Preliminary Evaluation of Washington State’s Early Intervention Program for First-Episode Psychosis

The New Journeys treatment model includes individualized medication management, family psychoeducation, individual resiliency training, and supported employment and education components of NAVIGATE (8, 9). Individualized medication management focuses on a shared decision-making approach to reduce symptoms and prevent relapses while managing side effects. The goal of individual family psychoeducation is to teach families about psychosis and its treatment while reducing family stress through communication and problem-solving skills.

Individual resiliency training is an individually delivered psychotherapeutic intervention based on positive psychology, illness self-management, and cognitive-behavioral therapy. It focuses on helping clients achieve their personal goals and develop resiliency and skills to target ongoing challenges. Supported employment and education are evidence-based models that support clients in developing and achieving employment and education goals related to their career interests (10). All components of the New Journeys intervention are delivered in person. On the basis of client and family needs, clinicians determine whether family psychoeducation, individual resiliency training, and supported education and employment are delivered on site, within the home, or in the community. Each site utilizes a team-based approach facilitated by weekly meetings to coordinate treatment planning and engagement for each client. To serve a maximum of 30 clients, each team consists of one psychiatrist or advanced registered nurse practitioner (0.2 full-time equivalent [FTE]) and bachelor’s- or master’s-level mental health professionals in the following roles: program director or family education specialist (1.0 FTE), individual resiliency training clinician (1.0 FTE), supported employment and education specialist (1.0 FTE), and case manager (0.5 FTE). A peer specialist role (0.5 FTE) was added to the New Journeys model after the initial implementation period.

Consistent with the elements of a learning health system, and to increase the efficiency of data collection from clients, the implementation and evaluation team along with program developers, clinicians, and administrators modified and developed an existing Web-based dashboard (Toolkit) that would provide both data infrastructure and support a clinical feedback system. Before the implementation of New Journeys, clinicians and administrators provided feedback on the usability and acceptability of the Toolkit prototype, which was used to improve the Toolkit prior to implementation. The Toolkit facilitates the implementation, training, and supervision of evidence-based treatments, and it is accessible by log-in only.

Each year, Washington State distributes a request for information for FEP sites, using funds from the Community Mental Health Services State Block Grant, to all behavioral health organizations and managed care entities in the state. Interested community-based sites are required to apply in response to the request for information. Applications are assessed on population need, experience, community and clinical resources, agency capacity, and sustainability. Determinations are based on the best fit for providing services. Through this mechanism, New Journeys has been implemented sequentially across five counties throughout Washington State. The evaluation team had no direct contact with program clients. The Washington State Institutional Review Board determined that this program evaluation was not human subjects research.

Between October 15, 2015, and December 31, 2018, a total of 201 potential clients were referred to five clinical sites, of which 112 clients were eligible after screening. Program eligibility included clients ages 15–40; a primary diagnosis of schizophrenia, schizoaffective disorder, schizophreniform disorder, brief psychotic disorder, or other specified psychotic disorder, as assessed by the Structured Clinical Interview for DSM-5 Disorders, Clinician Version (SCID-5-CV; 11); and presence of psychotic symptoms for greater than 1 week and less than 2 years and less than 6 months of treatment with antipsychotic medications. In 2018, a diagnosis of delusional disorder was added to the primary diagnosis inclusion criteria. All insurance types are accepted, and clients who are uninsured are also eligible to receive services. Exclusion criteria included psychosis due to substance intoxication or withdrawal or known to be the result of a medical condition, and more than one episode of psychosis (i.e., a psychotic episode that was followed by full symptom remission and then relapse to another episode).

During the first 2 years, New Journeys team members participated in yearly in-person training conducted by NAVIGATE treatment developers and the implementation team. Trainers conducted a 2-day workshop that included large group presentations on current and past treatment practices and breakout sessions on specialty roles (i.e., program director, family clinician, individual resiliency training clinician, supported employment and education specialist, and prescribers) and corresponding treatment components, which consisted of demonstrations and behavioral rehearsals. In addition to the NAVIGATE training, team members participated in a 2-day SCID-5-CV training course conducted by a certified SCID trainer. New Journeys team members also participated in a 1-day of training that provided an overview of the Toolkit, information on clinical and functional measures, and an interpretation of results conducted by two members of the evaluation team. Training involved role play, mock data entry, and didactics. To address the training needs of new staff because of turnover, program directors conduct orientations to New Journeys training. Subsequently, the implementation and evaluation team coordinates with program directors to schedule role-specific measurement and data input training.

Two separate universities hold contracts with the state to provide implementation and evaluation, respectively. In year 1, sites participated in biweekly (for the first 6 months) and monthly phone-based consultations conducted by the NAVIGATE treatment developers. The implementation and evaluation teams provided continuous technical assistance and data monitoring. New Journeys site directors participated in individual monthly calls focused on missing data, site performance reports, barriers to data collection, and staff turnover.

