Text Mining and Grounded Theory for Appraising the Self-Management Indicators of Diabetes Mobile Apps

Background: Understanding diabetes mobile apps functionality is fundamental to diabetes self-management because of the reliance of many patients with diabetes on these apps.

Objectives: The aim of this study is to perform a review of diabetes mobile apps to discover users’ sentiments and qualitatively examine the review comments to understand the perceptions of positive, neutral, and negative sentimental users of the apps.

Method: A total of 2678 user review comments obtained from the google play store were analysed from 47 diabetes mobile apps to understand user sentiments following clinical Self-management Indicators (SMIs) shown in previous research. Pearson correlation analysis was conducted to determine the association between the SMIs present in the apps’ and user review indicators such as rating score, user sentiment and the number of downloads. The users’ review comments were thematically screened using grounded theory to establish the themes to describe their perception of the apps.

Results: After evaluating SMIs such as weight tracking/BMI, sugar level monitoring, diet/Calories management, medication reminder, etc., 74.47% of the apps were found to have Sugar Level Monitoring(SLM) capabilities with 10.64% designed to track weight/BMI. There are 53.19% of the apps that can manage diet/calories and have data storage and security SMIs, however, less than 30% of them provide medication adherence, exercise management, doctor's appointment scheduling, and diabetes information repository. The number of the SMIs included in apps did not influence users’, but the value derived from the functionality of the apps.

Conclusions: Users are satisfied with the apps that are easy to use, setup, provide good analytics for blood sugar monitoring and have uncrowded graphical outputs and user interface. Proper data management and contemporary information about diabetes are among the identified challenges of the apps that were found to crash relentlessly on downloading, uploading, installing, and setup.