The STEP Learning Collaborative: A statewide Learning Healthcare System (LHS) for First Episode Psychosis

Abstract

The STEP Learning Collaborative (STEP-LC) is Connecticut’s statewide learning healthcare system (LHS) for individuals aged 16–35 with recent-onset schizophrenia-spectrum disorders. The goal of the STEP-LC is to ensure that all individuals with schizophrenia have the opportunity to lead fulfilling lives by facilitating rapid access to high-quality treatment during the critical early stages of illness. STEP-LC emphasizes improving local pathways to care, reducing the Duration of Untreated Psychosis, and enhancing population health outcomes.

The STEP-LC is a model implementation of a statewide learning health system for recent-onset schizophrenia spectrum disorders. This paper provides a detailed account of the rationale, design and ongoing implementation of the LHS. Several core workstreams organize the activities of the statewide system: workforce development, early detection, informatics, stakeholder engagement, sustainability and care model refinement.

The STEP-LC has demonstrated notable initial success in creating a robust system of care for early psychosis across Connecticut. It has implemented comprehensive workstreams to address critical gaps in workforce readiness and information. By leveraging informatics and continuous quality improvement workflows, STEP-LC aims to meet or exceed international benchmarks for access and outcomes of care.

Ongoing efforts are focused on further refining care models, sustaining best practices, and strengthening integration with state services through targeted policy advocacy and coordination. This paper serves as a practical guide for those planning to build LHSs in their regions, offering detailed implementation insights to take readers ‘behind the scenes’ and providing accessible tools for operationalization.

1. Background

The STEP Learning Collaborative (STEP-LC) is Connecticut’s statewide learning healthcare system (LHS) for individuals aged 16–35 with recent-onset schizophrenia-spectrum disorders. The goal is to ensure that all persons with schizophrenia can live fulfilling lives by providing rapid access to high-quality treatment for those with recent onset illness. STEP-LC focuses on improving local pathways to care (PTC), reducing the Duration of Untreated Psychosis (DUP), and enhancing population health outcomes. The STEP-LC is a model implementation of a statewide learning health system for recent-onset schizophrenia spectrum disorders. This paper describes the rationale and design of STEP-LC. Details about implementation are provided to take readers ‘behind the scenes’ and provide tools for operationalization elsewhere, alongside links to continuously updated online assets

1.1. History and Development of STEP’s Early Intervention Service: Research to Practice

The Specialized Treatment Early in Psychosis (STEP) program (www.step.yale.edu) began as a pilot clinic in 2006 within a Department of Mental Health and Addiction Services (DMHAS)-Yale partnership at Connecticut Mental Health Center (CMHC)(Srihari, et al., 2009). Between 2007–13 STEP (with Donaghue Foundation and NIH grants) completed the first U.S. randomized trial demonstrating the effectiveness of a specialty team-based service for early course schizophrenia spectrum disorders (Srihari et al., 2015). The clinic has since been recognized as a model for ‘Coordinated Specialty Care (CSC)’ by CMS and NIMH, and it was influential in the launch of a federal funding initiative that has catalyzed the rapid growth of similar clinics across the U.S since 2016 (Dixon et al., 2018).

STEP subsequently completed the first U.S. experimental test of early detection (2014–19) (www.mindmapct.org). The Mindmap campaign successfully halved access delays, as measured by DUP, across a 10-town catchment (Srihari et al., 2022).

STEP has deployed the results of this research into a population health based Early Intervention Service (EIS) that delivers both rapid access (ED) and high-quality care (CSC) to all residents of a defined geographic catchment (Srihari et al., 2015). See Figure 1 for STEP’s Early Intervention Service (EIS) Care Pathway.

Figure 1.

STEP’s Early Intervention Service (EIS) Care Pathway, developed from research trials of Coordinated Specialty Care (CSC) and Early Detection (ED) is comprised of 4 modules.

STEP’s operational costs were initially supported within CMHC funds, and subsequently by specific first-episode psychosis (FEP) allocations from the CT legislature (2013-present) and SAMHSA Mental Health Block Grant (MHBG) set-aside funds (2014-present). All research activities were supported by grants from the NIH and research foundations. This Public-Academic model has thus been able to provide high value care that is stably supported by recurring implementation funds, while competing for time-limited research grants to study and further refine the service model.

1.2. Outline

This paper serves as a practical guide for those planning to build Learning Healthcare Systems in their regions, offering detailed implementation insights to take readers ‘behind the scenes’ and providing accessible tools for operationalization. The following provides an outline of sections that will be covered in this paper.

2. Development of the STEP Learning Collaborative (STEP-LC)

There has been growing interest in expanding access to best practice care for FEP to target unmet need in Connecticut. STEP’s EIS aimed to engage all new individuals experiencing FEP living in its 10-town catchment (population ~ 400,000) in the Greater New Haven area of southern Connecticut, but based on combined census and predicted incidence data, this represented less than 10% of estimated new cases in the State. The following timeline outlines the steps leading to the development and implementation of the STEP Learning Collaborative in February 2024.

2.1. Workforce Development for Pediatric Mental Health

In May 2020, the CT Department of Children and Families (DCF) approached the program to assist with educating clinicians in the evaluation and care of FEP. In Connecticut, responsibilities for the public mental healthcare system are divided between DCF (under 18) and DMHAS (adults). STEP delivered a workforce development initiative, targeting key health care providers with the goal of increasing the capacity of Connecticut-based, DCF supported child guidance clinics to provide best-practice care to individuals experiencing FEP. STEP hosted regular ECHO- based case discussions and brief didactics to educate and support the care of front-line providers working with these individuals. Example topics included CBTp and other evidence-based therapy techniques for treating FEP, as well as guidelines for prescribing antipsychotic medications for FEP. A website was established (https://www.ctearlypsychosisnetwork.org) to promote trainings, events and resources. The initiative grew and began to provide additional trainings and workshops across the state targeting all stakeholder groups, including mental health providers, PCPs, families, young adults, advocacy organizations, and other community members. Although formal measures of provider attitude and behavioral changes were not assessed, ongoing needs assessments were conducted, which highlighted a variety of unmet, regionally specific needs across Connecticut. Such feedback included: improved access to care and referral lists for individuals experiencing FEP across the state and in specific resource-bare areas; increased availability of educational resources for providers including webinars and virtual resources such as tip sheets; opportunity for consultation with early psychosis experts; resources to support families both directly (workshops for families) and indirectly (trainings on family support and education for clinicians); and resources for Spanish speaking individuals. This feedback was informative for the subsequent formation of the learning collaborative.

