Coordinated specialty care (CSC) is the gold standard of care for the treatment of individuals at clinical high risk for psychosis (CHR-p) and with first-episode psychosis (FEP) and has demonstrated efficacy in reducing hospitalization visits and improving symptoms and functioning. CSC teams generally include (1) team leadership, (2) case management, (3) supported employment and education, (4) psychotherapy, (5) family education and support, and (6) pharmacotherapy and primary care coordination [1]. However, this also makes the CSC program staff intensive and limits their scalability. Recent research has suggested that CSC care delivery may be supported through digital mental health interventions (DMHIs) in FEP care [2] and CHR-p care [3]; however, its feasibility and effectiveness are yet to be demonstrated.

Patients with FEP have high rates of technology ownership and use as evidenced by a 2022 study of an FEP clinic that found that patients’ access to smartphones was 100% and that 66% of the patients used social media [4]. A variety of new technologies have been studied in CSC including apps, web-based videos, games, and social media platforms [5]. Although these studies had smaller sample sizes, they show high levels of interest from patients. However, more formal studies of technology and apps for CSC have shown more disappointing results likely due to low engagement [6,7] that may be related to implementation challenges such as acceptability and usability for participants [8].

However, low engagement and uptake need not be barriers. These are common challenges across all digital health platforms, meaning that there are common solutions. Proposed solutions beyond patient-centered design include ensuring that technology use is customized to patient needs and that clinicians have the support they need to implement and troubleshoot the technology. An ongoing study in an FEP program in Spain [9] demonstrated that when apps are integrated into care and are used in conjunction with a clinician, engagement can be high. Integrating apps in this manner requires consideration of CSC-specific challenges and thinking beyond just the technology. To address implementation challenges, implementation science offers guidelines to support (1) the identification of barriers and facilitators across different implementation targets and (2) the development of an implementation strategy to address barriers and enhance facilitators [10].

Although there are several implementation frameworks to guide efforts, one practical model to apply is the Access, Alignment, Connection, Care, and Scalability (AACCS) [11] framework, which uses digital navigators to measure and intervene upon specific implementation facilitators and barriers of technology in a CSC setting.

Given that recent meta-analyses show promise that digital technologies can effectively enhance care in psychosis treatment [12], it is of significance to explore the effectiveness of digital technologies on symptomatic and functional outcomes in concert with implementation outcomes. Assessing these outcomes in concert may shed light on the necessary next steps needed to bring the benefit of digital technologies to the clinical setting.

This quality improvement project aims to develop and use a novel protocol, Implementing Technologies to Enhance Coordinated Specialty Care (iTECSC), to assess the implementation outcomes for digital technologies in CSC. iTECSC is a hybrid type-2 [13] protocol intended to analyze implementation outcomes and the effectiveness of digitally supported treatment in a CSC setting. Guided by implementation science, the outlined protocol was informed by the AACCS framework and supported by a small quality improvement study at 2 local CSC clinics that attempted to identify, measure, and intervene upon implementation barriers and facilitators.

Having developed the blueprint, we conducted a small quality improvement project at 2 local CSC clinics to gather feedback that would further inform the design. Two local clinics were included in the project so that we could explore different DMHI configurations ranging from hyperpersonalized to more standardized configurations. Given the quality improvement design of the study, this project did not analyze symptomatic or functional outcomes. Future iterations of iTECSC would benefit from integrating both implementation outcomes and effectiveness outcomes.

The participants were patients and clinicians at CSC Boston–based clinics for CHR-p care (Center for Early Detection Assessment and Response to Risk [CEDAR]; [17]) and FEP care (Advancing Services for Psychosis Integration and Recovery [ASPIRE]). Patients at CHR-p met broadly defined CHR-p criteria [18], and patients with FEP had at least 1 documented incident of a first episode of psychosis. Clinicians included psychiatrists, a psychologist, a social worker, and a supported education and employment specialist.

At CEDAR, a patient-needs–identifying approach was used when implementing the DMHI into treatment. Patients selected from a “menu” of activities that the DMHI could support including a thought record, mindfulness activities, positive affirmations, relaxation techniques, and psychoeducation materials. Their app was tailored to include their specific “prescription” of activities.

At ASPIRE, a modular approach was presented to patients and clinicians, in which all patients received a preprogrammed app that included surveys measuring mood, sleep, sociality, psychosis, anxiety, and medication that were delivered daily in the morning and evening and as interventions targeting mindfulness, psychoeducational tips, and cognitive restructuring activities.

The review and analysis of data collected through this quality improvement project was approved by Beth Israel Deaconess Medical Center Institutional Review Board (#2023P000231). Participants were not compensated for participation in this project as it was originally implemented as an augmentation to current care.

