MétaCan
Menu
← Back to cohort
Record W4412611124 · doi:10.2196/57279

Grouping Digital Health Apps Based on Their Quality and User Ratings Using K-Medoids Clustering: Cross-Sectional Study

2025· article· en· W4412611124 on OpenAlexvenueno aff
Maciej Hyzy, Raymond Bond, Maurice Mulvenna, Lu Bai, Jorge Martínez Carracedo, Simon Leigh

Bibliographic record

VenueJMIR mhealth and uhealth · 2025
Typearticle
Languageen
FieldHealth Professions
TopicMobile Health and mHealth Applications
Canadian institutionsnot available
FundersDepartment for the Economy
KeywordsDigital healthmHealthCluster analysisComputer scienceMetric (unit)Bonferroni correctionHealth careData miningMedicineMachine learningStatisticsPsychological interventionNursingMathematicsEngineering

Abstract

fetched live from OpenAlex

Background: Digital health apps allow for proactive rather than reactive health care and have the potential to take the pressure off health care providers. With over 350,000 digital health apps available on the app stores today, those apps need to be of sufficient quality to be safe to use. Discovering the typology of digital health apps regarding professional and clinical assurance (PCA), user experience (UX), data privacy (DP), and user ratings may help in determining the areas where digital health apps can improve. Objective: This study has two objectives: (1) discover the types (clusters) of digital health apps with regards to their quality (scores) across 3 domains (their PCA, UX, and DP) and user ratings and (2) determine whether the National Institute for Health and Care Excellence (NICE) Evidence Standard Framework's (ESF's) tier, target users of the digital health apps, categories, or features have any association with this typology. Methods: Data were obtained from 1402 digital health app assessments conducted using the Organisation for the Review of Care and Health Apps Baseline Review (OBR), evaluating PCA, UX, and DP. K-medoids clustering identified app typologies, with the optimal number of clusters determined using the elbow method. The Shapiro-Wilk test assessed normality of user ratings and OBR scores. Nonparametric Wilcoxon rank sum tests compared cluster differences in these metrics. Post hoc analysis examined the distribution of NICE ESF tiers, target users, categories, and features across clusters, using Fisher exact test with Bonferroni correction. Effect sizes were calculated using Cohen w. Results: A total of four distinct app clusters emerged: (1) apps with poor user ratings (220/1402, 15.7%), (2) apps with poor PCA and DP scores (252/1402, 18%), (3) apps with poor PCA scores (415/1402, 29.6%), and (4) higher quality apps with high user ratings and OBR scores (515/1402, 36.7%). While some statistically significant associations were found between clusters and NICE ESF tiers (2/3), target users (0/14), categories (4/33), and features (6/19), all had small effect sizes (Cohen w<0.3). The strongest associations were for the "Service Signposting" feature (Cohen w=0.24) and NICE ESF tier B (Cohen w=0.19). Conclusions: The largest cluster comprised high-quality apps with strong user ratings and OBR scores (515/1402, 36.7%). A significant proportion (415/1402, 29.6%) performed poorly in PCA despite performing well in other domains. Notably, user ratings did not consistently align with PCA scores; some apps scored highly with users but poorly in PCA and DP. The 4-cluster typology underscores areas needing improvement, particularly PCA. Findings suggest limited association between the examined app characteristics and quality clusters, indicating a need for further investigation into what factors truly influence app quality.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.

How this classification was reachedexpand

Full frame machine prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.

metaresearch head score (Codex)0.010
metaresearch head score (Gemma)0.030
Version: metacan-v3-hybrid-931329e0061cValidation status: machine_predicted_unvalidated
Candidate categoriesnone
Consensus categoriesnone
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Observational · Consensus signal: Observational
GenreCandidate signal: Empirical · Consensus signal: Empirical
Teacher disagreement score0.010
Threshold uncertainty score0.051

Distilled classifier scores by category (both heads)

CategoryCodexGemma
Metaresearch0.0100.030
Meta-epidemiology (narrow)0.0000.001
Meta-epidemiology (broad)0.0010.001
Bibliometrics0.0030.003
Science and technology studies0.0010.001
Scholarly communication0.0020.002
Open science0.0010.001
Research integrity0.0010.001
Insufficient payload (model declined to judge)0.0030.001

Machine scores (provisional)

The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.

Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.

Opus teacher head0.165
GPT teacher head0.530
Teacher spread0.365 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it

Classification

machine, unvalidated

Machine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.

The models applied no category: nothing in the taxonomy fit this work.
Study designObservational
Domainnot available
GenreEmpirical

How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".

Quick stats

Citations0
Published2025
Admission routes1
Has abstractyes

Explore more

Same venueJMIR mhealth and uhealth→Same topicMobile Health and mHealth Applications→French-language works237,207→