Prior to this effort, the STEP clinic had served as a resource for clinicians, healthcare organizations, advocacy groups and policymakers across the country. However, this DCF-funded effort, to meet the educational needs of healthcare providers serving minors, was the first organized effort to scale impact across healthcare organizations in Connecticut.

2.2. The American Rescue Plan Act (ARPA)

ARPA (also referred to as the COVID-19 stimulus package) was enacted on March 11, 2021to address the economic and health impacts of the pandemic. The $1.9 trillion allocation included supplements to the federal Mental Health Block Grant sent to each US state. Since 10% of these funds were already stipulated for FEP care, this supplement presented an opportunity to target unmet need in Connecticut. With funding limited to 4 years, it was necessary to make strategic investments that would maximize long-term impact on health outcomes and minimize the disruptions expected with the loss of funding in October 2025. The existing Public-Academic collaboration between the STEP Program and the two State behavioral health agencies provided a natural platform for developing a statewide strategy.

2.3. Establishment of Statewide FEP Planning Committee

In September 2021, representatives from various agencies including DMHAS, Young Adult Services (a division of DMHAS focused on the behavioral needs of 18–25 year olds many of whom are transitioning from DCF to the adult mental health system), DCF, and STEP began meeting to develop a strategy to address the unmet needs of individuals experiencing schizophrenia spectrum disorders in the state. Presentations by leaders of other statewide efforts to improve FEP care helped advance discussions. There was strong consensus on a shared goal of enabling full lives for all individuals experiencing FEP across the state. The ensuing mission for the STEP Learning Collaborative was developed: To transform access and quality of care for all individuals with recent onset schizophrenia spectrum disorders (or FEP) (Breitborde et al., 2009). Population outcomes would be measured and used to drive performance improvement and benchmarking to international standards (Srihari, et al., 2015, Mathis et al., in current issue). The target population was defined to include all residents of CT ages 16–35 and within 3 years of psychosis onset. No exclusions would be made for health insurance status and active efforts would be made to address disparities based on membership in vulnerable sub-groups (including race, ethnicity, gender, immigration status and socioeconomic status). The committee has continued to meet on a regular basis to monitor progress with quarterly reports that include metrics of access and quality and prioritize outcomes relevant to patients and families (Srihari et al., 2016).

2.4. Clinical Infrastructure of the STEP Learning Collaborative

The planning committee proposed that the STEP-LC utilize the state’s existing network of 13 Local Mental Health Authorities (LMHAs) that are operated or funded by DMHAS, the state mental health department. The LMHAs are clinical sites that provide therapeutic programs and crisis intervention services for individuals experiencing mental health and substance use disorders throughout the state. Consistent with STEP’s population health-based approach, the LMHAs cover 5 regions of Connecticut and each LMHA has a designated catchment area offering the network the ability to cover every eligible individual in the state, regardless of their location (https://portal.ct.gov/dmhas/programs-and-services/dmhas-directories/local-mental-health-authorities). The STEP Program, housed within the Connecticut Mental Health Center, one of the LMHAs, acts as the organizing and expert hub (STEP-Hub) within the collaborative, while also hosting a clinical site (STEP Clinic) which models best practice care for the rest of the partner sites (the 12 other LMHAs). Following this catchment-based approach, individuals who engage in the STEP-LC are referred to care at the LMHA responsible for serving their zip code of residence. The acronym LMHAs will be used throughout this paper to refer to the clinical sites within the STEP-LC.

STEP is the only Coordinated Specialty Care (CSC) clinic within this statewide network of diverse clinics. The design of this network has taken an explicitly social-ecological view that seeks to leverage the strengths and expertise within the public mental health infrastructure of the State and begin with explicit measures of population health to guide process improvements, as opposed to starting with changing care models. There is already substantial variation within established CSCs that likely reflects this need to allow for innovative and diverse ways to meet population needs that may vary across the catchments of the individual spokes/healthcare agencies (Keshavan, et al., 2022; Srihari, et al., 2016). This is indeed a novel aspect of our learning health system, versus existing models that either envision CSCs as spokes (e.g., NIH EPINET) or utilize a franchise-oriented model that aims to replicate the same clinical model across all sites with fidelity (e.g., OnTrack NY).

2.5. Engagement with Network of Clinical Sites

Members of the FEP Planning Committee began conducting in-person engagement meetings (Jan-March 2023) with the LMHAs to understand current practices, strengths, needs, and interest in order to better serve individuals experiencing FEP within their agencies and communities. Several staff who expressed strong enthusiasm were involved as “early psychosis champions,” across all agencies to coordinate with the Hub. Agencies were also asked to capture data on their existing FEP census, including how many individuals met FEP criteria (16–35 years old, within first 3 years of primary psychotic disorder) and on which clinical teams these individuals are being served.

2.6. Overview of Early Intervention Services (EIS) for Schizophrenia Course

The STEP-LC leadership hosted a 6 session, “Overview of Early Intervention Services (EIS) for Schizophrenia” course which oriented participants to STEP’s Early Intervention Service Care Pathway to gain an understanding of the structures, processes and outcomes that can guide implementation (see Figure 1). The course targeted leaders and managers who wish to consider and adapt this evidence-based model of EIS for delivery in their healthcare agency. Attendees included at least two early psychosis champions from each of the LMHAs (Spring 2023). The course is publicly available here: https://www.ctearlypsychosisnetwork.org/overview-of-eis-course1.html.