Demographic information was analyzed to examine sample diversity. Descriptive statistics, including means and frequencies, was used to examine the baseline and 1-month survey data to assess implementation outcomes. Survey data were qualitatively reviewed to assess for feedback that may provide insights into improving future implementation of projects. Descriptive statistics from the mindLAMP app and clinical data was also analyzed to examine which specific parts of the app had the highest participation to inform future development. One provider from each clinic provided case examples that were analyzed to demonstrate app use over the 1-month period.

The initial presentation included 4 clinicians from the ASPIRE clinic and 5 clinicians from the CEDAR clinic. Two clinicians from the ASPIRE clinic—JT (MD) and MK (MD)—and 3 clinicians from the CEDAR clinic—Caroline Howland (licensed independent clinical social worker), Amanda Weber (PhD), and JG (BA; supported education and employment specialist)—expressed interest in the DMHI opportunity. Moreover, 1 clinician from ASPIRE (MK) and 3 clinicians from CEDAR (Caroline Howland, Amanda Weber, and JG) completed the baseline assessment battery. Furthermore, 2 ASPIRE clinicians (JT and MK) and 2 CEDAR clinicians (Caroline Howland and JG) worked with patients with the DMHI, and 1 ASPIRE clinician (MK) and 1 CEDAR clinician (JG) completed the follow-up assessment battery. A case study was provided from 1 clinician at each site to describe their clinical use of the DMHI (JT and JG).

The project included 5 participant dyads (CEDAR: n=2 and ASPIRE: n=3). Patients were 40% (2/5) Black or African American and 60% (3/5) White; 40% (2/5) of the participants identified as cis-male, 40% (2/5) of the participants identified as cis-female, and 20% (1/5) of the participants identified as transmasculine; the participants had a mean age of 19.6 years. Moreover, 20% (1/5) of the participants were in high school, 60% (3/5) of the participants attended university, and 60% (3/5) of the participants were employed. Furthermore, 3 patients completed the baseline and follow-up assessment batteries.

The AACCS quantitative survey data (Table 1) demonstrated that clinicians and patients at both clinics on average agreed that they had high levels of access to technology, felt connected to the technology, and believed that technology could be sustained over time, as responses to the AACCS survey for all categories ranged from 5 (somewhat agree) to 6 (agree).

In assessing how technology could align with care offered by the clinics, clinicians rated supporting psychotherapy as the highest, but overall responses were diverse, as clinicians reported that the DMHI could support unique therapeutic interventions addressing specific treatment needs and all aspects of FEP or CHR-p care. This is best illustrated in the cases presented in Textbox 1.

Implementation outcomes data suggested that patients and clinicians at both clinics had moderately high levels of acceptability, reflected by the moderately high mean scores from the EBPAS ranging from 2 (moderate extent) to 3 (great extent) for adherence to requirements, appeal, openness, and low levels of divergence to new interventions (Table 1). Regarding adoption and penetration, 40% (2/5) of the patients and 40% (2/5) of the clinicians at CEDAR participated in the intervention, whereas 100% (3/3) of the patients and 50% (2/4) of the clinicians at ASPIRE participated in the intervention. The total level of adoption and penetration at both clinics for both patients and clinicians was 53% (9/17), suggesting moderate levels of adoption and penetration, slightly favoring the modular version of the app for patients that was presented to ASPIRE. The implementation pathways from the initial presentation to adoption and penetration outcomes are shown in Figures 6 and 7.

The Counselor Assessments of Training and Adoption Barriers survey responses that were collected during the 1-month follow-up suggested that on average clinicians and patients rated the app with some (3) appropriateness to a lot (4) of appropriateness. Post–1-month follow-up survey responses demonstrated that all 3 patients preferred different aspects of the app (Table 2). These results also suggested that patients had varying difficulties and challenges with the app, ranging from being unsure of what to do outside the scheduled activities, using the technical aspects of the app, and reminding themselves to use the app. These data demonstrate moderate feasibility, given the wide range of needs and challenges while using the DMHI.

Deidentified case examples are provided from 1 provider from each clinic to describe their experiences with mindLAMP (Textbox 2). These case examples demonstrate how DMHIs can be deployed to address the unique needs of each patient.

