HealthMag: Inclusive Digital Health Evaluation

• HealthMag is a systematic, evidence-based framework that models and evaluates health-related inclusivity in digital health software.
• It uses facet-aware cognitive walkthroughs and empirical metrics to identify and remedy usability challenges for users with chronic conditions and fluctuating health statuses.
• The dual-lens approach, integrating AgeMag, enables tailored evaluations and design improvements for heterogeneous user populations.

HealthMag

HealthMag refers to a systematic, evidence-based evaluation framework designed to elicit, model, and assess health-related inclusivity requirements in digital health (DH) software. By operationalizing health-relevant “facets” (interaction-impacting user attributes grounded in empirical evidence) and leveraging cognitive walkthroughs that interrogate software tasks through these lenses, HealthMag aims to uncover and remedy inclusivity failures affecting users living with chronic conditions, fluctuating symptoms, or diverse health statuses. Originally inspired by the GenderMag method, HealthMag is a direct member of the InclusiveMag family, providing both core health-related and, in its dual-lens variant, age-related evaluative capabilities to support inclusive DH software design and evaluation for heterogeneous populations (Xiao et al., 30 Jan 2026).

1. Theoretical Foundations and Motivation

The prevailing digital health paradigm often presumes normative user characteristics, neglecting the variability in physical, cognitive, and psychosocial capacities arising from chronic illness, persistent symptoms, or aging. This has resulted in DH software that, despite meeting clinical objectives, fails to support real-world interaction needs, leading to "inclusivity bugs." HealthMag is defined as a “compact, validated lens that (i) models health-related requirements in ways engineers can act on and (ii) detects developer-assumption bias before software reaches actual users.” Its core function is to ensure that software accommodates varying levels of health self-efficacy, motivation, received care, trust/privacy attitudes, technical proficiency, and accounts for breakdowns stemming from comorbidities or fluctuating symptoms (Xiao et al., 30 Jan 2026).

The Intellectual provenance of HealthMag derives from GenderMag, a method that leverages evidence-driven facets and personas for software inclusivity evaluation along gender-difference axes. HealthMag abstracts the same meta-methodology to the health domain: it identifies critical health facets, devises research-based personas spanning these facets’ value ranges, and structures facet-aware walkthroughs, systematically surfacing interaction breakdowns tied to health status (Xiao et al., 30 Jan 2026).

2. Facet Identification, Mapping, and Calibration

HealthMag’s development followed the InclusiveMag pipeline comprising four iterative phases:

1. Mapping: Sourced from systematic literature reviews (SLRs) spanning software engineering (SE), human–computer interaction (HCI), clinical, and social science, HealthMag’s candidate pool included 16 health-related facets. A support-frequency metric was maintained (e.g., "Received Care" appeared in 44 papers, "Cultural Competency" in 4).
2. Foundation: Candidate facets were clustered and consolidated with operational definitions grounded in primary and review-literature (e.g., the linkage between social support and patient activation).
3. Developing: Inclusion criteria were imposed: (C1) empirical evidence of UX impact, (C2) design relevance, (C3) wide value range, (C4) plain-language usability. This filtered the pool to a tentative set of seven one-word-labeled facets, including “Motivation,” “Received Care,” “Tech Proficiency,” “Cognitive Load,” “Trust & Privacy,” “Co-morbidities,” “Health Self-Efficacy.”
4. Calibration: Through semi-structured interviews and forced-ranking protocols with a multidisciplinary expert panel (N=10), facet mean ranks (μ) and standard deviations (σ) were collected, resulting in a final calibrated set of five facets. For example, “Motivation” (μ=3.3, σ=2.1) and “Health Self-Efficacy” (μ=4.3, σ=2.5) were consistently retained (Xiao et al., 30 Jan 2026).

The final selection yields a facet set in which usability can often be conceptualized as an “AND-gate” logic among facets (e.g., $$Usability = A_{\mathrm{facet}} \land B_{\mathrm{facet}}$$).