2.7. Regional Events (Fall 2023)

Following the education and engagement of LMHA leadership through previous meetings and the EIS course, regional events were held to kick off LMHA involvement in the STEP Learning Collaborative statewide initiative. These events were held regionally, bringing together all of the LMHAs in the surrounding area. The events were used to foster support and anticipation prior to full public implementation of the STEP-LC.

Background information was provided to orient LMHAs to the evidence-base and rationale for the STEP-LC, with an emphasis on the importance of rapid access to care and shortening the Duration of Untreated Psychosis (DUP). A phased outline of plans to implement a statewide learning health system was discussed (1: Access, 2: Care Quality). Regions were also introduced to the Early Detection and Assessment Coordinator (EDAC) assigned to their region. The EDACs act as the main STEP-LC contact in the region for both the LMHAs and individuals experiencing psychosis. EDACs lead early detection efforts in the region across stakeholder groups (e.g., hospitals, schools, community agencies, advocacy, judicial), capture new cases via a central referral line, assess eligibility, and support engagement into care at the LMHA.

These meetings helped to establish a culture of community and collaboration around shared values, aims, mutually agreed upon outcomes, and to cultivate a shared responsibility and agreement to accept into care ALL individuals experiencing FEP, regardless of insurance status, as early in the course of their illness as possible. Agencies were also offered incentives (EDAC support, free training and consultation opportunities, financial incentives from state agencies to help offset costs of clinical staff no providing billable clinical care during training opportunities and STEP-LC meetings).

At the conclusion of the regional events all LMHAs had agreed to be a part of the STEP Learning Collaborative and to begin accepting eligible individuals by the initiative’s official statewide implementation date of February 1^st, 2024.

2.8. Launching the Early Detection Campaign Mindmap Statewide

On February 1^st, 2024, the STEP Learning Collaborative launched the statewide Early Detection campaign, Mindmap. This included the launch of the state’s centralized referral number for early psychosis screening and referrals, the readiness of LMHAs to accept all eligible individuals experiencing FEP as quickly as possible, and the start of statewide data collection for this population.

This occasion was marked by a virtual launch event, which outlined the goals and components of the Mindmap campaign and STEP Learning Collaborative (Available to view here: https://www.youtube.com/watch?v=mZlGBDbfziE). The event included testimonials from various stakeholders including government officials and commissioners of leading state agencies to demonstrate buy-in from vast stakeholder groups. Attendees included various community members, as well as providers and representatives from each of the agencies in the STEP Learning Collaborative.

3. STEP Learning Collaborative Design Principles and Goals

3.1. Dissemination of STEP’s Population Health Model

Infused throughout the aforementioned “Overview of EIS for Schizophrenia” course and all meetings with STEP-LC members was an explicit focus on using STEP’s Population Health Model as a common framework to address unmet need across Connecticut. The STEP-Hub team oriented and collaborated with leaders across State LMHAs to develop a statewide strategy to deliver EIS that responds to the following core Population Health principles:

• Social-ecological: adaption of empirically-based early detection and treatment approaches to local needs and resources

• Inclusion of all relevant stakeholders within and outside the healthcare system (e.g. education, clergy, consumer organizations, criminal justice, etc.) to intervene on all determinants of health (beyond healthcare), including social and structural determinants of health

• Focus on addressing disparities by measuring patient outcomes in order to guide improvements in access and care quality.

The box below outlines the strategic plan for implementation of the STEP-LC, a statewide learning healthcare system for all individuals experiencing recent onset schizophrenia across Connecticut.

STEP-LC Overall Goal: all persons with schizophrenia will live fulfilling lives

The target population includes: all residents of CT 16–35 years old and within 3 years of psychosis onset. No exclusions for health insurance status.

What: To continuously improve access and quality of care for all individuals with recent onset schizophrenia or FEP. To do this by measuring outcomes benchmarked to international standards and addressing gaps with quality improvement and research.

Who: All clinical and community stakeholders who can help improve pathways to or through care or any determinants of health

How: Via a statewide network of behavioral healthcare agencies built on a public-academic ‘backbone’ organized to deliver a population-based learning health system

Phase 1: Access (Summer 2023- Oct 1^st, 2024)

Phase 1 of the STEP Learning Collaborative statewide roll-out is focused on transforming Access to care. Specifically, the goal of Phase 1 is to improve local pathways to care (PTC) and reduce the Duration of Untreated Psychosis (DUP) across all 5 Regions of the state, via the following deliverables:

• Establish a statewide network of LMHAs

• Establish workflows for referral pathways

• Launch Mindmap 2.0

• Establish workflows for data collection, analysis and review

Phase 2: Quality (Oct 1, 2024 – present)

Phase 2 of the STEP Learning Collaborative is focused on transforming care quality. Specifically, to improve population health outcomes across all 5 Regions of the state, via the following:

• Establish standardized collection of outcomes across key population health metrics

• Share outcomes: leverage expertise across network for quality improvement.

3.2. Guiding Principles of LHS

The STEP-LC exemplifies how to create a statewide learning health system for recent onset schizophrenia spectrum disorders or FEP. The STEP-LC is guided by the principles of a learning health system (LHS), as outlined by the Institute of Medicine including utilizing evidence-based science, informatics, incentives and culture with a goal of continuous improvement and innovation (Friedman et al., 2016). The STEP-LC also views stakeholder inclusion as a foundational principle that pervades all elements of the LHS.

The Institute of Medicine (2007) provides a useful framework to guide development and implementation of a learning health system for recent onset schizophrenia, including attention to the following interrelated principles:

• Science: through the collection of valid and reliable measures, innovative statistical and translational research technologies, and the identification, generation and free exchange of knowledge and expertise between researchers and clinicians

• Informatics: robust, user-friendly and collaboratively designed software that allows for a range of data collection methods, integrating with existing databases/health records and interfacing with all stakeholders

• Incentives: by facilitating accountability to key stakeholders, fostering creativity, growth and optimism and empowering members with support, tools and resources

• Culture: via the formation of non-judgmental, diverse communities of agencies which embrace both local ownership of care and a shared core mission of measuring and learning from patient outcomes

See Figure 2 for a visualization of the STEP-LC LHS.