The Toolkit provided the online data management support for all sites. The evaluation team selected reliable and valid measures that captured areas of assessment in previous literature, provided meaningful clinical cut-points, were brief, and were at no cost to New Journeys sites (12, 13). The measures selected were a mixture of self-reported and clinician-rated instruments to reduce the burden associated with collecting data. Self-reported measures were completed during appointment times, either directly by the client or delivered by trained providers at the request of the client. Demographic information, psychiatric symptoms, information on real-world functioning, and medical history were collected at intake by site staff.

To examine differences in service utilization among New Journeys clients 24 months before and up to 24 months after intake, we obtained data using the Integrated Client Database, which contains individual-level administrative data from multiple state agencies on client service risks, history, and outcomes (27). The 24-month window was selected to match the maximum timeframe of services clients could receive. Prescribed medications, emergency room (ER) visits, inpatient hospitalizations from three different facilities (i.e., medical hospitals, community psychiatric programs, state-funded psychiatric programs), and the use of behavioral health organization mental health services were assessed for only New Journeys participants with Medicaid claims and encounters. Administrative data were used to identify clients who had any services from Washington State’s Economic Services Administration, which includes Temporary Assistance for Needy Families, Basic Food, and other public assistance programs (e.g., Diversion Cash Assistance, Supplemental Security Income–State, Disability Lifeline Unemployment). Administrative data from different sources have varying lengths of delay, and clients admitted into the program after the latest data update were excluded from the analysis. As a result, the final sample sizes for these measures differ by data source.

Descriptive analyses of demographic variables were conducted. Separate generalized estimating equations (GEEs) were used to assess all measures at baseline compared with the first 12 months of treatment. Twelve months was chosen as the cutoff point to maximize the number of available pairs used in GEE and to reduce the effect of missing data. For example, in measuring PHQ-9 scores, 83 clients (74%) were eligible for analysis when we used the 12-month window, which was consistent across measures. GEEs were also used to assess administrative data obtained from Medicaid-eligible participants 24 months preintake and up to 24 months postintake. Administrative data were retrieved for publicly insured eligible clients from four New Journeys sites (N=84). One site was omitted because of its brevity in the New Journeys network, and an additional seven clients were missing intake dates and were excluded from analyses. Given the various intake dates and lengths of treatment among clients, those who had an intake date after the reporting period for each agency were also excluded. Unstandardized regression coefficients and 95% confidence intervals (CIs) are presented for continuous outcomes. Analyses were performed with SPSS, version 25.0, and statistical significance was set at p<0.05.

Demographic and clinical characteristics for the 112 clients eligible for New Journeys are presented in Table 1. Of the eligible clients, 90 clients (80%) were actively (one or more services monthly) receiving services as of December 2018, and the mean±SD duration in the program was 10.62±8.58 months. Of the remaining clients not receiving services, 12% (N=13) discontinued services after 8.10±8.50 months, and 8% (N=9) successfully graduated or completed after 9.88±9.76 months.

Preliminary analyses on clinical outcomes found significant decreases (or improvements) on the CRDPSS (N=90; β=−1.47, 95% CI=−2.47 to −0.46, p<0.05), CAPE-P15 (N=82; β=−3.37, 95% CI=−4.89 to −1.84, p<0.05), GAD-7 (N=87; β=−2.48, 95% CI=−3.74 to −1.22, p<0.001), and the number of unhealthy days per month (N=84; β=−5.95, 95% CI=−8.85 to −3.05, p<0.001) during the first 12 months of treatment after controlling for site (Figures 1 and 2). There were no significant differences in depression scores assessed by the PHQ-9 or substance use during the first 12 months of treatment. Clients (N=70) were significantly more likely to attend school (odds ratio=1.42, 95% CI=1.02 to 1.96, p<0.05) in the first 12 months of treatment. However, there was no significant change over time for employment.

Of the 112 eligible clients, 18% (N=20) had no data beyond intake, and 82% (N=92) received some services beyond intake. However, an additional 15 clients did not receive any supported employment and education sessions, and nine clients did not attend any individualized medication management sessions. Among family members, 14 did not receive any family psychoeducation (Table 2). Of 92 clients with at least one scheduled appointment beyond intake, clients attended two individual resiliency training sessions (1.84±1.07), one supported employment and education session (1.18±1.06), and one individualized medication management session per month (0.66±0.44), whereas family members attended approximately one family psychoeducation session per month (0.84±0.70).

Administrative data before and after patients became New Journeys clients are presented in Table 3. The number of clients (N=67) who were prescribed antipsychotic medication (β=1.53, 95% CI=1.17 to 2.00, p<0.05) significantly increased from pre- to postintake. There was a significant decrease in ER visits with a psychiatric diagnosis (β=0.22, 95% CI=0.08 to 0.56, p<0.05) and without a psychiatric diagnosis (β=0.18, 95% CI=0.06 to 0.52, p<0.001) after intake. Client (N=84) utilization of state-funded community psychiatric hospitals (β=0.31, 95% CI=0.16 to 0.60, p<0.001) significantly decreased from pre- to postintake. In addition, client (N=84) utilization of public assistance significantly decreased (β=0.71, 95% CI=0.57 to 0.88, p<0.05) after participating in New Journeys.