Cornerstone program for transition-age youth with serious mental illness: study protocol for a randomized controlled trial

Unmet mental health needs among transition-age youth and young adults (defined as 17- to 25-year-olds) with serious mental health conditions, who are predominately low-income, minority, and system-involved (i.e., public welfare, foster care) are well known; however, they have not been satisfactorily addressed. The onset of mental health conditions often occur in individuals during adolescence and young adulthood [2]. Epidemiological reports reveal that rates of mental health disorders among 18- to 24-year-olds are high when compared to older age cohorts [3] and studies of youth who have been involved with public systems of care reveal even more elevated rates of mental health conditions [4, 5]. Beyond prevalence, research has documented the extreme difficulties these young people face, such as poor education, underemployment, increased risk of harm to self and/or others, and substance abuse problems [6]. There are at least two major reasons this is of concern: (1) many transition-age youth and young adults stop receiving professional help during this critical developmental transition [7], and (2) the mental health system has few evidence-based psychosocial interventions designed to address their needs [8].

Studies have shown that the transition to adulthood leads to discontinuation of, or at best, sporadic use of mental health services [7, 9]. We know that consistent treatment can decrease symptoms and increase functioning, whereas inconsistent utilization is likely to be less effective [10]. Thus, it is not surprising that studies show that early discontinuation is associated with an increased likelihood of recurrence of illness [11]. One study recently interviewed 60 young adults who had used Medicaid-funded mental health services, had a formal psychiatric diagnosis during childhood, and reported still struggling with emotional problems, to learn about their experiences with mental health services over the transition to adulthood. Results revealed that 42 % of the young adults were no longer using any mental health services and another 37 % reported gaps in service use during the transition [9]. Another study found that approximately 60 % of a sample of youth aging out of foster care discontinued services 1 month after exiting the system [7]. Similarly, longitudinal studies of child welfare-involved youth report decreasing use as young adults; one stating that while 47.3 % participated in mental health services as adolescents, only 14.3 % continued as young adults [12, 13]. Disengagement in services over the transition is a clearly documented problem, which we are still struggling to address.

Research has identified barriers to mental health service use among transition-age youth, for example, lapses in insurance, mistrust, and stigma, among others [7]. For example, the belief that there is stigma associated with participating in treatment remains common among transition-age youth with mental health conditions, [14, 15] and it may play a role in the reluctance to continue treatment [16]. Also, studies have found that a lack of understanding and knowledge, sometimes referred to as ‘mental health literacy,’ of mental health and treatment can be a significant barrier to service use [17]. Further, these young people report that pessimistic messages from providers, irrelevant services, and mistrust are barriers to treatment continuation [18‐20]. While the Affordable Care Act (ACA) can expand coverage for young adults, if implemented, without attention to perceptual obstacles that may be part of young adults’ beliefs, along with negative affective associations, continuous care is unlikely to be achieved. Such barriers must be addressed if we are to minimize dropout, increase investment, and improve long-term health and functioning outcomes of transition-age youth with SMHC.

In sum, the present article provides a description of the background, theoretical framework, protocols and the study aims for this intervention development trial, which will refine and examine the feasibility, acceptability, and preliminary impact of a theoretically guided intervention focused on improving the mental health, service use, and young adult life outcomes of low-income individuals with SMHC. The project will develop revised manuals (i.e., treatment, training, and implementation) and protocols with enhanced ‘real world’ validity in the mental health clinical context, and test its preliminary efficacy among transition-age youth with SMHC.

Clark’s Transition to Independence Process Model (TIP) is one of the most promising practices for transition-age youth and young adults with SMHC [21]. Outcome studies of the TIP have documented that it is associated with improved education and employment [22, 23]. TIP is delivered in different service platforms with a ‘transition facilitator’ (TF) at its core. TFs are case managers who use a set of guidelines to provide assistance to youth and young adults in transition living with SMHC. TFs are taught and coached in the following guidelines: (1) strengths and needs assessment, (2) future planning, (3) rationales, (4) in vivo teaching, (5) problem solving, (6) prevention planning, and (7) mediation. While promising, TIP is difficult to implement in ‘real world’ settings, as it requires communities to hire purveyors to provide extensive training and certification, which is costly. Further, Weisz and colleagues [24] have argued that effectiveness studies, including those of TIP, often do not involve clinic-referred youth and often are not delivered by clinicians employed in urban, stressed, community-based settings [25]. The intervention that will be tested in this trial draws on some of the strengths of TIP, among other models, to test a components-based intervention within an urban ‘real world’ setting.