Figure 2.

The STEP Learning Collaborative learning healthcare system (LHS) for recent-onset schizophrenia delivers a system of care that will organize local networks of stakeholders to transform pathways to and through care to improve defined, benchmarked outcomes. FES = First Episode Services. Q/A = Quality Assurance. QI =Quality Improvement.

4. Clinical and Assessment Infrastructure for the STEP-LC

As discussed, the STEP-LC partners with 13 Local Mental Health Authorities (LMHAs) across 5 regions of Connecticut. In instances in which the designated LMHA only serves adults (18+), the local DCF-affiliated child guidance clinics act as the referral source for those under 18. Each region has an Early Detection and Assessment Coordinator (EDAC) who captures new cases via a central referral line, assesses eligibility, and supports engagement in care. The STEP Program acts as the organizing and expert hub (STEP-Hub) within the collaborative, while also hosting a clinical site (STEP Clinic) which models best practice care for the rest of the partner sites.

In addition to the EDACs, members of the STEP-Hub operate in multiple roles, including offering clinical services at the STEP clinical site, as well as outward facing roles within the STEP-LC such as overseeing engagement with different stakeholder groups (e.g., education, judicial, healthcare/hospitals, workforce development, community partnerships).

4.1. Screening and Rapid Eligibility Determination

Each inquiry to the STEP-LC is fielded by the EDACs, who screen all callers for eligibility and imminent safety concerns. Screening aims to be brief and inclusive to prevent further delays in access. EDACs review cases within daily “huddles” and adjudicate ambiguous cases with the STEP-LC Director. Screening form is available upon request.

Eligible individuals are offered participation in the STEP-LC, which includes a streamlined referral to care within the STEP-LC network, support with engagement, education to individual and involved family/collateral, assessment and ongoing monitoring of outcomes. Those who are ineligible are still directed to care and offered relevant resources, including a consultation service and workforce development offerings for providers, educational workshops and resources for families. See Figure 3 for details of the Referral and Intake workflow.

Figure 3.

STEP LC Referral and Intake Workflow. EDAC = Early Detection and Assessment Coordinator. FEP = First Episode Psychosis. PTC = Pathway to Care. DUP = Duration of Untreated Psychosis

Inclusion Criteria:

16–35 years old; must live within the state of Connecticut and have had their first episode of psychosis within the past 3 years, regardless of insurance status or type.

Exclusion Criteria:

Greater than 3 years since psychosis onset, established diagnosis of Affective psychosis (i.e., non-ambiguous Bipolar Disorder or MDD with psychotic features) or psychosis secondary to substance use or a medical illness, eligible for DDS (Department of Developmental Services), unstable serious medical illness.

4.2. Measurement

Individuals who agree to participate in the STEP-LC complete a baseline assessment with the EDACs. A summary form is shared with the receiving clinical agency. Participants then receive follow-up assessments via REDCap surveys every 3 months for ongoing monitoring of health outcomes and connection to care. Surveys are automated and sent to the designated individual’s email, with an option to text their phone. EDACs regularly prompt survey recipients to fill out the surveys and/or send the survey to the next approved designee if the patient is not responding (e.g., clinician and/or family support person). Assessment completion is reviewed at a weekly internal meeting to track progress and troubleshoot issues. All 56 eligible individuals completed their STEP-LC baseline assessments. Completion rates for follow-up assessments are as follows: 3 months (16/38 = 42%), 6 months (16/27=59%), and 9 months (12/16 = 75%). Note, “complete” indicates that some source of data has been used for the participant’s survey including self, family/collateral, or clinician report (not missing). The assessment battery and procedures were finalized during the first few months of implementation thus we expect these rates to continue to improve as further quality improvement measures are used to increase response rates. Payment has been considered but not yet implemented to incentivize response.

Prior to administering assessments, EDACs were trained on all measures and complete reliability assessment to Gold Standard ratings. Reliability assessments are repeated annually with correct answers and rationale reviewed as a team. If an individual varies from the Gold Standard ratings in a way that changes the score for a participant’s rating (e.g., from mild to moderate), then additional training and review is offered until acceptable reliability is achieved. Intraclass correlation coefficients (ICC) demonstrated a high degree of interrater reliability. Scores were “excellent” for PANSS with ICC= 0.914 and a 95% CI [.864 to .952] and “good” for GAF with ICC= 0.803 and a 95% CI [.334 to 1.00]. See Appendix Table 1 for details on baseline and follow-up assessment measures.

4.3. System Specification - Population Health

The STEP-LC utilizes transparent measures of impact on population metrics that prioritize patient centered outcomes while considering shared stakeholder values. This approach ensures efficient resource allocation, supports quality improvement, and prioritizes the health status of the target population over fidelity to specific care models. The ‘deliverables’ of the STEP-LC focus on (1) Access Outcomes, including objectives such as Rapidity, Equity, Coverage, and Pathway to Care; and (2) Patient Outcomes, including Hospitalizations, Remission, Vocational Engagement, Cardiovascular Risk, and Suicide Prevention. Additional items under development include Caregiver Burden and Judicial Involvement. A detailed system specification of these population health outcomes is outlined in Srihari et al., 2016 and are based on prior work at STEP wherein we: (i) elicited broad stakeholder input to inform us on which outcomes were of greatest importance; (ii) converted these to Objectives and Measures that could then be analyzed and displayed to allow real time comparison to prior (achievable) and international best practice (aspirational) Standards.

A prior version of this list was published and continues to be refined based on this iterative multi-stakeholder process. The current version is designed as a working list, with the aspirational standards to be modified as new evidence emerges, while the achievable are based on a prior year’s performance. Critically, this means services can be transparently evaluated based on patient-centered outcomes and be supported in quality improvement (for lagging outcomes) or in sharing successful practices. STEP has also developed care process metrics (available upon request) to assist with traditional quality assurance activities. The STEP Learning Collaborative will emphasize a shared focus on patient outcomes within a culture that incentivizes use of such data for meaningful feedback to drive improvement and values innovative adaptations of empirically based care models to local needs and resources. This is consistent with learning health systems deployed for other chronic illnesses.

