly understood, while “care” seemed to be limited to pharmaceutical interventions that caused disruptive side effects. Expensive emergency hospitalizations offered very little material, moral, or therapeutic support.
After hospitalization, many felt worse. Being advised not to drive, to abstain completely from alcohol and drugs while taking medications, and to avoid returning to work and school to reduce stress was frustrating. The number-one priority for young adults and key supporters was getting back to normal. To do so, participants wanted help to address the financial challenges of mental health care, to meet increased transportation needs for appointments, to understand the role of substance use in mental health crises and recovery, and to repair relationships strained by issues that arose prior to and during their hospitalization and that remained unresolved.
Clearly, young adults with early psychosis need to have access to care prior to a crisis point (
16). For example, education campaigns can encourage people to seek help before they need emergency services (
16,
36). Many families also struggle with whether or not to seek out mental health services for what may be typical teen behavior (
16). Offering families good information that includes positive, nonconfusing messages can have a positive effect (
32,
37) and may reduce the uncertainty that young people and families feel (
38). Involving families is also critical for engagement (
3,
39,
40), a conclusion that is supported by our findings. This may be especially true for racial-ethnic minority groups, who often depend on family more (
41) and may be more sensitive to stigma-related shame (
31,
42). One-fourth of our key supporters were siblings, thus supporting studies showing that siblings may need more support (
43). Therapeutic offerings for families can work to repair relationships and help people manage a sense of loss and grief around the diagnosis (
16,
32).
Best practices for ideal engagement should include a strong therapeutic alliance, mutual acceptance, focus on the client’s life goals, and shared decision making (
7,
9). Shared decision making supports young adults’ autonomy and healthy development (
7,
44,
45). A good therapist can help people seek meaning at these early stages and affirm private understandings that help people move past their distrust of biomedical diagnoses, promote hope, and are essential for successful treatment (
46,
47).
Age-appropriate support is needed to address substance use or abuse, which is reported to be a robust predictor of treatment disengagement in the literature (
3,
39,
40,
48–
50). Experimentation with drugs and alcohol may be especially attractive to young adults with early psychosis because friends who are intoxicated are less judgmental of unusual behaviors (
32,
51). Specialized programs that address the social pressure to use drugs and alcohol and that offer advice on how to make good choices surrounding medications and substances used socially are needed.
Methodological limitations of our study included difficulty retaining participants and a small sample size. Yet, retention issues were expected, given that up to 40% of young adults with early psychosis drop out even from specialty programs (
3). Efforts to increase retention included avoiding the use of clinically oriented questions and language; contacting participants by using cell phones, social media accounts, and their key supporters (with permission); and offering financial incentives to follow-up. The sample size was small, but adequate for a purposive sample of interviews, which has been suggested to require a minimum of nine participants (
52).
This study is a first step in researching engagement longitudinally in a phase-based approach, as has been recommended (
8). Our study supports the suggestion that more options are needed (
10), and the presence of a youth-oriented, early intervention specialty program that uses best practices may help. More research is needed to test the relevance of the factors we identified for early psychosis programs and the racially and ethnically diverse young adults and key supporters who use them.
Acknowledgments
The authors thank Thomas M. Collins, Kanlaya Ditthakassem, M.N.S., R.N., Gail Barnes, M.S.W., Miguel Gutierrez, Megan Bartrum, M.A., Kelli Bassett, M.A., Brenna Raney, M.A., and Jordan Goldstein, M.A., for their assistance. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health or NIMH.