3. Dual-Lens Extension: Integrating AgeMag for Elderly Populations

To differentiate between health- and age-driven failures and reveal intersectional breakdowns, HealthMag was structurally merged with a recalibrated version of AgeMag, yielding the “Elderly HealthMag” dual-lens tool. The dual-lens taxonomy comprises eight non-overlapping facets organized across three interaction-time layers:

Layer	Facets (non-overlapping)
Intrinsic Capacity (“Can Use”)	Visual Impairment, Physical Difficulties
Health Drivers (“Want to Use”)	Motivation, Health Self-Efficacy, Trust & Privacy
Contextual Enablers (“How to Use”)	Willingness, Tech Proficiency, Received Care

AgeMag facets were imported and recalibrated via expert ranking for users aged 65+:

• Visual Impairment (e.g., contrast, text size)
• Physical Difficulties (e.g., dexterity, tremor)
• Tech Proficiency (device and app comfort)
• Willingness (risk aversion, fear of error)
• Education Level (digital literacy proxy)

Structurally overlapping facets (e.g., Tech Proficiency) were merged, while distinct constructs (e.g., Trust vs. Willingness) were preserved (Xiao et al., 30 Jan 2026).

4. Methodological Workflow: Personas, Cognitive Walkthroughs, and Bug Tagging

HealthMag operationalizes evaluation via persona-driven, facet-aware cognitive walkthroughs (CW):

1. Persona Construction: LLM-assisted, evidence-driven workflows generate three or more personas (e.g., Margaret, Zhao, Kamala), each spanning the range of HealthMag and AgeMag facet endpoints with embedded use scenarios.
2. Walkthrough Procedure:
   • Evaluators are trained on the framework and personas.
   • Given a typical DH use case (e.g., medication management: adding drugs, dose-logging, record-sharing), each subgoal is interrogated using facet-specific questions.
   • Example questions include “Does the display scale enough for a user with poor vision?” or “Does this step require medical jargon they won’t understand?” (targeting Visual Impairment and Health Self-Efficacy, respectively).
   • Outcomes (Yes/Partial/No), task ease, errors, and inclusivity bugs (with facet tags) are logged.
   • Aggregated adjudication and design remedies address tagged issues.

Facet tagging enables the diagnostic separation between age- and health-driven breakdowns and systematically exposes intersectional vulnerabilities, such as compounded exclusion due to low vision and low trust when privacy prompts are unlabeled (Xiao et al., 30 Jan 2026).

5. Application to Digital Health Software: Empirical Findings and Design Guidance

Applying Elderly HealthMag to two widely used medication-management apps (Medisafe and Apple Health’s Medications), the walkthroughs exposed a consistent usability gradient. Personas with “high” facet values (e.g., strong motivation, high self-efficacy, high technical skill) completed 90% of tasks with ease, while those at lower endpoints exhibited increased failure and confusion rates. Specific findings included:

• Failures to discover icons due to small size (Physical Difficulties, Visual Impairment).
• Task hesitancy when terminology mismatched self-efficacy or educational background.
• Success despite low contrast, when motivation and trust were high.

These evaluations led directly to interface modifications (larger touch targets, plain-language prompts, proxy-sharing options), demonstrating the causal linkage between facet-driven analysis and software remediation. Design guidelines emergent from the project include:

• Early application of health- and age-aware personas across prototyping iterations.
• Composition of dual-lens analyses to distinguish cause of failure.
• Facet specialization in walkthrough prompts to permit traceability of bugs to requirements.
• Organization of facets in layered “gating” logic (Can Use → Want to Use → How to Use) (Xiao et al., 30 Jan 2026).

6. Limitations and Extensibility

The HealthMag method, while systematic and empirically grounded, does not mandate quantitative formulas beyond rank-based metrics and AND-gate logic for usability prediction. Persona and facet calibrations are necessarily subject to available literature and expert judgment; as populations and health technology contexts evolve, periodic recalibration is warranted. HealthMag artifacts (facet definitions, prompts, personas) are openly reusable and support ongoing extension to other DH contexts and diversity dimensions (e.g., gender, cognitive ability) within the InclusiveMag family (Xiao et al., 30 Jan 2026).

7. Conclusion and Research Trajectory

HealthMag represents a structured mechanism for systematically modeling, diagnosing, and remedying inclusivity gaps in digital health software for users with heterogeneous health experiences. The dual-lens Elderly HealthMag method enables discrimination of intersecting age and health barriers. By embedding evidence-driven facet analysis, persona-centric scenario walkthrough, and explicit tagging and remedy of inclusivity bugs, DH teams can address developer-assumption bias and deliver more effective, inclusive systems. Methodological transparency, reusability, and documented impacts on concrete design improvements position HealthMag as a canonical reference for inclusivity assessment and requirements engineering in contemporary and future digital health software engineering (Xiao et al., 30 Jan 2026).