These outcomes will make transparent the value of the State’s investment in EIS, allow efficient resource allocation to improve lagging indicators, and continue to be open to stakeholder-driven refinements. For example, in response to family feedback, a measure of family caregiver burden is in development at STEP for planned inclusion in a future version of the core outcomes set.

5. Core Workstreams and Deliverables of the STEP Learning Collaborative

To achieve the overall goals of the LHS, the STEP-LC utilizes 6 main workstreams: workforce development, early detection, stakeholder engagement, informatics, sustainability, and care model refinement. This section will discuss in detail the design and goals of each of these workstreams. Corresponding boxes for each section report on progress and deliverables and in some cases aspirational deliverables.

5.1. Workforce Development

Workforce development initiatives, delivered by experts at the STEP-Hub, aim to enhance the capacity and quality of early psychosis treatment through structured curricula, case-based discussions, virtual resources, webinars, and the STEP Consultation Service. Free Continuing Education Credits are available to incentivize busy providers and agencies to attend trainings. If providers are unable to join trainings live, resources are available on a centralized website: https://www.ctearlypsychosisnetwork.org/.

The STEP-LC Provider Training Series.

The STEP-LC provides a monthly training series intended for front-line clinicians of various disciplines working to support individuals and families impacted by early psychosis. It is also open to interested community members. The series provides research based, practical information and strategies on topics core to delivering best practice care for individuals with recent onset schizophrenia spectrum disorders.

The STEP Consultation Service.

This is a free provider-to-provider service, available to clinicians, administrators and leaders of healthcare systems who would like to consult about their continuing care of young people with recent onset psychosis between the ages of 16 and 35. The consultation service is designed to be a low-barrier, curb-side discussion. A brief online request form is publicly available (https://www.ctearlypsychosisnetwork.org/consultation-service1.html) and experts at the STEP-hub aim to get back to the requester within 1 business day to discuss the identified case. The consultation service is flexible in its implementation, including the option to have the STEP-hub experts discuss cases at requesters’ existing team meetings, as well as to invite outside agencies to observe STEP-hub operations.

STEP-LC Family and Community Educational Workshops.

The STEP-LC hosts periodic family and community educational workshops, consisting of expert-led discussions on core curriculum and specialty topics of interest to young people and families impacted by early psychosis.

STEP-LC ECHO Case Consultations.

The STEP-LC hosts as needed, ECHO (Extension for Community Healthcare Outcomes) - based case discussions, providing community providers the opportunity to present and receive feedback on cases of individuals experiencing FEP to fellow community providers in the learning community and STEP-hub experts.

To keep the above educational content responsive to ongoing learning needs, the STEP hub hosts speakers to their internal clinical and research rounds to discuss the latest research and topics of interest related to early psychosis. Upon vetting presenters, they are invited to present at the STEP-LC Educational Series to share the latest updates in the field with both provider and/or community member audiences. The STEP hub team regularly utilizes feedback from presentation attendees about topics of interest and need to inform development of new content. We also aim to update foundational presentations yearly to include the most up-to-date information.

Offerings of the workforce development workstream are advertised widely through our websites, list-servs, physical flyers, targeted emails, and offered, as appropriate, to every referring provider and family member who calls the referral line

Deliverables:

Workforce development: through various offerings by the STEP-LC hub experts, Connecticut has made strides in creating an engaged and educated workforce. To date the following deliverables have been produced.

• Overview of EIS for Schizophrenia Course: https://www.ctearlypsychosisnetwork.org/overview-of-eis-course1.html

• Provider Trainings: Monthly trainings targeted at front-line clinicians working with individuals experiencing FEP. https://www.ctearlypsychosisnetwork.org/provider-trainings.html

• Consultation Service: Providers of varying disciplines have received expert consultations on topics related to FEP including: medication, psychotherapy, managing risk and access to firearms, and building a new service for FEP.

• Family and Community Workshops: the STEP-LC provides regular family and community educational workshops, consisting of expert-led discussions on core curriculum and specialty topics of interest to young people and families impacted by early psychosis.

• Family/Caregiver Course: https://www.ctearlypsychosisnetwork.org/families.html

• PowerBI Training Dashboard: An interactive dashboard displays the number of stakeholders reached based on training offered, discipline, agency, region, and other key variables. https://www.ctearlypsychosisnetwork.org/offerings.html#dashboards

• Virtual Resources: Tip sheets, educational blogs, and other evergreen educational materials have been developed and made available to the public (https://mindmapct.org/blog/ and https://www.ctearlypsychosisnetwork.org/early-psychosis-resource-library1.html

5.2. Informatics

The STEP-LC informatics workstream focuses on assessing the burden of illness and analyzing population health outcomes across the network. A core set of population health outcomes (outlined in Srihari et al., 2016) is evaluated against international standards and benchmarks. The STEP-LC prioritizes making data accessible to stakeholders, enabling them to ask diverse questions that drive continuous quality improvement. Regular data reviews facilitate sharing lessons across sites, fostering a culture aligned with Learning Health System (LHS) principles.

This workstream aims to deliver robust, user-friendly, and collaboratively designed software that supports various data collection methods, including individual and program-level data. It integrates with existing databases and health records—such as public data (census, community surveys), claims databases, and other state-collected data—while engaging stakeholders effectively. Since 2015, the STEP-LC has iteratively developed informatics platforms tailored to the needs of an LHS for first episode psychosis (FEP). The latest platform, Health Outcomes Network Education (HONE), is described further in Mathis et al. (this issue). Additionally, the informatics platform will include region-specific asset maps to guide efforts in other workstreams, such as early detection, stakeholder engagement, and sustainability.

