CO-DESIGNING ALIVIADO CAREGIVING - An AI-POWERED MOBILE APP FOR SUPPORTING CARE PARTNERS IN MANAGING BPSD

Sep 28, 2026

Uploaded by Putri

Co-design of Aliviado Caregiving addresses the daily care-partner burden in dementia, especially managing behavioral and psychological symptoms of dementia (BPSD) such as agitation. Care partners often lack timely access to evidence-based non-pharmacological interventions and decision support, limiting effective self-management. The work develops a precision heuristic machine-learning approach using an XGBoost model trained and tested on patient data, then embeds it into a mobile health app. A community advisory board co-designed and iteratively refined the app through focus groups and interviews, yielding four usability and prioritization themes supporting acceptability and future pilot testing.

LanguageEnglish

CategoryHealthcare

FormatPDF

Pages1

Reading Time3 min

AccessFree to download

Abstract Overview

Care Challenges in Dementia

AI Model Development and Performance

Human-Centered Co-Design Process

Key Qualitative Themes

Implications for Pilot Testing

740 Innovation in Aging, 2025, Vol. 9, No. S2 Abstract citation ID: igaf122.2518 CO-DESIGNING ALIVIADO CAREGIVING: AN AI-POWERED MOBILE APP FOR SUPPORTING CARE PARTNERS IN MANAGING BPSD Moroni Fernandez Cajavilca1, Shih-Yin Lin2, Benedict Guzman3, Kimberly Cheng4, Denise Lawson5, Narges Razavian4, and Abraham Brody2, 1. University of Utah, Salt Lake City, Utah, United States, 2. New York University Rory Meyers College of Nursing, New York, New York, United States, 3. New York University Langone Health, New York, New York, United States, 4. New York University, New York, New York, United States, 5. Pop Services, White Plains, New York, United States Care partners (CP) of persons living with dementia (PLWD) often face complex, unpaid caregiving responsibilities, particularly in managing behavioral and psychological symptoms of dementia (BPSD) such as agitation. CPs lack day-to-day access to evidence-based non-pharmacological interventions (e.g., music therapy) and decision support for daily BPSD management. Mobile health interventions offer a promising solution to aid CPsin managing BPSD. This presentation details the development of a precision heuristic machine learning algorithm and human-centered design of Aliviado Caregiving, a mobile health application designed to support CPs’ self-management and prioritization of BPSD. We used the National Alzheimer’s Coordinating Center Uniform Data Set to create an XGBoost model, training it on 2360 patients (80% of the final cohort) and testing it on 590 patients (20%) . The model achieved an area under the curve of .829 and an average precision-recall curve of .905. We then embedded the algorithm into the Aliviado Caregiving app and established a community advisory board of diverse CPs of PLWD (N = 8). The board members participated in two focus groups and eight individual interviews to co-design and iteratively refine the Aliviado Caregiving application. We conducted a reflexive thematic analysis on open-ended responses collected from all focus groups and interviews. We identified four qualitative themes: 1) User-Interface Design Preferences and Suggestions, 2) User-Friendly Language, 3) User-Concern Prioritization, and 4) Future Application Features. Applying co-design principles and partnering with diverse CPs improved our ability to enhance the acceptability and usability of Aliviado Caregiving for future testing in a pilot project.

FAQs

What problem does Aliviado Caregiving target for dementia care partners?

It targets care partners’ challenges in managing behavioral and psychological symptoms of dementia (BPSD), including limited access to evidence-based non-pharmacological interventions and daily decision support.

How was the AI component developed for the app?

The study used the National Alzheimer’s Coordinating Center Uniform Data Set to build an XGBoost model trained on 2360 patients and tested on 590 patients.

How did co-design shape the final app?

A community advisory board of diverse care partners participated in focus groups and interviews, which informed iterative refinements identified through reflexive thematic analysis.

Sep 28, 2026

Uploaded by Putri