Young Adults’ Perspectives on Factors Related to Relapse After First-Episode Psychosis: Qualitative Focus Group Study

Relapse after first-episode psychosis (FEP) is a major clinical challenge for mental health services (1–3). Relapse of psychosis is defined as a recurrence in the severity of positive symptoms, followed by partial or full remission (4). A review of the literature (3) found that mean relapse rates among individuals receiving early intervention services (EIS) were 17% at 9 months, 38% at 24 months, and 54% at >10 years. Consequences of relapse, such as psychiatric hospitalizations, psychological trauma, and impacts on functioning and quality of life, are major concerns (1, 5–7).

Multiple studies (1, 3, 7, 8) have identified several risk factors for relapse after FEP, including medication nonadherence, substance use, caregiver criticism (i.e., expressed emotion), poor premorbid adjustment, poor insight, diagnosis of schizophrenia spectrum disorder, and affective psychosis. A small but growing body of qualitative literature (9–13) has begun to identify risk and protective factors for relapse and recovery drawn from research conducted with individuals diagnosed as having schizophrenia and FEP. In several of these studies (11–13), researchers have identified specific relapse risk factors—stopping medication; stress related to school, work, or social environment; life events (past and present); and isolation (e.g., living alone)—and protective factors—family support, medication adherence, and employment or studying.

However, limited research has used focus group methodology to examine the perspectives of young adults recovering from FEP on relapse. Such knowledge could contribute to a more comprehensive understanding of the relapse process, help improve assessment of relapse risk, and inform preventive strategies and interventions.

Complementing previous qualitative focus group research on relapse, from the perspectives of family members of individuals receiving EIS for psychosis (14), this study investigated factors that contribute to and prevent relapse by collecting and analyzing the experiences and perspectives of young adults receiving services for FEP.

Methods

This study was part of a larger qualitative project examining the experiences and perspectives of family members and patients with FEP on the subject of psychosis relapse and its prevention, the ultimate purpose of which was to inform development of a relapse risk assessment tool. The study first received central approval from the research ethics board of the Douglas Mental Health University Institute (Montreal) and then from the ethics review boards of three additional recruitment sites in Halifax (Nova Scotia), Chatham-Kent (Ontario), and London (Ontario). All participants provided informed consent before participating in the study. Participants were recruited and data were collected from June 2014 through January 2015. This study was based on a qualitative descriptive research methodology approach (15) and used focus groups as its main method for data collection. Focus groups are useful for stimulating discussion on complex matters, obtaining a range of perspectives, and supporting and empowering participants in exploring and articulating their perspectives through a group process. As such, focus groups are perceived as particularly relevant for giving voice to marginalized populations (16). Therefore, focus groups represented an appropriate data collection methodology to facilitate discussion in this exploratory study of relapse among a population receiving services for FEP. In this study, we focused on the data obtained from the focus groups conducted with patients.

Participants were recruited from four EIS for psychosis sites in Canada—Montreal, Halifax, Chatham-Kent, and London. Inclusion criteria were age ≥18 years, diagnosis of a nonaffective or affective psychotic disorder within the first 2–5 years of initial treatment for FEP, symptomatically stable and capable of participating in a 2-hour discussion (as determined by the primary treating psychiatrist), and experience of psychosis relapse or interest in contributing to a discussion on relapse. Participants without relapse experience were included because their concerns and opinions about the possibility of relapse, perceptions about what may contribute to or prevent relapse, and conceptualization and understanding of relapse were considered useful in providing insight for the development of relapse prevention materials for all patients receiving EIS for psychosis.