Thus far, LMHAs implemented a new “estimating psychosis onset” question to their intake process to support calculating DUP, an example of how the LMHAs have incorporated population health outcomes into their usual practices. The question, adapted from the EPINET Core Assessment Battery, is as follows “Provide a best estimate* of when frank (not prodromal) psychotic symptoms (e.g., delusions, hallucinations, or disorganized speech/behavior) began. (Month/Year).” This supports the STEP-LC in capturing data on the burden of illness and access outcomes for those individuals who are entering care directly at the LMHAs. Each agency submits a quarterly report on the census of individuals aged 16–35 within their service, which allows for the capture of individuals entering care directly at the LMHAs in addition to the centralized referral line.

5.2.1. Informatics to drive continuous Quality Improvement (QI)

Critical as part of the culture of the LHS is to congratulate the agency on being willing to assist with the collection and use of data to guide their performance. The advantage of our population health driven LHS based design is that feedback with the LMHA or healthcare agency begins with discussion of deviations in patient outcomes, rather than focus on lack of fidelity to internal care processes. Although STEP acts as the coordinating hub, there is also the expectation that the local agency staff are the experts on what changes may be necessary to improve the measured outcomes. A continuous Quality Improvement (QI) framework is used to guide local QI efforts and as needed assistance will be offered to the site from the STEP Hub as well as State QI staff. The site may need more than process refinements, and we will offer any necessary training to staff and communicate with State agencies on possible resource needs to help fill the performance gap.

Relatedly, existing strengths of LMHA sites, team structures, and local expertise are highlighted and leveraged. For example, if one site has an effective strategy of engaging with local police departments through an embedded mobile crisis clinician, strategies and processes are shared across other LMHAs to see if lessons learned can be implemented by other LMHAs across the state to improve their outcomes (e.g., reducing unnecessary criminal justice involvement).

Deliverables:

Informatics: The development and implementation of the HONE platform has addressed critical challenges in establishing a LHS for FEP, including:

• Creation of a user-friendly digital platform, HONE, focused on a priori outcome measures and employing a question-based framing of data, to capture burden of illness and population health outcomes across the network

• HONE prioritizes streamlined data collection and data interpretability for clinicians.

• Support quality improvement across the network by empowering clinicians to make evidence-based decisions

• Create regionally specific asset maps to help direct the efforts of other workstreams (e.g., early detection, stakeholder engagement, sustainability).

• These innovations collectively contribute to improved patient care and exemplify the principles of an effective LHS. See Mathis et al., in this issue for detailed evolution and description of the STEP-LC Informatics platform.

*A demonstration of the HONE Platform, including the Population Health Outcomes specification is available upon request.

5.3. Early Detection

The STEP-LC developed and launched a statewide Early Detection campaign, Mindmap (www.mindmapct.org), which aims to reduce the DUP across the state. The campaign utilizes a three-pronged approach: (1) Public education campaign (PEC) via mass media outreach, including news appearances (television), press releases, transit, billboards and radio ads, targeted social media ads, google search terms, focused on optimizing Search Engine Optimization (SEO) of the Mindmap website and related web content; (2) Professional outreach and detailing (POD) of both clinical and non-clinical stakeholders – guided by the ACME framework (Audience, Channel, Message, Evaluation) (Noar, 2012), including activities such as outreach presentations and events, individual detailing meetings with potential referral sources and community stakeholders, as well as dissemination of branded promotional items (e.g., brochures, stress balls, pens, stickers, water bottles); (3) rapid access to service (RAS). RAS is achieved through use of a centralized referral line staffed by Early Detection and Assessment Coordinators (EDACs) who, upon screening individuals for eligibility, collaborate with the receiving clinical agencies (LMHAs) to rapidly refer and admit individuals experiencing FEP to appropriate outpatient care (see Figure 4 for path to treatment workflow).

Figure 4.

Overview of STEP LC’s path to treatment workflow from time of initial inquiry to the centralized referral line to ongoing follow-up of eligible participants. Individualized workflows have been developed in collaboration with each LMHA. EDAC = Early Detection and Assessment Coordinator.

The Mindmap campaign aims to deliver pathways to and through care that are faster and less aversive, limiting unnecessary criminal justice involvement and involuntary hospitalization. The Mindmap campaign was developed in collaboration with a social marketing company, Red Rock Branding. For detailed protocol on the Early Detection Campaign, Mindmap, see Srihari, et al., 2014.

The STEP-LC leadership meets weekly with the social marketing partners to implement the Mindmap campaign across the state. The approach is a complex intervention, including several interactive components (Srihari & Cahill, 2020; Srihari et al., 2014) and is responsive and iterative based on available data (e.g., inquiries to the referral line by region of the state) and stakeholder feedback (e.g., utilizing advertising materials in Spanish to target heavily Latine regions of the state).

Operational Meetings to Support Early Detection Efforts.

STEP-LC leadership and the regional EDAC have monthly meetings with LMHAs to review (1) workflow of referrals from STEP-LC to LMHAs,(2) Rapid Access to Service: troubleshoot any issues with delays to care and engagement in care, and (3) Collaborate with LMHAs on identifying and engaging in other early detection efforts, including outreach and detailing of key referral sources and stakeholder groups within their catchment area (e.g., local hospitals, schools, community organizations).

Meeting attendees from the LMHAs include key point people who are directly involved in the intake process (e.g., intake coordinators, mobile crisis team leaders, peers/recovery support specialists), designated early psychosis champion for the site, as well site leadership (e.g., CEO), and leaders of teams likely to receive individuals with FEP into care (e.g. Young Adult Services teams). Given the diverse makeup of organizations and processes across the LMHAs, exact attendees vary.

