Paper Type
Short
Paper Number
PACIS2026-1647
Description
Medication adherence remains a major healthcare challenge, particularly for patients with chronic conditions requiring long-term treatment. While AI-enabled interventions have shown potential in supporting adherence, many existing solutions lack behavioral-theory grounding and user-centered design. This study applies Action Design Research (ADR) to explore how AI-enabled interventions can support long-term behavior management, using medication adherence as the application context. Informed by the COM-B model, meta-requirements and design principles were derived to address patients’ capability, opportunity, and motivation for adherence. An alpha mockup was evaluated through walkthroughs and interviews with older adults managing chronic conditions. Findings revealed that adherence challenges were related to participants’ medication-taking experience and adherence stage, leading to a stage-sensitive framework aligned with initiation, implementation, and discontinuation. Based on these insights, the beta cycle artifact was redesigned as an AI-enabled conversational intervention. This study contributes early-stage design knowledge for theory-informed AI-enabled healthcare self-management interventions.
Recommended Citation
TZENG, YUN CHEN and Lin, Yi-ling, "Designing AI Interventions for Medication Adherence: An Action Design Research Study" (2026). PACIS 2026 Proceedings. 14.
https://aisel.aisnet.org/pacis2026/ishealthcare/ishealthcare/14
Designing AI Interventions for Medication Adherence: An Action Design Research Study
Medication adherence remains a major healthcare challenge, particularly for patients with chronic conditions requiring long-term treatment. While AI-enabled interventions have shown potential in supporting adherence, many existing solutions lack behavioral-theory grounding and user-centered design. This study applies Action Design Research (ADR) to explore how AI-enabled interventions can support long-term behavior management, using medication adherence as the application context. Informed by the COM-B model, meta-requirements and design principles were derived to address patients’ capability, opportunity, and motivation for adherence. An alpha mockup was evaluated through walkthroughs and interviews with older adults managing chronic conditions. Findings revealed that adherence challenges were related to participants’ medication-taking experience and adherence stage, leading to a stage-sensitive framework aligned with initiation, implementation, and discontinuation. Based on these insights, the beta cycle artifact was redesigned as an AI-enabled conversational intervention. This study contributes early-stage design knowledge for theory-informed AI-enabled healthcare self-management interventions.
Comments
14-Healthcare