Mental Health

Evaluating an AI-Enabled Mobile Mental Health Monitoring Tool Among Family Caregivers of Adults Living With Cancer: Single-Arm Feasibility and Acceptability Trial Protocol.

TL;DR

This protocol describes a single-arm feasibility and acceptability trial evaluating the eCARE speech-based AI-enabled mobile app for monitoring depression and anxiety symptoms among family caregivers of adults living with cancer over an 8-week period.

Key Findings

Psychological distress, particularly depression and anxiety, is highly prevalent among family caregivers of individuals living with cancer, frequently equaling or exceeding rates observed in patients themselves.

  • Caregivers often assume central roles in care coordination, treatment adherence, symptom monitoring, and emotional support.
  • Despite increased attention to caregivers' mental health needs, routine distress screening remains limited in oncology care settings.
  • The paper identifies this gap as a primary motivation for developing and testing mobile health monitoring tools for caregivers.

The eCARE app (Ellipsis Caregiver Assessment Enhancement) is a speech-based, AI-enabled mobile application designed to screen and monitor symptoms of depression and anxiety in caregivers.

  • The app collects brief voice recordings and in-app survey data.
  • It is described as offering 'a scalable approach for integrating caregiver distress monitoring into cancer care.'
  • The tool was developed by Ellipsis Health, Inc.
  • eCARE is positioned as a low-burden, caregiver-centered approach to expand access to psychosocial support.

The study is a single-arm feasibility and acceptability trial with two phases targeting family caregivers of cancer patients.

  • Phase 1 involves 60 US-based family caregivers who are primary caregivers of patients diagnosed with cancer within the past 5 years.
  • Participants complete 6 eCARE sessions over an 8-week period.
  • Phase 2 invites 20 caregivers to participate in semi-structured online interviews exploring user experience.
  • Recruitment sources include community health clinics, cancer and caregiving advocacy groups, and online postings.

The study has three specific aims: determining feasibility based on platform completion rates, assessing acceptability using validated measures, and identifying barriers and facilitators to uptake and sustained use.

  • Feasibility will be evaluated based on the proportion of participants who complete at least 66% of weekly assessments.
  • Acceptability will be assessed using the Acceptability of Intervention Measure (AIM).
  • Pre- and posttest surveys assess depression, anxiety, caregiving burden, and relational processes.
  • Qualitative data from Phase 2 interviews will be analyzed thematically to inform tool refinement.

The study received IRB approval from the University of Houston, with recruitment beginning in June 2024 and data collection expected to conclude by August 2025.

  • Data analysis is scheduled to begin in December 2025.
  • Preliminary results are anticipated by May 2026.
  • The study is registered with the identifier DERR1-10.2196/83276.

The study is designed to generate preliminary evidence to inform the design of a larger, fully powered trial and guide future implementation of remote psychological distress monitoring in oncology care.

  • Findings are intended to guide 'future implementation of remote psychological distress monitoring strategies in oncology care.'
  • The tool is described as having 'the potential to expand access to psychosocial support and facilitate timely identification of needs and coordination of services across cancer care settings.'
  • The protocol explicitly frames this as a feasibility study preceding a larger trial.

Have a question about this study?

Citation

Acquati C, Aratow M, Nazreen T, Bhattacharjee A, Marra I, Alexander A. (2026). Evaluating an AI-Enabled Mobile Mental Health Monitoring Tool Among Family Caregivers of Adults Living With Cancer: Single-Arm Feasibility and Acceptability Trial Protocol.. JMIR research protocols. https://doi.org/10.2196/83276