The results of this quality improvement study shed light on the current barriers and possible facilitators for the implementation of DMHIs in CSC. Primarily, this quality improvement project demonstrated that clinicians and patients reported high levels of access to DMHIs and believed in the sustainability of DMHIs. In addition, the results suggest that clinicians and patients express an interest in a diverse range of use cases for DMHI-augmented care. Furthermore, patients and clinicians self-reported moderate to high levels of acceptability on the AACCS survey, demonstrating initial acceptability and interest in DMHIs generally. When implemented into care, adoption and penetration proportions ranged from 40% to 100% in each setting, with an average of 53%, suggesting moderate and variable adoption and penetration across different use cases. At 1-month follow-up, clinicians and patients reported some to a lot of appropriateness of DMHIs in CSC care after using the intervention, suggesting that the DMHI was moderately appropriate for use in these care settings. Post–1-month follow-up survey results further suggest variable feasibility of the DMHI, given the diversity of use cases and challenges that arose while using the DMHI. Finally, the case examples demonstrated how and when DMHIs can be used and designed uniquely so that they can address the specific needs of each patient to inform care. This suggests that flexibility and versatility are important when considering the use of DMHIs in care. These results indicate that barriers to implementation are exemplified during the adoption and penetration phase of implementation, which may be associated with the moderate feasibility and appropriateness of the DMHI in its presented form. To improve implementation outcomes of DMHIs in CSC, barriers and facilitators to adoption and penetration, feasibility, and appropriateness must be investigated and targeted.

When reviewing these findings, it is evident that both clinicians and patients can access and believe in the sustainability of DMHIs, demonstrating their readiness to enhance care with DMHIs, which is reflective of mainstream findings in the literature [3].

While considering what may have facilitated higher levels of adoption and feasibility, this project suggested that the versality and multicomponent design, rather than the complexity, of DMHIs is important as evidenced by the diversity of responses to how clinicians and patients imagined the DMHI could be used to align with treatment goals. This is reflective of similar findings, which found that approaching DMHI development from a versatile rather than a complex perspective may be preferable [21]. Results from the case examples and follow-up survey responses also suggest that DMHI autonomy is important to allow the DMHI to appropriately align and augment the unique clinical styles of clinicians and the needs of patients, which has also been evident in other recent findings [22].

Congruent with other trends in the literature [23], patient feedback and higher rates of fidelity and feasibility in the modular version of the DMHI suggest that a structured approach may maximize acceptability, adoption, appropriateness, fidelity, and feasibility. In addition, findings from a recent qualitative study gathering clinician feedback on another DMHI suggested that adaptable DMHIs may work to improve implementation outcomes such as feasibility, appropriateness, adoption, and penetration [24]. Prior research using smartphone apps in the care of patients with early psychosis has also shown feasibility [9,25], which is similar to our results; however, the use of digital navigators in this study is novel. The implementation methodology described in this project heavily relied upon digital navigators, thus reinforcing the essential role of a digital navigator in the initial and sustained implementation of DMHIs in clinical settings [15].

Given the findings from the quality improvement project as well as leading trends in mainstream literature, we suggest that developing DMHIs that are versatile rather than complex in their design, promote patient and clinician autonomy to meet unique clinical needs, and provide patients with a structure to complete modules and activities may improve the evident implementation pitfalls of DMHIs exemplified in the adoption, penetration, appropriateness, and feasibility outcomes.

Although this quality improvement project was limited by its protocol design, sample size, and absence of symptomatic or functional outcome data, a recent study implementing a DMHI into an FEP program showed promise in reducing relapses, hospitalization, and visits to urgent care units [25]. Our results suggest that the greatest barriers to adoption and penetration may be associated with feasibility and appropriateness. Developing DMHIs that are adaptable, versatile, structured, and patient- and clinician-centered may improve feasibility and appropriateness, thus improving adoption and penetration outcomes, which may in turn allow for patients to access the identified benefits that technology-supported CSC has to offer.

This quality improvement project was limited by many factors, as it was conducted as natural quality improvement project in real-world clinical settings. Our primary limitation included a small sample size, given our goal to establish feasibility. In addition, the authors recognize that some of the findings may be biased, given that the authors of the protocol design have participated in the project as clinicians; however, many of the suggestions that had been leveraged to improve the protocol came from patients and clinicians who were not directly involved in the development of the protocol. In addition, this study did not assess symptomatic or functional outcomes. Future iterations of iTECSC may benefit from assessing implementation outcomes and effectiveness outcomes as primary co-outcomes.

In summary, the feasibility findings from this project have informed the development of the iTECSC protocol by suggesting the need for a patient-centered approach that is guided by digital navigators and provides versatility, autonomy, and structure. Such a new study protocol has the potential to build on growing insights regarding the need for versatility, autonomy, digital navigator support, and structured applications [26-28]. We anticipate that by continuing to research and improve implementation barriers impeding the adoption and penetration of DMHIs in CSC, accessibility and uptake of DMHIs in CSC will improve, thus connecting patients to the demonstrated benefits of technology-augmented care [25].