Deliverables:

Early Detection: The STEP-LC has successfully built upon a regionally successful early detection campaign (Srihari et al., 2022) and has launched a statewide Early Detection campaign (www.mindmapct.org). The Mindmap campaign aims to create pathways to care that are faster and less aversive. To date the following deliverables have been produced:

• Launched a centralized referral number and assessment hub to screen and engage referrals for FEP across Connecticut

• Curated referral pathways and workflows for each participating clinical agency (13 LMHAs) to quickly engage young people experiencing FEP into care (RAS)

• Creation of Mindmap Website (www.mindmapct.org) hosting various digital assets, including a downloadable “Digital Tool Kit”, educational blogs, and an archive of news related to the campaign efforts

• Public education campaign (PEC) efforts across the state via mass media

• Social media engagement with key stakeholders across platforms (Facebook, Instagram, LinkedIn, Youtube)

• Developed and ongoing dissemination of videos amplifying voices of those individuals and family members with lived experience of psychosis

• Utilization of asset maps to guide outreach and detailing (POD)

*See Appendix 2 for Sample of STEP-LC quarterly report

5.4. Stakeholder Engagement

As previously mentioned, the STEP-LC views stakeholder inclusion as a core element of a LHS, and this principle is embedded throughout all workstreams. Specifically, the STEP-LC will work to include clinical and non-clinical stakeholders (e.g. education, criminal justice, clergy, social services, consumer groups, youth organizations) to help advise, inform and implement the mission of EIS. It is important that the LHS be responsive to feedback from all stakeholders, particularly those it aims to directly serve – young people experiencing FEP and their families. Examples of consumers and family member inclusion in the design and rollout of STEP-LC include:

• 1.Feedback about difficulties along the pathway to care within the STEP-LC is taken seriously and addressed using a quality improvement framework, such that any issues which could be more widespread can be addressed and shared across the network.
• 2.Regular conversations with alumni and current parents as part an effort to develop language for a bill to reform access and care for FEP in Connecticut
• 3.Recording and interviewing of individuals and family members to create videos to amplify voices of lived experience and to assist with the Early Detection campaign. See examples here: https://mindmapct.org/lived-experience/
• 3.Conversations with state-based (NAMI) and national (Students with Psychosis) youth-led groups to explore ways advocates can provide feedback on and support the Early Detection and Workforce Development workstreams of the STEP-LC
• 4.Active efforts are underway to formalize an advisory board that will include representatives from all stakeholder groups including many who have been active participants in the first Mindmap campaign or have over the past almost two decades demonstrated commitment to efforts that are very consistent with the LC.

Asset Mapping.

The STEP-LC engaged with each LMHA around an “asset mapping” exercise, in which various LMHA members were supported in identifying key “assets” in their local catchment area. The Red Rock team created an informational video to facilitate the asset mapping process, which can be viewed here: https://bit.ly/3TrZ8Gx. LMHA team members were encouraged to identify various organizations (hospitals, social & cultural organizations, educational, neighborhood resources), individuals (government officials, police chiefs, community influencers), and physical spaces (libraries, popular bus stations) that may be useful in targeting with messaging, outreach, and/or detailing efforts. The ideas generated in each asset mapping exercise were then processed by the STEP-Hub team and used in collaboration with the LMHAs to guide Early Detection outreach efforts.

Deliverables:

Stakeholder Engagement: Stakeholder inclusion is infused within all other workstreams. The ultimate goal of this workstream is broad ownership and governance of the statewide system of care. Progress on activities and deliverables are detailed below:

• Created detailed asset maps across all regions of the state

• Utilized asset maps to guide ongoing Early Detection efforts, including active outreach and regular detailing

• Engaged with a wide range of outreach and educational requests across stakeholder groups, including advocacy and consumer organizations (e.g., NAMI, Students with Psychosis)

• Curated an expansive email list-serv of engaged stakeholders, of whom we can subsequently use for targeted messaging

• The STEP-LC is actively working to identify key members across stakeholder groups to be a part of a statewide FEP Advisory Board.

5.5. Sustainability

No model of care, however effective, can have meaningful impact without a plan to sustain itself. This domain includes not only fiscal health but also the ability to retain and avoid demoralization of the necessary workforce, and inclusion of a broad range of supporters within the legislative and executive branches of government to encourage ongoing investment in such services. The design of the STEP-LC intentionally includes continuous re-assessment of resource allocation to limit waste and maximize value. The STEP-LC team has a weekly meeting amongst the leadership and staff involved in all workflows to review structures, processes and preliminary outcomes that suggest opportunities to save staff time or cost for resources. Examples include:

• Informatics: moving to REDcap based surveys facing patients and family to reduce the time burden of collecting patient-level outcomes data from clinicians or medical record

• Informatics: developing conventions and stopping rules for when to discontinue patient follow-up in the LC after repeated attempts to reach them have failed (current protocol is to stop after 3 successive failures to reach patient or collateral over 3-monthly assessments or if there is clear evidence the patient has moved out of the state).

• Workforce development: archiving cuts of webinars on website and directing individuals and agencies to view these to limit the number of in-person or on-demand presentations

• Early Detection: coordinating efforts with mobile crisis teams when they are already engaged with a patient, to avoid replicating efforts to guide patient into care.

• Coordination with existing resources: identification and coordination with existing services that are funded across the state to avoid redundant resourcing and empower other agencies and stakeholders, such as crisis services and supported employment.

The issue of fiscal viability remains an ongoing target. It is well known that comprehensive EIS are impossible to mount within the current fee for service billing regimes and alternative payment models will be necessary. The recent approval of a CMS code for CSC is a positive step; state level negotiations will be necessary to align budgets and limit the financial risk to motivated healthcare providers who want to provide EIS and reduce overall societal costs, but for whom doing so may incur unacceptable costs to operational budgets.

Deliverables:

Sustainability: We will continue to work with all stakeholders to build a fiscally sustainable LHS with decreasing reliance on State/Federal grants that remains aligned with the overall public health goal. Progress and directions on sustainability deliverables are detailed below:

• The STEP-LC is engaged in ongoing efforts toward developing Alternative Payment Models (APMs) that would permit reimbursement or bundling of more services to provide necessary empirically-based services.

• In the meanwhile, concrete progress has been made in supporting the longer-term viability of the LC. Several LMHAs have received limited support from DMHAS to offset clinical staff time spent on work to coordinate with the Hub and the wider LC activities that are currently not reimbursable.

