Designing Adaptive User Interfaces for mHealth Applications Targeting Chronic Disease: A User-Centered Approach
Wei Wang, John Grundy, Hourieh Khalajzadeh, Anuradha Madugalla, Humphrey O. Obie · 2026 · ACM Transactions on Software Engineering and Methodology · doi:10.1145/3731750
Summary
This paper from Monash and Deakin Universities addresses a persistent gap in mHealth design: most chronic disease self-management apps treat users as a homogeneous group, ignoring the wide variation in capabilities, health status, cultural background, and technological literacy among people who need these tools most. The authors argue that chronic disease patients are significantly underrepresented in mainstream accessibility research — which tends to focus on blindness, low vision, and physical impairment — despite facing substantial and evolving usability barriers. The study has two stages. In Stage 1, the team built an Adaptive User Interface (AUI) prototype covering three adaptation types — presentation, content, and behaviour — and evaluated it through iterative focus groups, semi-structured interviews, and an independent survey of 90 participants with chronic conditions (cardiometabolic diseases 52%, respiratory, immune-related, and mental health). Data were analysed using Socio-Technical Grounded Theory (STGT), identifying four challenge categories around what to adapt, who initiates adaptation, how to adapt, and when. Three contextual factors shaped individual preferences: user involvement level (active vs. passive), prior experience with mHealth apps, and health condition severity. In Stage 2, Stage 1 insights were synthesised with existing literature into an initial set of guidelines. These were evaluated by 20 end-users and 43 software practitioners via sequential surveys, refined iteratively, and validated through a case study of four widely used diabetes management apps (mySugr, Gluroo, Health2Sync, LibreLinkUp). User reviews from over 4,000 app store comments were analysed to corroborate expert evaluation findings. The final output is nine guidelines organised into four groups: User Support and Interaction, Context-Aware Adaptations, Caregiver Collaboration and Adaptation, and Empowerment and Autonomy.
Key findings
Among 90 survey participants, content complexity (59%) and graphic design (58%) were the most preferred adaptation types — a notable divergence from prior AUI research, which has focused predominantly on graphic design. Fifty-four percent preferred mixed-initiative adaptation (a balance of system automation and user control) over fully automatic (38%) or fully manual systems (8%). Statistically significant associations were found between demographic factors and adaptation preferences: Chinese participants were significantly less likely to prefer content complexity (OR = 0.125) and add-on function adaptations (OR = 0.215), consistent with Hofstede's Uncertainty Avoidance dimension. Participants over 45 showed a markedly higher preference for multimodal interaction (OR = 5.824) and additional functions (OR = 3.764). Those with cardiometabolic diseases had significantly stronger preferences for goal-aligned data adaptations (OR = 13.948). In the case study, the test guidelines identified 33 issues compared to 49 from the Xcertia control guideline, but the test guideline issues were rated substantially higher in severity (average 3.2 vs. 1.5 on a 5-point scale). Ninety-one percent of software practitioners preferred the proposed guidelines over existing standalone mHealth usability guidelines. User review analysis corroborated expert-identified issues: 59 of 131 filtered reviews reflected empowerment and autonomy concerns, 46 raised caregiver collaboration issues, and 26 highlighted user support problems. A recurring user complaint was the absence of role differentiation between patients and caregivers, leading to confusion about whose data was being displayed.
Relevance
This paper is directly relevant to digital accessibility practice on multiple fronts. It explicitly frames chronic disease management as an accessibility problem, noting that chronic conditions — which affect 41 million deaths annually and 74% of all global fatalities — remain underrepresented in accessibility research despite generating substantial usability barriers. The nine finalized guidelines constitute a practitioner-ready framework that extends WCAG principles (WCAG 2.0, 2.1, and 2.2 are discussed in context) into the specific domain of adaptive mHealth design. For accessibility practitioners, the finding that user preferences vary significantly by age, nationality, and disease type has direct implications for how inclusive design must go beyond one-size-fits-all compliance. The paper's treatment of caregiver collaboration as a design axis — including privacy-conscious data sharing, role-based customisation, and audit trails — is particularly valuable for teams building health apps used by people with support networks. Limitations include a sample skewed toward Australia (49%) and younger adults, and the case study covers only four diabetes apps, so generalisability across other chronic diseases needs further validation.
Tags: mHealth · adaptive user interfaces · chronic disease · accessibility · user-centered design · mobile accessibility · guidelines · self-management
Standards referenced: WCAG 2.0 · WCAG 2.1 · WCAG 2.2