Although tele–mental health and parent coaching services are increasingly available, it remains unclear which types of support programs are best suited to various populations and how to best deliver these support programs. One population particularly well suited to telehealth support programs is caregivers of children with neurogenetic conditions (NGCs). Here, we defined NGCs as genetic conditions that affect cognitive, psychosocial, and physical development; examples include named conditions, such as Angelman syndrome, Prader-Willi syndrome, and Williams syndrome, as well as rare and ultrarare conditions with lower prevalence rates. These NGC caregivers are vulnerable to elevated stress and mental health challenges due to often extreme caregiving burden; for example, they may experience geographic isolation from specialty support clinics, exhibit elevated levels of parenting stress, and manage complex child behavioral and medical needs with limited structural resources [1-3]. Project WellCAST (Supporting Well-Being of Caregivers via Telehealth) is a National Institutes of Health–funded clinical trial that aims to prospectively develop and test the efficacy of a precision health decision tree algorithm for routing NGC caregivers to evidence-based telehealth supports, based on their unique clinical and lifestyle profiles. Project WellCAST also tests the efficacy of community-engaged peer-to-peer coaching protocol in enhancing the feasibility and efficacy of digital supports. Here, we provide the rationale and protocol for Project WellCAST, including anticipated outcomes and impact within and beyond NGC caregiving communities.

Although telehealth support programs have existed for some time, telehealth services expanded rapidly in response to the COVID-19 pandemic [4], providing new opportunities for increased access to mental health programs [5,6]. One population for whom telehealth services offer critical benefits are NGC caregivers. Although the elevated mental health needs of NGC caregivers predated the COVID-19 pandemic [7], the pandemic caused them to face crisis-level breakdowns in support and intervention services that have continued to impact families [8]. These COVID-19 pandemic–related impacts were, in part, driven by social isolation, financial strain, and reductions in face-to-face support programs, which compounded an already fragmented and challenging landscape of care [2,3]. The COVID-19 pandemic–related interruptions required caregivers to take on even more responsibility as they managed urgent educational, vocational, and functional needs of their children [3]. The negative impact on caregivers’ health and well-being continues to this day [8]. However, despite their elevated need, caregivers rarely receive support for their own mental health. Any psychological support they do receive is typically disconnected from their child’s care plan and delivered by practitioners with little-to-no expertise in the needs of NGC families [9,10]. Consequently, accessible and scalable digital interventions are needed to support the intersectional needs of NGC caregivers and prevent an accelerated breakdown in their mental health, particularly in the wake of the COVID-19 pandemic.

In part due to the rapid and dramatic uptake in telehealth methods during the COVID-19 pandemic, many questions remain regarding the circumstances and populations for which telehealth services are most effective [6], including but not limited to services for NGC caregivers. Although a variety of evidence-based telehealth support programs are available, it is not always clear which programs should be provided to which patients. This issue has long motivated the psychiatry and psychology literature, though many randomized controlled trials continue to examine broad treatment effectiveness [11,12] rather than investigating treatment personalization protocols [13]. Precision health literature guides our efforts to move toward support program personalization for NCG caregivers, with promising examples of prognostic models in other populations [14].

One challenge to the effective prospective assignment of patients to treatments is the issue of measurement. Often, mental health providers base selection on limited intake information and through subjective assessment guided by broad symptom checklists or patient-reported symptoms. As an expansion of this approach, precision mental health care aims to optimally match patients to existing evidence-based treatments through the use of demographic, clinical, and dispositional information [14]. Tools such as daily diaries and ecological momentary assessments (EMAs) hold promise for providing temporally dense information about participants’ experiences, beyond the contributions of clinical assessments alone. EMAs are brief questionnaires aimed at gathering real-time data on the participants’ experiences that reduce the challenges of retrospection [15]. Using EMA data can provide insights into fluctuating daily experiences and momentary affects, as well as information about practical day-to-day constraints within their lives that may affect support program uptake (eg, caregiving burden, health behaviors, and social support). However, to date, tools such as EMA have not been prospectively used to match patients to support programs based on their personal health profiles.

