Investigating the Quality of Mobile Apps for Drug-Drug Interaction Management Using the Mobile App Rating Scale and K-Means Clustering: Systematic Search of App Stores
Bibliographic record
Abstract
Background: Drug-drug interactions (DDIs) pose a significant risk to patient safety and increase health care costs. Mobile apps offer potential solutions for managing DDIs, yet their quality and effectiveness from the user's perspective remain unclear. Objective: The aim is to evaluate the quality of publicly available mobile apps for DDI management in the US using the Mobile App Rating Scale (MARS) and to identify patterns that reflect user satisfaction and preferences. Methods: A structured review was conducted to identify mobile apps for DDI management, resulting in 19 eligible apps. Two health care-affiliated evaluators independently assessed each app using the mobile app rating scale (MARS). Dimensionality scores were calculated, and correlation analysis was conducted to examine relationships among dimensions. K-means clustering was applied to group apps based on their MARS scores. Scatter plots visualized app distributions across clusters. To validate the clustering model and assess alignment with user satisfaction, mean weighted user ratings were compared with mean MARS scores per cluster. Correlation analysis was also performed between individual MARS dimensions and user ratings within each cluster. Results: The mean MARS score was 3.54 out of 5, with the Information dimension scoring the highest (mean 3.68, SD 0.51) and Engagement the lowest (mean 3.42, SD 0.80). The Kruskal-Wallis test revealed no significant differences in median scores across the four dimensions (χ²3=2.109, P=.55). All MARS dimensions were positively correlated (r=0.65 to 0.92), indicating interrelated quality characteristics. K-means clustering identified three app groups with varying quality profiles: Cluster 1 (n=7, mean MARS=2.86), Cluster 2 (n=7, mean=3.57), and Cluster 3 (n=5, mean=4.44). Cluster 1 apps showed strongest correlations between user satisfaction and functionality (r=0.74) and engagement (r=0.53). Cluster 2 users prioritized information (r=0.41) and aesthetics (r=0.58), and Cluster 3 exhibited balanced influence from information (r=0.62), aesthetics (r=0.58), and functionality (r=0.39). Scatter plots indicated that engagement, functionality, and aesthetics were key drivers of user perception, while information, though consistently strong, played a lesser role in differentiating the apps. The weighted user ratings aligned with MARS scores, supporting the validity of the clustering model. Conclusions: This study assesses the quality of mobile apps for DDI management by integrating MARS with K-means Clustering. This approach enabled a structured classification of apps based on the MARS scores, identifying distinct clusters that reflect overall app quality profiles across key usability dimensions. The study revealed that the influence of MARS dimensions on app ratings varies by cluster, highlighting that the significance of these dimensions shifts according to the specific needs and preferences of different user groups.
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 imitationNot 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.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.032 | 0.146 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.003 | 0.006 |
| Bibliometrics | 0.019 | 0.012 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.002 | 0.003 |
| Open science | 0.002 | 0.003 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
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.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
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".