• The Connecticut legislature passed a bill that requires the Commissioner of DMHAS to present an annual report on the statewide LC. This offers a mechanism for public oversight of the work but also for sustaining and expanding the measurable impacts on population health.

5.6. Care Pathway Refinement

The STEP-Hub serves as an ‘incubator’ for new evidence-based treatment approaches. This includes both the implementation of novel empirically based treatments that require the acquisition of new knowledge and skills (‘knowledge to practice’) and the development of new interventions to address gaps in our current evidence base (‘practice to research’). The EIS will pilot interventions that emerge from these activities before deployment across the LHN and will lead grant applications for discovery-oriented research. This workflow reflects the commitment of the STEP-LC to the dual goals of LHS within an overarching aim of ‘knowledge translation’ wherein it should be possible to both do what we know works (best practice implementation) while also seeking to learn what to do better (discovery-oriented research) (Srihari & Cahill, 2020).

While the workforce development, informatics, and early detection workstreams are most mature, future efforts will continue to develop these areas while increasing focus on sustainability and care pathway refinement. Results related to population health-based patient outcomes will be available with continued implementation of the STEP-LC. A sample of the STEP-LC quarterly report is available in the Appendix (see Appendix 2. Sample of STEP-LC Quarterly Report).

Deliverables:

Care Pathway Refinement: Ongoing goal of continuously updated best practice care pathway ‘incubated’ at STEP-Hub for dissemination across STEP-LC. Progress on activities and deliverables is outlined below:

• STEP continues to use data collected during care to improve outcomes via Quality Improvement (QI) methodologies. Ongoing projects are:

  • The development of processes to ensure better monitoring of two key cardiovascular risk factors: weight gain and smoking and the responses with evidence-based interventions (including better ways to route eligible patients to new treatments e.g., GLP-1 agonists)
  • Development of processes to apply best available approaches for Cannabis use disorders, including integration of motivational interviewing, risk reduction
• STEP is currently engaged in collaborative research that spans the translational continuum (funded by extramural research grants) and that if successful will be easier to disseminate across the network. These include:

  • Basic efforts to understand impairments in social interaction using novel functional near infrared spectroscopy (fNIRS)
  • Clinical trials of a novel supplement to target cognitive deficits

6. Discussion

6.1. Lessons Learned

Expect Variation in Clinical Operations Among Sites in the Network:

As an example, it became apparent that agencies in the STEP-LC had different definitions of what constitutes “admission” to their clinic, necessitating the STEP-LC to be transparent about the Access Outcomes that would be measured. To do so, the STEP-LC defined how it constitutes “admission” (when an individual has access to the full range of clinical services offered by the agency) and then asked each agency to define at which point an individual would meet those criteria. Instead of attempting to dictate admissions procedures, which is beyond the purview of the STEP-LC, we continue to encourage rapid engagement within existing frameworks, working with the LMHAs to troubleshoot barriers and periodically reviewing access outcomes to ensure efficiency and equity. The STEP-LC aims to create a culture in which the LMHAs are eager to engage in quality improvement and ask about existing processes in order to meet international benchmarks (e.g., admission within 7 days). Other common factors contributing to delay include lack of staff availability, lengthy intake assessments, and limited access to youth-friendly communication (e.g., texting or other e-communication).

Expect the Need for Re-orientation:

Despite invested leadership and local FEP champions, it should be expected that it will take time for staff to adopt a new initiative, especially with high rates of staff turnover and if the initiative involves a change to longstanding workflow (e.g., accepting individuals of all insurance statuses). The hub team should proactively work with each site and/or designated FEP champion to help keep staff oriented to and invested in FEP related efforts.

Sharing Lessons Across Sites:

Each site in the network will have existing strengths as well as successful adaptations to address issues as they arise. These lessons are important to learn from and share across the network. For example, some sites may have a particularly strong relationship with local mobile crisis response teams, enabling rapid identification and admission of eligible individuals . Others may have particularly strong ties with local universities and schools, or ways to engage individuals and families into care, such as having a designated peer support specialist as the first point of contact for this population. Such exemplary efforts should be highlighted in regional meetings, shared in newsletters, and ultimately adopted across the network if contributing to improved population health outcomes and applicable for the needs of that region.

Transparency about LHS Process Expectations:

It is important to be clear with agencies in your network about what is expected of them, however, inherent in the learning healthcare system ethos is that the system is responsive as new data and needs arise. To avoid misperception about changing expectations, it is important to clarify to agencies upfront that while some processes may evolve based on ongoing assessment and feedback, the core set of valued population health outcomes will remain constant across the network.

Data Collection:

Collecting data from clinical sites can be challenging and its quality variable of varying quality. To improve adherence, it is best to utilize existing data collection entities whenever possible such as state reporting mechanisms, publicly available data, or data collected directly by the Hub team.

7. Conclusion

In summary, the STEP Learning Collaborative (STEP-LC) provides a model for development and implementation of a statewide learning health system for recent-onset schizophrenia spectrum disorders. The STEP-LC has demonstrated initial success in creating a robust system of care for early psychosis across Connecticut. The STEP-LC has implemented vast workstreams to address gaps in information and workforce readiness. By leveraging informatics and continuous quality improvement workflows, STEP-LC aims to meet international benchmarks for access and outcomes of care. STEP has also helped stand up new services in other U.S. states, provided informatics and learning health system support to two national networks and collaborated with other state and international initiatives. Ongoing efforts will further refine care models, sustain best practices, and integrate with state services through policy advocacy and coordination and expand upon the STEP-LC’s public-academic ‘backbone,’ to include private healthcare providers.

Supplementary Material

1

Appendix Table 1. Measures and objectives for STEP-LC baseline and follow-up assessment batteries

Glossary:

STEP-LC

STEP Learning Collaborative

STEP

Specialized Treatment Early in Psychosis Program

LHS

Learning Healthcare System

FEP

First-Episode Psychosis

DUP

Duration of Untreated Psychosis

PTC

Pathway to Care

PEC

Public Education Campaign

POD

Professional Outreach and Detailing

RAS

Rapid Access to Service