Families of children with autism often face significant challenges following diagnosis, including difficulty accessing reliable information, navigating complex systems of care, identifying appropriate services, and managing the emotional demands associated with caregiving. These challenges may be particularly pronounced in rural, underserved, and resource-constrained settings where access to specialty autism services and family navigation support is limited.
This study will evaluate an AI-enabled autism caregiver support tool designed to provide evidence-based autism information, caregiver emotional support, and service-navigation guidance for caregivers of children with autism in Indiana, United States, and western Kenya. The intervention was developed through a reciprocal innovation approach involving partners in both settings and was informed by caregiver, clinician, educator, and community stakeholder input. The tool is designed to provide plain-language information about autism, answer frequently asked caregiver questions, help caregivers identify relevant services and resources, and provide supportive coping guidance. Safety guardrails are incorporated to address crisis, medical, diagnostic, treatment-related, and other high-risk questions. The tool is not intended to diagnose autism, replace clinical care, or provide emergency services.
The study will use a prospective single-arm pre-post pilot design. Approximately 60 caregivers of children with autism will be recruited in Indiana and Kenya and provided access to the AI-enabled caregiver support tool. Participants will complete baseline and follow-up assessments and will have access to the intervention throughout the study period. Usage data, including engagement with the tool, use of resource-navigation features, and safety-related interactions, will also be collected.
The primary focus of the study is to evaluate feasibility, acceptability, safety, engagement, and clinical-trial readiness. Specific outcomes will include recruitment, retention, assessment completion, intervention uptake, participant engagement, safety events and escalations, usability, and implementation outcomes such as acceptability, feasibility, and appropriateness.
Additional exploratory outcomes will include autism knowledge, caregiver self-efficacy, access to reliable information, emotional well-being, service navigation, unmet caregiver needs, service engagement, preparedness to take next steps following diagnosis, satisfaction with the intervention, and trust in the AI-enabled caregiver support tool. These outcomes will be used to estimate outcome variability and inform selection of measures for a future fully powered effectiveness-implementation study.
By evaluating an AI-enabled caregiver support intervention in both Indiana and Kenya, this study will generate important information regarding implementation, safety, usability, engagement, and caregiver support needs across diverse cultural and resource settings. Findings will inform future efforts to develop scalable approaches for improving access to autism information, caregiver support, and service navigation for families of children with autism.