Peer support models are gaining support nationally. For example, in New York State and throughout the country peer support models are being explored for marginalized populations of youth in transition [26, 27]. Peer support services, in most cases are services provided by individuals who also live with a mental illness and therefore, possess “lived experience” [28]. Lived experience can make peers particularly credible, trustworthy, and influential in communicating to youth and young adults who often are skeptical of professionals [29]. Ideas for the recovery role model position emerged out of previous research on understanding how peers can make a significant impact within the mental health system specifically through the role of mentorship [30]. Whereas recovery role models (RRMs) align with the peer support movement by taking advantage of the benefits of shared lived experience, they are distinct in that the relationship with the transition-age youth is intentionally hierarchical (e.g., mentor and protégé), not mutual. Also, they move beyond providing emotional support and companionship, as RRMs are recruited, screened, trained, and supervised as mentors who are older, wiser, and have something (knowledge, skills, and experience) to offer a younger, less-experienced protégé [31]. The role models employed in the Cornerstone trial mentor transition-age youth on the following: (1) consistent use of mental health services, (2) recovery (modeling hope for the possibility of a full life with a serious mental health condition), (3) managing disclosure decisions of mental health challenges in a stigma-laden society, and (4) life outcomes (i.e., education, housing, employment), among other areas of young adulthood unique to each participant. Recovery role model mentors serve as trusted advisors who possess specific knowledge, experiences, and skills associated with living with a SMHC. They also, by definition, are individuals who have some distance from their last episode of significant symptoms, whereas the participants who enroll in Cornerstone are in the midst of a significant episode of illness.

Mentoring can be particularly relevant for transition age youth and young adults with serious mental health conditions. Research has shown that the ‘transition years’ are often the developmental period when young people experience their first onset of a mental disorder, or serious mental health challenges [2]. Given that these years are a time of critical identity development, mentors may be particularly relevant in assisting youth in making sense of their life with a mental health condition [32]. Previous research suggests that transition-age youth report recovery role models are engaging and helpful in a myriad of ways that feel important to youth in the process of getting better after an episode of illness [33]. Beyond representing a conceptual innovation the recovery role model fits with our transforming care system, which emphasizes outcomes-based reimbursement for non-professionals. Recovery role models can be a cost-effective strategy in the current health care landscape. The position also provides employment opportunities for adults in recovery from a mental health condition. Finally, the position builds upon current efforts throughout the nation to certify peer support specialists in providing services to individuals with serious mental health conditions [34].

Group work is another approach that has been shown to be promising for transition-age youth who commonly have had experiences that have left them lacking in relational trust [35]. A group approach where young adults in transition living with SMHC are allowed to process experiences, alongside a recovery role model who comes from a similar community who has been able to engage professionals, as needed, and move forward in their lives, may prove effective. Previous research has found that the participants find the group process to be particularly helpful and that it decreases feelings of isolation [33]. The groups in Cornerstone focus on areas of development that key stakeholders identified as critical and not well addressed, such as substance use prevention, healthy relationships, safety planning, employment skills, education planning, and safe housing, among others (See Table 1 for Cornerstone Group Manual Table of Contents). Group content will be flexible based on participant needs as well, but the manual provides a set of curriculum for providers to use.

Critical Time Intervention (CTI) is an evidence-based practice that has been shown to effectively prepare adults with serious mental illness for critical transitions through the use of three distinct phases of practice [36]. Herman and colleagues have documented that time-limited case management focused on ‘linking’ individuals to needed services can lead to effects that are enduring for adults with mental illness through careful transition planning and the development of ongoing supports in the community [37]. In this randomized controlled trial, Cornerstone utilizes some of the guiding principles of the CTI approach by identifying key developmental and service transitions participants are experiencing while in the program. Cornerstone also builds on CTI through the development of “phase-based practice” for transition-age youth with SMHC; each youth in the trial can receive services for up to 9 months, yet services are offered at a high level of intensity in the first phase with decreasing intensity and increasing linkage by the third phase. We are piloting protocols to examine participants’ “transition readiness” and develop “transition planning”. Cornerstone staff use the assessment tool(s) to identify which areas should be the focus of services and linkage efforts. We will follow the CTI principle of reducing the role of staff in delivering services, gradually preparing the participants with linkages in the community, and working to secure a support team for young adults once they are discharged from Cornerstone [38]. We hypothesize that this will reduce abrupt transitions to adulthood, while improving overall young adult outcomes.