Another consideration for the effective assignment of support programs is the impact of sociocultural and practical issues on a patient’s willingness and ability to engage in tele–mental health treatment. For example, several studies have documented that telehealth does not universally improve accessibility and may, in some cases, enhance preexisting disparities by disrupting trust and rapport, distancing communities with poor technology literacy or access, and introducing concerns about data security and privacy [16]. These health care gaps are amplified among Black and other minority families who are less likely to receive mental health care due to discrimination, racism, and historic injustices against Black people in health care settings [17].

One potential solution to the disrupted rapport in telehealth-based programs is to implement community-centered and motivational support programs, such as participation enhancement interventions aimed at addressing barriers to support program participation and enhance motivation and adherence to treatments [18,19]. These support programs may be particularly beneficial for caregivers of children with NGCs due to isolation and disconnection from social support that may accompany their child’s complex needs [9,10]. Peer support is known to provide benefits to caregivers [20,21], though to be most useful, perceived need and received support must align [22]. There is also some evidence that peer support may have differential impacts on certain subgroups of parent caregivers (eg, individuals with higher baseline anxiety [23]), though more evidence is needed to determine which participants might benefit most from peer support. Thus, peer support interventions may positively impact willingness and engagement in tele–mental health programs for some caregivers while simultaneously being an extra unneeded hassle for others. Although these programs have demonstrated broad efficacy across several neurodevelopmental populations, they have not been explicitly tested in substantial NGC caregiving samples. Given that coaching programs require extra time and resource commitments from both families and professionals, and many participants effectively engage in intervention without peer support, it is important to determine which participants would benefit from this extra layer of support.

Project WellCAST (NCT05999448) is an early-phase clinical trial designed to prospectively test which evidence-based telehealth interventions best meet the unique needs of NGC caregivers. We accomplish two specific aims: (1) develop and optimize a personalized health decision tree algorithm to match NGC caregivers to digital health programs using a combination of standard clinical tools and smartphone-based EMA data, and (2) test the efficacy of peer-to-peer coaching, deployed by trained NGC caregivers using an evidence-based motivational interviewing protocol, in enhancing support program uptake and clinical outcomes. Here, we describe the protocol methods and anticipated impact of this project, both for NGC caregivers and the broader field of precision mental health.

Project WellCAST uses a randomized control design with waitlist control to test the effectiveness of both (1) a precision health algorithm and (2) a participation enhancement intervention on improving clinical outcomes for NGC caregivers. It is anticipated that between 500 and 750 caregivers will enroll in WellCAST across waves, with a minimum enrollment of 500 participants, including 50 waitlist controls, to adequately power primary analyses. SPIRIT (Standard Protocol Items: Recommendations for Interventional Trials) reporting guidelines were used in the preparation of the Project WellCAST protocol [24].

Project WellCAST is deployed across 4 waves. During waves 1 and 2, participants are randomly assigned to be routed to support programs using either a preregistered precision health algorithm (algorithm 1) or quasi-random assignment. During waves 3 and 4, participants are assigned to support programs using optimized algorithms (algorithms 2 and 3, respectively). Across waves 1 to 4, participants are randomized to either a participation enhancement intervention delivered by an NGC caregiver coach (peer coaching) or similarly timed check-ins by non-NGC caregiver staff (staff coaching).

At the time of the initial protocol submission, Project WellCAST has completed waves 1 and 2 (fall 2023 and spring 2024); waves 3 and 4 will be conducted in fall 2024 and summer 2025, respectively. Thus, across waves, 25% of participants are assigned to each algorithm category (quasi-random, algorithm 1, algorithm 2, and algorithm 3), and 50% of participants are assigned to each peer coaching category (peer coaching and staff coaching).

Project WellCAST was registered on ClinicalTrials.gov on August 21, 2023, and the first algorithm was registered on Open Science Framework (OSF) Registries on August 30, 2023, before running the first algorithm assignment. The first participant consented to a support program on August 31, 2023, and the first participant started the support program on September 11, 2023. Supplemental analytic plans were preregistered on OSF Registries on October 18, 2023, to further detail the ClinicalTrials.gov registration. Wave 1 participants consented to baseline data collection starting June 28, 2023, and started baseline data collection on August 7, 2023. Baseline data were collected before any participant consented to support programs and were not analyzed or used to route participants until after registration was complete.