A semistructured interview guide was used to collect data on conceptualization of relapse, impact of relapse, recognition of relapse and its precursors, and factors that contribute to and prevent relapse. This study was mainly focused on data pertaining to the conceptualization of relapse, and relapse factors, stemming from the following questions: “What comes to mind when you hear the word relapse? How would you define it? What helps to prevent a relapse from happening? What is not so helpful in preventing relapse?” Following these open-ended questions, the guide also included questions on the role of substance use, family, and medication (e.g., “What role does medication play in contributing to and/or preventing relapse?”). Topics were based on documented risk factors for relapse after FEP. Additionally, participants were asked to reflect on the role technology plays in relapse. Discussions ranged from 60 to 90 minutes and were facilitated by one author (S.L.), with logistics facilitated by a research assistant. In total, we conducted four focus groups (one focus group per recruitment site), which were audio recorded and transcribed verbatim. Research (17) suggests that data obtained from three to six focus groups, with a specific population group and a semistructured interview guide, are likely to achieve saturation of themes. We analyzed the four focus groups for any participant mention of a contributory or preventive factor. We set our threshold for reporting a factor at four separate mentions, meaning that if at least four individuals from the total sample of 25 identified a factor, we would report it, because it might warrant future exploration with larger sample sizes.

Data were analyzed with a descriptive content analysis approach (18). Specifically, all four transcripts were first reviewed and coded inductively and independently by three coders (S.L., A.C., and a research assistant). Each coder developed a coding framework (with preliminary codes and categories) reflective of the data set, which was then discussed and finalized through consensus meetings. The final coding framework was systematically applied to the entire data set by one author (A.C.). This coding framework was broad, in that it addressed several topics that were raised in response to the interview guide (e.g., conceptualization of relapse, impact of relapse, how to recognize relapse, and contributory and preventive relapse factors). The data coded in relation to specific categories (e.g., contributory factors) were then extracted into an Excel sheet by one author (A.C.) to identify additional codes and subcodes and then reviewed in their entirety by another author (S.L.) (including a review of the original transcripts in relation to applied categories, codes, and subcodes). At this stage, minor revisions were made to the final coding of participant quotations through discussions between the two coders, mainly about naming and collapsing categories. Category response type frequencies were calculated to illustrate factors commonly ascribed as contributing to or preventing relapse.

Results

Twenty-six participants attended one of the four focus groups (Montreal, N=7; Halifax, N=4; Chatham-Kent, N=8; London, N=7). One participant was excluded from the analyses after withdrawal of consent after >1 year of study participation. The sociodemographic characteristics of the 25 participants, including their experience with relapse, are presented in Table 1. Of the 25 study participants, 10 acknowledged that they had experienced a psychosis relapse, and two indicated that they thought they had a relapse but were not sure. Four stated they had not experienced relapse, and nine did not respond or did not provide sufficient information on this topic. Nonetheless, all participants could convey their perspectives on what might contribute to or prevent relapse on the basis of past experiences of relapse or on other sources of information (e.g., factors contributing to their FEP, factors contributing to their recovery, information received from service providers, and observations from their social network—such as the relapse and hospitalization of a family member).

Conceptualization of Relapse

During recruitment, participants were provided with a brief definition of psychosis relapse in the patient consent form: “By relapse we mean a return in symptoms that can result in a decline in daily functioning.” During the focus group discussion, however, the facilitators refrained from repeating this definition or elaborating on it. Instead, as an icebreaker activity, the facilitators asked participants to share what came to their minds when they heard the word relapse, and the discussion ensued from there. When asked to define relapse, a dominant theme across the focus groups was uncertainty about the meaning of relapse; some participants returned the question to the facilitator (IDs of anonymized participants are given at the end of quotations): “So what is relapse?” (P7); “I’ve had instances where it’ll come back, but it might only be for a few hours. Is that considered a [relapse]? I’m still a little iffy on the definition” (P2); and “I still haven’t really had a clear definition” (P5). The uncertainty was often related to which symptoms and behaviors constituted relapse, as well as to their severity and duration: “Like I heard loud voices and became extremely paranoid about them. But then I slept it off, and it went away. So I’m not sure if that’s a relapse” (P1). In general, however, most participants defined relapse in relation to a change in symptoms (i.e., psychosis symptoms, change in mood, and increase in anxiety); thoughts and behavior (e.g., “behaving really weird”); functioning (“not functioning normally anymore”); or physical state (“extra energy that’s causing you not to sleep”). They also conceptualized relapse in relation to hospitalization (“going back to the hospital”) and reuse of substances (e.g., marijuana).

Factors Related to Relapse or Relapse Prevention

Participants identified 12 unique factors related to relapse or relapse prevention, four of which were discussed as both contributory and preventive. The next section elaborates on each factor; Tables 2 and 3 provide further details, including illustrative quotations and percentages of the sample identifying each factor.

