Evaluation of mHealth Applications Quality Based on User Ratings

The presented study covers the evaluation of ratings of a sample set of 46,430 medical applications available on Google's Play Store. It was discovered that the distribution of user ratings given to applications has a log-normal form and has a correlation with many application characteristics one would not expect to be directly related, among others the time of the last update of the app, the app vocabulary as well as descriptions. Popular applications with a large number of downloads and reviews tend to have average ratings, while the ratings of rarely downloaded apps tend to be either highly positive or negative. Despite the huge diversity and number of applications available in the market, it is highly dominated by just a few apps: 90.7% of the overall number of user ratings assigned to the apps in our sample is distributed among just 1,601 apps, corresponding to 4.1% of all apps provided within the "Medical" and "Health and fitness" categories at the time of our evaluation.