To date, interventions for transition-age youth and young adults with mental health conditions have largely been focused on principles of practice. This randomized controlled trial moves the field towards components-based interventions, which can be examined for feasibility, acceptability, impact, and fidelity. The four major components are boundary-spanning case managers, recovery role models, knowledge and skills-based groups, and in vivo community practice. We describe components 1 and 2, which have components 3 and 4 embedded within their description, as they are delivered by the intervention clinical staff.

Cornerstone has increased relevance as the mental health system is transforming with the implementation of the Affordable Care Act (ACA) and New York State’s Medicaid Redesign. In particular, changes will transform payment structures (connecting them to outcome performance), expand the workforce beyond professionals, and expand coverage for transition-age youth and young adults with serious mental health conditions, particularly Medicaid coverage. Further, the ACA is moving the nation into a healthcare landscape that will integrate care for mental health, substance abuse and physical health issues. Our proposed study is closely aligned with these changes, and a project aim is to research implementation questions related to the sustainability of Cornerstone within the urban outpatient mental health clinic context.

In our small trial, we apply an innovative framework of young adult mental health and mental health service use that examines multilevel factors that influence mental health service use among transition-age youth and young adults with serious mental health conditions [7] (See Fig. 1). This mid-level conceptual model illustrates the empirically based underlying mechanisms of change for study outcomes. The conceptual model, in part, focuses on cognitive processes that are part of an individual’s decision to use (or not use) professional mental health services. The proposed study will examine if Cornerstone impacts three key mediating factors that emerged as salient in previous research on the study population: (1) behavioral beliefs, defined as the perceptions of the advantages/disadvantages of using mental health services, i.e., whether services are effective, whether providers are trustworthy; (2) image considerations, i.e., stigma, and (3) efficacy, i.e., one’s perceived ability to access and engage in mental health services. The conceptual model also proposes that Cornerstone, in particular, the providers of services can influence theses salient cognitive mediating factors discussed above, along with directly influencing service engagement and young adult mental health and life outcomes. These relationships are depicted in Fig. 1 above.

Aim 2 will be accomplished with a two-group (random assignment, treatment versus control) by four-assessment (baseline, 3-month, 6-month, 9-month post) design. Participants will be involved for approximately 9 months. We will use an intention-to-treat design, thus, continuing to collect data on individuals who decide to drop out of the program before clinically indicated.

The partner agency provides clinical treatment to transition-age youth and young adults with serious mental health conditions (SMHC). These services provide a safety net for clients, facilitate access for clients to a wide range of services in the community, coordinate care to ensure that clients receive psychiatric treatment, and other concrete services. Youth in the TAU condition will receive these services as delivered at the agency. Participants in this condition will transition from services following agency protocols.

The names and identification numbers (IDs) of participants who consent to become involved will be entered into a tracking form. IDs will be substituted for names on all questionnaires and study forms. Confidentiality of data will be assured by maintaining locked file cabinets and protected electronic files that require a password. The PI and project coordinators will be responsible for maintaining IDs and data files. We have created a tracking system using established strategies [43]. The investigative team has considerable experience with recruitment and retention of transition-age youth with serious mental health conditions in a variety of settings, and with individuals in poverty experiencing multiple stressors. Our strategies have resulted in high retention rates (90 %) in our previous studies [4, 44]. We will use these experiences to implement the tracking system and to help minimize logistical barriers for participation. Furthermore, some of the young people targeted for enrollment will already be receiving care at the clinic. Due to the close relationship between the staff and the investigators, we anticipate few participants will be lost to follow-up. Keeping a list of participant names, addresses, phone numbers, and contacts will enable us to make phone calls and send appointment reminders in an effort to maintain engagement. We will also ask each participant on a regular basis to update their phone numbers and addresses so that we have alternative ways of reaching them. This tracking system will be locked in a separate file cabinet from all study data that will only be identified by a study ID number. Based on our experiences, we do not anticipate many participants will be lost due to relocation. We are recruiting 60 transition-age youth (TAY) to achieve a final sample size of 55.