Factors That Were Both Contributory to and Preventive of Relapse

Social environment (supportive or unsupportive).

Nineteen participants identified an unsupportive social environment, specifically pertaining to family, peers, and health care providers, as a factor that could contribute to relapse. Participants perceived unsupportive behaviors when family and peers were judgmental; told them what to say, do, and think; were impatient with the recovery process; constantly inquired about their well-being; showed a lack of understanding toward the illness and its symptoms; or expressed their own stress. Participants also reported that intimate relationships and related breakups contributed to relapse. In interactions with health care providers, examples of behaviors that participants perceived as unsupportive typically revolved around information and explanations provided (or not provided) regarding symptoms and treatment: “The least helpful thing for me would be the nurses and doctors insisting that everything bad that has happened to me is delusion. Just like it’s over and over, . . . and it’s really hurtful” (P22).

Concurrently, all participants identified supportive behaviors from family members, peers, and health care providers as helpful in preventing relapse. Supportive behaviors included being trustworthy, listening, providing validation, having patience, expressing genuine concern, and not being judgmental. In terms of relationships with health care providers, participants explained that having a positive relationship with their case manager, nurse, or psychiatrist was important for preventing relapse. In particular, they described the following as being helpful: being understood and validated; being given clear information regarding their illness, symptoms, and treatments; and having timely access to appointments.

Five participants also mentioned societal education when asked about factors that could prevent relapse: “The greatest barrier to a successful and happy and fulfilling life with somebody with a mental illness is the stigma that’s attached to [mental illness]” (P9). Participants recommended that more education in formal school settings is needed to help the public empathize with patients and to reduce mental health–related stigma.

Technology use.

Fifteen participants described technology-related activities as potentially contributing to relapse (e.g., playing video games with others at a distance, listening to music, and using social media). They explained that technology can trigger concerns regarding being monitored (e.g., by cameras), hacked, tracked, and manipulated. One participant elaborated that the use of the Internet and technology could worsen symptoms:

Conversely, 22 participants mentioned that the Internet and social media activities could help prevent relapse (e.g., to consider other explanations for their thoughts and observations, learn about psychosis, overcome fear of treatment, or seek additional or alternative solutions for recovery). They also described the role of technology in staying connected with family and friends on social networks. Social media platforms, such as Facebook, YouTube, and Twitter, were considered important sources of peer support and assistance with symptom monitoring:

Participants also described technology as facilitating help-seeking efforts when experiencing symptoms of social withdrawal:

Medication (taking or not taking).

Fifteen participants identified not taking medication or going off medication as potentially contributing to relapse. In addition, they described how the dosage or type of medication, prescription change, or adverse effects of medication could contribute to relapse.

Nineteen participants also believed that taking medication was helpful in preventing relapse. Other participants expressed ambivalence toward a medication’s effectiveness in prevention, questioned the need to take medication long term, highlighted the importance of a holistic approach, and expressed concerns regarding adverse effects: “Medication wasn’t the total answer, because I had a lot of symptoms of fatigue, and there were symptoms from the medication that were really unpleasant” (P3); “I’m not always for it either, because I find there’s other things, like keeping busy and having an active life and doing things, that are good and positive, that help you” (P5).

Lifestyle behaviors (sleeping or resting, exercising, eating).

Ten participants described lifestyle-related behaviors, such as those leading to a lack of sleep and rest, as potentially contributing to relapse. They explained that lack of sleep could affect mood and ability to think the next day, leading to stress and symptoms:

At the same time, 11 participants also described the importance of lifestyle behaviors in preventing relapse, explaining that sleeping, exercising, and eating well have a positive effect on coping with stress, building confidence, and regulating mood: “For me exercise is huge. The endorphins from that really make me feel so much better and less depressed and things like that. And [not] eating properly, that really affects my mood, which could be causing symptoms” (P22).

Participants also highlighted the importance of keeping a routine and having a balance between work and other daily activities: “It’s not too much work, it’s also sort of play and enjoying myself and enjoying my life. So the balance is super important, right, so that you feel nourished and not depleted” (P3).