Caregivers of children with NGCs are recruited through various sources to maximize the potential participant pool. Recruitment strategies include social media advertisements, sharing materials with genetic clinics and providers, mailers by patient advocacy groups, emails to condition-specific email distribution lists, and recruitment through registries. Inclusion criteria are (1) legal guardian of an individual aged 2 to 35 years with a documented NGC associated with intellectual or developmental disability that is not primarily characterized by neurodegeneration, (2) fluent in English, (3) interacts with the individual with an NGC most of the time, and (4) resides in the United States. Participants are excluded if they either (1) express severe psychiatric illness not indicated for the interventions offered (eg, suicidal, psychotic thoughts, mania, or significant substance use) or (2) are currently in active therapy or receiving support that would conflict with, or be redundant with, the interventions offered.

Interested participants may scan a QR code on recruitment materials or receive a link via email to a digital prescreening survey via REDCap [25,26]. Once complete, research staff members who are trained in clinical assessments contact interested caregivers by phone to conduct a screening interview and to determine eligibility. The research staff member verifies demographic information and assesses the participant’s recent mental health history using the Mini-International Neuropsychiatric Interview [27]. If the participant is determined eligible to participate, the research staff member provides instructions on how to sign electronic consent forms for the baseline phase of the project in REDCap and begins the baseline phase of the study. Participants separately provide consent to their specific support program after being assigned a program arm during the routing phase.

Participants are matched to a combination of digital health support programs using a multistep process indicated by the data provided in the baseline phase. First, research staff not involved in participant support programs generate an eligibility panel for each participant that contains feasible and appropriate programs based on broad rule out criteria. For example, participants who do not identify with minority groups will not be eligible for programs focused on race-related trauma. Participants with children aged >8 years are not eligible for certain child-focused programs designed for young children. The preregistered algorithm is then used to generate a sequential list of participants’ recommended programs, based on their baseline data.

During waves 1 and 2, the first randomization is to algorithm 1 or quasi-random assignment. Participants randomized to algorithm 1 are assigned their algorithm-indicated support program. If practical issues prevent them from engaging with their top program (eg, lack of availability during group times), they are offered the next indicated support program. Participants randomized to quasi-random assignment are assigned to a support program using a random number sequence generated by a member of the biostatistics team who is not involved in participant management or intervention deployment. Similarly, if practical constraints prevent their first program from being feasible, they are moved to their next program indicated by the random number sequence. Whether the participant received their top assignment is recorded and integrated into statistical analyses. Finally, all participants are randomized to either peer coaching or staff coaching protocols using a random number generator. During waves 3 and 4, participants continue to be randomized to peer coaching or staff coaching. All participants receive algorithm 2 during wave 3 and algorithm 3 during wave 4. Care providers are masked to the status of peer coaching and algorithm assignment. Peer coaches are masked to the status of algorithm assignment. Data managers and statisticians evaluating study outcomes are provided deidentified data and are not involved in support program–related study operations. Research team members evaluating study outcomes are also masked to assignments.

A subset of participants is also assigned to a waitlist control to allow for the evaluation of treatment efficacy and quantify potential attenuation of responses with repeated assessment [30]. Participants assigned to waitlist controls contribute identical data to active participants and receive the same remuneration. They are routed to support programs or resources during the next wave.

Precision health methods have the potential to improve the feasibility and efficacy of tele–mental health programs for high-need populations, including and beyond NGC caregivers. Project WellCAST tests whether a precision health decision tree algorithm can be developed and used to improve clinical outcomes among NGC caregivers engaging in brief, 12-week, evidence-based support programs. We plan to disseminate trial results through multiple avenues, including publication in peer-reviewed journals, conference presentations, community-based webinars, and other methods as appropriate. It is anticipated that our results will improve understanding of the personal and clinical factors that may predict suboptimal outcomes. If successful, we anticipate Project WellCAST will provide a model for rapidly virtually assessing and triaging populations in need of tele–mental health care, including during future public health emergencies, such as the COVID-19 pandemic.

