Background: The expansion of mobile health app or apps has created a growing need for structured and predictive tools to evaluate app quality before deployment. The Mobile App Rating Scale (MARS) offers a standardized, expert-driven assessment across 4 key dimensions-engagement, functionality, aesthetics, and information-but its use in forecasting user satisfaction through predictive modeling remains limited.
Objective: This study aimed to investigate how k-means clustering, combined with machine learning models, can predict user ratings for physical activity apps based on MARS dimensions, with the goal of forecasting ratings before production and uncovering insights into user satisfaction drivers.
Methods: We analyzed a dataset of 155 MARS-rated physical activity apps with user ratings. The dataset was split into training (n=111) and testing (n=44) subsets. K means clustering was applied to the training data, identifying 2 clusters. Exploratory data analysis included box plots, summary statistics, and component+residual plots to visualize linearity and distribution patterns across MARS dimensions. Correlation analysis was performed to quantify relationships between each MARS dimension and user ratings. In total, 5 machine learning models-generalized additive models, k-nearest neighbors, random forest, extreme gradient boosting, and support vector regression-were trained with and without clustering. Models were hypertuned and trained separately on each cluster, and the best-performing model for each cluster was selected. These predictions were combined to compute final performance metrics for the test set. Performance was evaluated using correct prediction percentage (0.5 range), mean absolute error, and R². Validation was performed on 2 additional datasets: mindfulness (n=85) and older adults (n=55) apps.
Results: Exploratory data analysis revealed that apps in cluster 1 were feature-rich and scored higher across all MARS dimensions, reflecting comprehensive and engagement-oriented designs. In contrast, cluster 2 comprised simpler, utilitarian apps focused on basic functionality. Component+residual plots showed nonlinear relationships, which became more interpretable within clusters. Correlation analysis indicated stronger associations between user ratings and engagement and functionality, but weaker or negative correlations with aesthetics and information, particularly in cluster 2. In the unclustered dataset, k nearest neighbors achieved 79.55% accuracy, mean absolute error=0.26, and R²=0.06. The combined support vector regression (cluster 1)+k-nearest neighbors (cluster 2) model achieved the highest performance: 88.64% accuracy, mean absolute error=0.27, and R²=0.04. Clustering improved prediction accuracy and enhanced alignment between predicted and actual user ratings. Models also generalized well to the external datasets.
Conclusions:
Background: Emergency department (ED) overcrowding and avoidable revisits challenge health systems, with approximately 20% of patients returning within 30 days. ED-based transitional care interventions, including automated SMS text messaging, offer scalable, cost-effective means to improve follow-up, though evidence remains limited.
Objective: This study evaluated a transitional care intervention combining SMS text messaging and virtual transitional care visits to reduce ED revisits and improve outpatient follow-up.
Methods: This retrospective observational cohort study included patients discharged from 4 EDs within a single US health system between September 2023 and September 2024. Patients were categorized into two groups based on intervention engagements: (1) completed (requested, scheduled, and completed a visit) and (2) noncompleted (requested, scheduled, and did not complete). The primary outcome was spontaneous, unplanned ED revisits within 90 days; secondary outcomes included outpatient follow-up and time to first outpatient evaluation. Between-group differences were assessed using descriptive statistics and multivariable regression models (with P<.05 considered statistically significant).
Results: Of 68,115 discharged patients, 42.72% (29,100/68,115) received an automated SMS text messaging for the virtual transitional care program, and 2.93% (853/29,100) accessed the scheduling link. Of these, 56.5% (482/853) requested a visit, 49.8% (240/482) scheduled, and 70% (168/240) completed the visit (completed group). Among 72 noncompleted patients, 57% (n=41) did not show, 32% (n=23) canceled, and 11% (n=8) scheduled 2 appointments but completed neither. Nearly half (35/72, 49%) of the noncompleted group had a subsequent ambulatory follow-up. Demographics, comorbidities, and acuity were similar. The noncompleted group was nearly twice as likely to return to the ED within 90 days (21/72, 29% vs 28/150 18.7%; χ21=4.20, P=.04; odds ratio 2.11, 95% CI 1.02-4.33), while the completed group was more likely to complete outpatient follow-up (35/72, 49% vs 51/168, 30.4%; χ21=6.60, P=.01; odds ratio 2.15, 95% CI 1.03-4.77). Time to first outpatient visit did not differ significantly between groups (mean 15.7, SD 19.0 d vs mean 19.8, SD 20.7 d; Δβ=-1.93, 95% CI -10.09 to 6.42; P=.65).
Conclusions: A combined SMS text messaging and virtual transitional care program lowered 90-day ED revisits and increased outpatient follow-up, but engagement was low (2.9%). Future work should focus on optimizing care delivery and developing strategies to expand reach across the broader ED discharge population.
Background: Accessible mental health support is essential for military members (MMs), veterans, and public safety personnel (PSP) who are at an increased risk of mental health challenges. Unique barriers to care, however, often leave these populations going untreated. Mental health treatment delivered via apps or websites (ie, digital mental health interventions [DMHIs]) offers an accessible alternative to in-person therapy.
Objective: We aimed to synthesize the current literature on apps and web-based programs focused on enhancing MMs', PSPs', and veterans' resilience or well-being. A multidimensional well-being model, I-COPPE (interpersonal, community, occupational, physical, psychological, economic, and overall well-being), was used as a framework guiding the scoping review.
Methods: A search of 6 databases was conducted using key terms related to (1) population, (2) resilience and well-being constructs, and (3) web- or mobile-based programs. At all levels of screening, at least 2 researchers (RRA, MAM, and CA) reviewed each paper independently. Data were extracted and recorded to include relevant study characteristics including program name and description, target population, number of participants, therapeutic approach, results, limitations, and I-COPPE dimension supported. A narrative synthesis was performed to summarize the eligible studies.
Results: In total, 44 papers were included in the study and 39 unique resilience or well-being apps or web-based programs identified for MMs, PSP, or veterans. The programs largely focused on veteran populations (28/44, 64%). In total, 51% (20/39) of programs relied on cognitive behavioral approaches and most aimed to support posttraumatic stress disorder-related symptoms. In consideration of the I-COPPE model, a majority supported psychological well-being, followed by interpersonal and physical well-being. Most apps were believed to support more than 1 domain of well-being. The main methodologies used in the literature to evaluate digital mental health interventions include randomized controlled trials, secondary analyses, and pilot randomized controlled trials with evaluations of feasibility, acceptability, satisfaction, or qualitative feedback. Generalizability of findings was commonly limited by attrition rates and small sample sizes.
Conclusions: DMHIs for MMs, PSP, and veterans appear promising due to their accessibility and scalability. More research is needed, however, to determine whether DMHIs are an effective alternative to in-person mental health care. The current review contributes to the literature by compiling evidence of DMHIs and the domains of well-being supported by, and the therapeutic orientation of, these programs. Our review revealed that more research is needed to determine the effectiveness and efficacy of DMHIs offered to these populations.
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：