Contributing Factors

Substance use.

Nineteen participants identified substance use as a factor that may contribute to relapse, often referring to cannabis (“weed” or “smoking pot”). Other, less frequently mentioned substances included alcohol, magic mushrooms, MDMA (3,4-methyl​enedioxy​methamphetamine [ecstasy]), cocaine, and bath salts (synthetic cathinone). Despite acknowledging substance use as a relapse risk factor, participants also expressed challenges and ambivalence toward abstaining from such use:

Participants further stated that individuals with a history of psychosis were more likely to experience relapse when consuming substances: “It doesn’t actually cause it; it just brings it out of you if you’re prone to it” (P11).

Work, career, school stress.

Ten participants described performance-related stress at work (e.g., meeting deadlines), unstable career status, being overworked, and school stress as risk factors for relapse, and explained their vulnerability toward these factors: “Students experience stress during big exams or midterms and things like that. If you had a psychosis, it’s more severe. You don’t deal with it as well as others” (P15).

Significant life events (positive, negative, and unresolved trauma).

Eight participants described life events as risk factors for relapse. These factors pertained to positive life events, such as having a baby; negative events, such as death of a family member; and unresolved traumas from the past (e.g., being bullied or abused): “If you suffered abuse during your childhood or otherwise . . . and there wasn’t any kind of psychological help, this [relapse] can happen again. And I don’t think medication helps in this situation” (P6).

Symptoms of depression or mania.

Seven participants described depressive or manic symptoms as potentially contributing to relapse. In relation to depression, one patient explained:

Participants also discussed precursors to depression, such as relationship issues with peers (e.g., a breakup). Those with manic experiences explained that such symptoms hindered their ability to feel a need for treatment, leading to a psychosis relapse.

Generalized worry.

Six participants described anxiety, particularly related to daily activities and responsibilities and generalized anxiety, as a potential contributor to relapse: “Always just worrying about, like, I don’t know. Sometimes it’s like every day is doomsday kind of thing. Sky’s always falling, you know?” (P15).

Financial stress.

Four participants identified financial stress as a potential contributor to relapse. These participants used expressions, such as “being broke” (P8), or clarified that their financial concerns could cause stress leading to relapse.

Preventive Factors

Coping strategies.

Fifteen participants described a range of coping strategies for preventing relapse, including learning about the illness (i.e., what psychosis is, how it is treated, and its early warning signs), making sense of the illness within the broader context of one’s life, taking responsibility, caring for oneself before focusing on others, building confidence, and staying positive. Learning both to recognize psychosis symptoms and personal triggers and to seek help when noticing early warning signs were particularly useful strategies: “I think it’s very important that we detect our symptoms and some of the signs that can lead to relapse and prevent them” (P9). Participants explained that knowledge about the illness was not only important for themselves but equally important for caregivers and could be acquired through the Internet, by attending group education sessions, and through interactions with service providers: “I think being educated is the most important one because, again, it just goes back to being familiar with what’s going on so they can be more supportive” (P14).

Engaging in meaningful activities (productive, leisure, physical, or spiritual).

Ten participants discussed engagement in meaningful activities as an important relapse prevention factor. These included productive activities (e.g., working, looking for a job, or going to school) and spiritual, physical, creative, and social activities (e.g., praying, painting, spending time with pets, playing or listening to music, spending time with family and friends, watching a baseball game, being active, and meditating). Participants also emphasized the importance of engaging in activities outside work and school and investing time in enjoyable activities. They explained that meaningful activities could help prevent symptoms by building confidence and by keeping them occupied: “When I stay busy it helps me, it just keeps me going—I don’t hear voices, I don’t hear anything, and I’m just neutral” (P4). Participants also frequently highlighted creative outlets as ways to reduce stress and to channel negative emotions in a healthy way: “I had a lot of intense emotion, and instead of getting myself into all this mess, I painted it. I paint it all out, I let it out in the paintings, and I felt liberated, I felt free” (P3); “[I am] just kind of staying creative to eliminate the negative stress” (P19).