Addressing within-group randomization and potential for contamination: in the proposed study the experimental/control assignment will be done within site and therefore issues of potential contamination must be addressed. Our research has revealed that we must consider contamination on both the participant and the provider level(s). It is possible that the participants may discuss their involvement in Cornerstone with another young person who is in the control condition. However, we expect this to be minimal as these young people do not typically have involvement with each other either within or outside the outpatient clinic site. Contamination at the provider level is of some concern as well. It is possible that the Cornerstone staff might share materials from Cornerstone manuals and trainings; however, the facilitation skills require practice and supervision to incorporate the services into overall care. Thus, contamination at this level will be minimal. We will take precautions by asking providers not to share materials and information until the study is over and explaining why this is important. We will also remind providers during our regular project meetings of this sensitive issue. We will also collect data to assess the nature and degree of contamination by asking providers if they have discussed Cornerstone content with colleagues, and if so have they noticed service changes as a result of discussions.

Most of the measures proposed in the study have been piloted with the target population and have been found to have good psychometric properties. We are also including a few measures in the battery in order to examine their psychometric properties, as a cost-effective strategy in preparation for their use in future studies. Table 2 presents a summary of the major study measures and their psychometric characteristics. These measures include factors we hypothesize are going to make the greatest impact.

Data analysis for aims 1 and 3: aims 1 and 3 involve qualitative data analysis, specifically grounded theory coding techniques [46, 47]. The qualitative data derives from in-depth interviews and group meetings. With participant permission, qualitative research activities will be recorded on digital audio recorders and in field notes. Analysis will proceed in two steps. Interviews will produce large volumes of text to be content analyzed. We will use a data-reduction process in which emergent themes are identified and coded to yield a set of core themes. Step 1 involves writing analytic summaries, concise reviews of key findings in each interview and field note. Analytic summaries will allow us to discuss key findings during project meetings. Step 2 involves systematic coding using a well-defined thematic codebook. Codebooks for qualitative analysis are theoretically informed manuals of codes and subcodes, defined by specific definitional criteria that allow for systematic textual coding. We will derive an analytic codebook based on the analytic summaries produced during step 1 and emergent themes. Trained coders and project staff will conduct all coding. All data will be analyzed using Atlas.ti, which will allow investigators to develop thematic units, or frequently occurring sets of explanatory statements [47]. In addition, the data will be explored for negative incidents and divergent themes vis-à-vis negative incident analysis [48]. With coded data in Atlas.ti, we will examine the range of variation of individual codes across our sample to obtain information about aggregate tendencies. This “vertical” analysis decontextualizes segments of data by removing them from the larger transcript, thereby permitting the examination of code-specific responses. Then, we will examine individual cases by analyzing how instances of particular factors are related to the larger context of meaning, experience, and behavior. This “horizontal” analysis allows us to examine how contextual factors may vary for transition-age youth with serious mental health conditions. We also will apply constant comparison coding methods throughout the project [49, 50]. Results will assist in modifying protocols and procedures at the local level, while also developing a Cornerstone Implementation Manual, and improved Cornerstone Intervention Manuals and Protocols.

Data analysis for aim 2: due to limited sample size we will proceed cautiously to conduct basic omnibus analyses to examine whether Cornerstone leads to improved outcomes relative to TAU, as this study is intended for intervention feasibility and acceptability testing, and therefore we will conduct only exploratory statistical testing of study outcomes. Our primary goal in this proposal is to gather preliminary data to inform a future large-scale, fully powered, randomized control trial of Cornerstone [51]. For each study subject we will calculate the difference, relative to baseline, at the end of the study for each outcome specified (see Table 2 above). A simple t test will be used to assess whether these changes differ across treatment conditions. For example, we will examine whether the mean within-person difference between the 9-month assessment and baseline assessment of intention to use services differs for subjects in Cornerstone versus those in TAU. We will likewise examine whether within person changes in the proposed mediating variables differ across intervention groups.

The RCT protocol for all research aims has been reviewed and approved by the New York University Committee on Activities Involving Human Subjects as of November 30^, 2016 (IRB-FY2016-172). For protocol modifications the research team will submit an amendment to the Committee on Activities Involving Human Subjects. Consent (and assent for those under 18) to enroll in the study will be processed and obtained by the PI or trained project research staff who have all been trained on activities involving human subjects.