Several innovations uniquely situate Project WellCAST to contribute to the precision health and psychiatry literatures. First, we use digital health approaches that leverage smartphone-based data to monitor daily fluctuations in symptoms—rather than one-time clinical intake self-report data—providing a more detailed topography of initial patient needs and more sensitive metrics of acute, real-world change over time.

We also leverage community-academic partnerships that provide a scalable, cost-effective model for deploying community-centered interventions while simultaneously training the next generation of mental health providers to deploy digital evidence-based interventions with high-risk populations. To do this, we have aligned multiple digital support programs to a 12-week simultaneous delivery framework, providing a foundation for integrative support programs that can be individualized to families yet delivered with cohesion by trainees.

An additional innovation is that we incorporate extra support elements that we anticipate will boost clinical outcomes for participants who may face further barriers to treatment. For example, we include program options for caregivers experiencing comorbid racial trauma, led by highly qualified experts in the needs of minority families. We leverage community-based partnerships with peer support coaches to deploy evidence-based motivational interviewing protocols to boost program completion and efficacy, while simultaneously maximizing the involvement of patient communities.

These innovations are timely and necessary for NGC caregivers during and after the COVID-19 pandemic; however, this project also has the potential to inform improvements to the long-term landscape for how digital health programs are assigned and deployed far beyond this specific population or public health crisis. Indeed, the high baseline stress and scheduling constraints experienced by NGC caregivers make them an ideal user development group for innovating support program networks that are feasible and acceptable to variety of future users, including those experiencing trauma similar to what was experienced during the pandemic (eg, first responders, teachers, and frontline nurses) or full-time caregivers for other patient groups (eg, patients with dementias). In other words, our model, if successful, can be rapidly adapted and scaled to meet the needs of countless special populations as part of both general standard care and crisis response efforts.

Our study may also inform more cost-effective and sustainable clinical trial practices for several reasons. First, the design of the trial is built on community-academic partnerships, where support programs are deployed by graduate student clinicians with appropriate oversight from licensed supervisors. By engaging student-trainees to deploy support programs, we have minimized the overall cost of deploying programs while simultaneously providing high-quality training opportunities, including necessary clinical hours for licensure eligibility, for the next generation of clinicians. More broadly, our study aims to inform how to minimize mismatches between patient needs and provider offerings, which can reduce unnecessary costs associated with ineffective programming and patient attrition. Appropriately matching patients to support programs can save time, money, and frustrations for both the mental health provider and the patient, supporting sustainability and efficiency.

Project WellCAST was also designed to be intentionally noninnovative in 1 important way—the support programs being tested are already evidence-based and widely available to the general population. In other words, it is possible that Project WellCAST’s 2-week baseline data collection procedure and associated precision health algorithm can improve the field’s understanding of how best to route patients to already-available programs in their communities, in line with current directions in precision mental health [14]. For example, our efforts to identify “suboptimal responders” may inform which practical lifestyle factors indicate or counterindicate telehealth-based support programs, which patients may benefit most from additional coaching support, and whether a given patient should be offered mental health support versus parenting coaching. Additional studies on the generalizability of our final decision tree algorithm will be an important interim step toward this goal.

It is important to highlight that we expect Project WellCAST to be a strong match for some participants and a potentially poor match for others. Although telehealth has rapidly expanded in the wake of the COVID-19 pandemic, we and others have continued to caution that telehealth alone will not address the preexisting demographic health disparities that continue to persist across a variety of fields [16]. Our hope is that Project WellCAST can generate the data and results necessary to determine the contexts in which telehealth is most useful [6]. These contributions could be highly clinically relevant for other caregiver groups (eg, other special needs, severe mental illness, and older adults), including those experiencing race-related stress or trauma.

Project WellCAST tests whether a digital personalized health decision tree algorithm and peer coaching protocol can prospectively enhance telehealth program outcomes among NGC caregivers. This project is relevant to the specific population of NGC caregivers and may also inform how brief digital assessments, precision health tools, and community-academic partnerships can enhance public health response to mental health crises across other high-need populations, during and beyond the COVID-19 pandemic.