Discussion

This qualitative study contributes to a nascent body of research on factors related to relapse of psychosis from the perspectives of young adults receiving treatment for FEP. Such research is needed in different clinical settings and across geographical, cultural, and sociodemographic contexts to ensure that the development of relapse risk assessment tools and prevention interventions can best meet the needs of a diverse population. The conceptualization of relapse by the participants in this study was broader than how it is typically defined or measured in the relapse literature (i.e., recurrence of positive symptoms or hospitalization). Specifically, their definitions extended beyond symptoms of psychosis and included changes in mood, anxiety, and behavior (including reuse of substances). The participants appeared to have difficulty in distinguishing early warning signs from an episode of relapse. This finding was noteworthy, because participants’ uncertainty regarding the duration, severity, and criteria for symptoms and behaviors that would constitute relapse highlights a need for focused education and clarity on what relapse is, how to recognize it, and how to identify its early warning signs. The extent to which conceptualization of relapse may differ for patients with and without a history of relapse is unclear. This difference would be a relevant area to examine in future research (i.e., through individual interviews), because insights from such research could help inform relapse prevention education.

We identified 12 factors related to relapse, four of which were discussed as both contributory and preventive factors. Our findings agreed with or diverged from results of previous research in several ways. The finding that medication nonadherence, substance use, and type of caregiver communication contribute to psychosis relapse is in agreement with findings from a systematic review of quantitative research in a similar population (3) and with findings from qualitative studies of older adults recovering from schizophrenia (11, 13). Other studies (12, 13) have also identified medication as both a contributory and preventive factor in relapse and have reported patient ambivalence about the effectiveness of medication (12). Going beyond previous work, the results of our study illustrate that patients’ ambivalence about medication stems from the perceived impacts, adverse effects, dosages, types, and changes in regimen, which may represent important elements to consider in the context of relapse risk assessment. Similarly, participants also reported ambivalence toward substance use; thus, besides actual change in such use (which may reflect more imminent risk), attitudes toward medication and substance use warrant further consideration of their predictive potential in assessing relapse risk. Furthermore, our results drew attention to what concerns patients have about the behaviors of their caregivers (e.g., constant check-ins and overexpression of caregiver’s own stress) and to the role of relationships with other stakeholders in relapse (e.g., relationships with siblings, aunts and uncles, peers, and health care providers).

In contrast with previous qualitative studies that have identified work and studying as protective factors and coping strategies (10, 13, 19), a dominant theme in this study was the contributory roles of work and school in stress and relapse. This finding may have resulted from the fact that the participants were younger, in earlier phases of recovery, and had less experience in coping with these roles. More research is needed to better understand the role of factors contributing to relapse across different phases of recovery in psychosis.

Moreover, four participants in this study identified financial stress as a factor contributing to relapse; although these participants represented a small subset of the sample, this finding suggests that persons with different sociodemographic backgrounds may perceive relapse factors differently. A person’s sociodemographic situation may lead to a particular factor having a more critical role in the experience of stress and relapse. This is an important consideration for the development of relapse risk assessment tools, as well as for relapse education and prevention.

Implications for Relapse Risk Assessment

The findings of our study may inform development of tools for relapse risk assessment among individuals receiving treatment for FEP. Relapse risk is typically assessed through interviews and assessments completed by clinicians. Because this approach provides only a snapshot of an individual’s status, and many such assessments have been designed for research but not clinical purposes (5, 20), easily completable tools for patients to identify their risk for relapse, and that allow for both active and passive monitoring, are needed (20–22). Recent efforts have been made to address some of these challenges (e.g., development of the Relapse Risk Assessment for Schizophrenia Patients [RASP], a self-report tool consisting of six items that screen for increased anxiety and social isolation [21], and CrossCheck, an online tool that combines ecological momentary assessment and behavioral sensing of physical activity, speech, location, and technology use). CrossCheck (20) includes a 10-item self-report covering symptoms of psychosis, general mental health, and functioning.

