ABSTRACT
Barriers to mental health care, including stigma, costs, and mental health professional shortages, have been exacerbated by the COVID-19 pandemic. Smartphone apps have the potential to increase scalability and improve access to mental health information, support, and interventions. However, evaluating these apps and selecting ones for use in care remain challenging, especially as apps are often updating and changing. Recommending apps requires knowledge of how stable apps are as the experience of one user several months ago may or may not be the same. A sample of 347 apps of the 650 apps on the M-health Index and Navigation Database (MIND) https://mindapps.org were reviewed between September 1, 2021, and January 5, 2022. Apps were selected by time since their last review, with updates occurring on average approximately 4 months from the last review. Eleven trained app evaluators reviewed apps across 105 evaluation criteria in 9 categories. Results were compared to initial ratings, identifying the changes that occurred. The average app updates every 433 days, though 19% were updated in the last 3 months and some nearly weekly. Changes in privacy and features made up the highest percentage of changes, both at 38%. The most frequently observed privacy-related change was increased privacy policy reading level. Functionality parameters changed in 28% of apps. The most common functionality change was the removal of an accessibility feature. Clinical foundations changed in 18% of apps and 9% added supporting studies. Cost structure changed in 17% of apps, with 10% adding a fee for use of the app. Engagement features changed in 17% of the apps, with additions and removals of validated assessments or screeners most common. The dynamic nature of the app stores is reflected in app privacy, features, and functionality. These changes, reflected by the increased reading levels required to understand privacy policies, the decrease in accessibility features, and the additions of fees to access mobile apps, reflect the need to constantly review apps and understand how they are evolving. Patient and clinicians should use the most recent and updated possible when evaluating apps.
ABSTRACT
Background: COVID-19 has increased interest and uptake of mental health apps. As healthcare payers consider reimbursement for some apps and creation of app formularies, questions remain about the state of these apps and how to judge their value. As apps are not static and constantly changing and updating, we investigated the dynamics of the app marketplaces to better understand their privacy, efficacy, and safety features in the 'real world'. Methods: A sample of 250 apps were reviewed between September 1st 2021 and December 1st 2021. Apps were selected from mindapps. org based on their age of last review on mindapps.org, with apps approaching six months since being updated in this database reviewed first. Mindapps.org is the world's largest free-to-access database of mental health apps that currently curates 650 apps each rated across 105 dimensions. For this study, each of the 250 apps were again reviewed across 105 dimensions and results were compared to changes from the prior rating of the same app ∼6 months ago. Results: The average app updates every 364 days although some nearly weekly. Privacy features changed the most since the last update with 38% of apps reflecting a change. 37% of apps changed in regards to features offered with the most common change as adding physical exercise related content to the apps. Functionality and access changed in 27% of apps with the most common change being increase in accessibility features as well as offering offline access to the app. 20% of apps changed their cost structure with 12% adding a fee for use of the app. Reviewers noted that ads to support apps were in some cases intrusive. Discussion: The dynamic nature of the app stores is reflected in app privacy, features, and functionality. For payers to be able to offer app formularies, they need to create metrics to assess these changes in the app landscape or else tightly control app updates themselves. Other aspects necessary for valuation such as real-world engagement with the apps as well as mean effect size are not publicly available and thus not reported in our results. To truly understand the economic value of mental health apps, dynamic pricing models combined with transparency in data reporting from apps will be necessary.