Our findings supported items included in RASP and also highlighted the importance of attending to additional factors, such as perceptions of relationships with families, peers, treatment providers, employers, and teachers; attitudes toward and use of illegal substances and prescribed medication; use of technology; changes in mood; significant life events (positive or negative); and engagement in and experiential perceptions of productive and restorative activities. The risk assessment approach of the CrossCheck study was more strongly aligned with the range of factors we identified in this study, including the role of technology. Conversely, our study also highlighted additional factors that warrant consideration (e.g., lifestyle behaviors, such as changes in eating patterns). Importantly, there has been little to no description of patient involvement in the design and development of the aforementioned tools. Our study provided patients with a voice on important considerations pertaining to relapse factors that can advance this previous work. For example, attending to diversity in terms of sociodemographic characteristics and circumstances (e.g., financial situation, housing circumstances, language barriers, and living in a marginalized community) is important in relapse risk assessments.

Implications for Relapse Prevention

On the basis of these findings, we offer the following eight recommendations for the planning and implementation of relapse prevention interventions. First, a need exists for targeted education to address patient uncertainties regarding the duration, severity, and criteria of symptoms and behaviors that constitute relapse; contributory and preventive factors; and early warning signs. Although in theory these aspects are addressed by existing interventions, the provision of these interventions may need to be reexamined to determine whether they sufficiently and effectively address patient education needs. Second, routine discussion is needed on the role of technology in relapse, including ensuring that patients are aware of credible websites to access information on psychosis, recovery, and the lived experience of others. Research (23) has shown that patients being treated for FEP have difficulties navigating the Internet to find mental health information and support. Service providers may need additional training to help patients find appropriate resources. Third, we recommend ensuring that patients have access to focused interventions that help them cope with work- and school-related stress, as well as to support and advocacy for workplace and educational accommodations. Fourth, occupational therapy interventions should be provided to facilitate tailored access to productive and restorative activities (24). Fifth, technology should be used and innovations developed to increase patient access to interventions and resources that support healthy lifestyle behaviors. Toward this aim, examples for such technology, such as HoryzonsCa, show some promise (25). Sixth, psychoeducation should be provided to patients and their families and should also be considered for siblings and friends, including information on what constitutes psychosis relapse and how to identify and prevent it. Moreover, we suggest that caregiver education include strategies on how to best support patients in engaging in healthy lifestyle behaviors. Seventh, the extent to which diversity is addressed by (or is absent from) existing relapse prevention interventions and materials should be examined and tailored to address sociodemographic factors of risk. Finally, research should be conducted to examine the short- and long-term impacts of relapse prevention interventions.

Strengths and Limitations

This study had several strengths and limitations. First, it brought attention to the perspectives of patients receiving EIS in terms of factors contributing to and preventing relapse. Second, we collected data from participants in four EIS programs across three Canadian provinces. Third, three research team members were involved in the initial coding of transcripts and in the development of the coding framework to reduce coder bias. Fourth, we documented the frequency of each factor to minimize the risk for bias in emphasizing any one category.

In terms of limitations, because recruitment of participants was facilitated by clinicians who had regular contact with their patients, the perspectives of participants who were disengaged from care or those with infrequent follow-up were not represented. Additionally, because of the sensitivity of the topic, the focus group context may have prevented some participants from sharing their perspectives; for example, we were unable to obtain information about a history of relapse from all participants in the context of the focus group discussions. Therefore, we were unable to map differences in perspectives in relation to relapse history. Future studies should also consider conducting individual interviews. Moreover, collecting further data from a more diverse range of settings and participants (e.g., representing a diverse demographic background and social circumstances) may reveal additional insights into relapse factors. As such, our findings may not generalize to individuals in different recovery phases, those not receiving care in EIS (or an equivalent), or those representing marginalized communities or circumstances (e.g., young people experiencing concurrent homelessness and FEP). In addition, we did not seek participant feedback on the findings. Finally, the frequency numbers did not capture the perspectives of participants who contributed through nonverbal communication (e.g., head nodding to indicate agreement), because the discussions were not video recorded; it is therefore likely that the number of participants who agreed to the importance of each factor was underestimated.

Conclusions

Our examination of patient perspectives on factors related to psychosis relapse provides important insights relevant for risk assessment of relapse. This study’s results can therefore be used to better inform clinical and research interventions aimed at reducing such risk. Most factors mentioned by the participants were modifiable and have implications for service delivery programming.