Rating analysis and BERTopic modeling of consumer versus regulated mHealth app reviews in Germany

Marie Uncovska; Bettina Freitag; Sven Meister; Leonard Fehring

doi:10.1038/s41746-023-00862-3

Rating analysis and BERTopic modeling of consumer versus regulated mHealth app reviews in Germany

NPJ Digit Med. 2023 Jun 21;6(1):115. doi: 10.1038/s41746-023-00862-3.

Authors

Marie Uncovska¹, Bettina Freitag¹, Sven Meister^{2

3}, Leonard Fehring^{4

5}

Affiliations

¹ Faculty of Health, School of Medicine, Witten/Herdecke University, Witten, Germany.
² Health Care Informatics, Faculty of Health, School of Medicine, Witten/Herdecke University, Witten, Germany.
³ Department Healthcare, Fraunhofer Institute for Software and Systems Engineering, Dortmund, Germany.
⁴ Faculty of Health, School of Medicine, Witten/Herdecke University, Witten, Germany. leonard.fehring@uni-wh.de.
⁵ Gastroenterology, HELIOS University Hospital Wuppertal, University Witten/Herdecke, Wuppertal, Germany. leonard.fehring@uni-wh.de.

Abstract

Germany introduced prescription-based mobile health (mHealth) apps in October 2020, becoming the first country to offer them fully reimbursed by health insurance. These regulated apps, known as DiGAs, undergo a rigorous approval process similar to pharmaceuticals, including data protection measures and sometimes clinical trials. This study compares the user experience of DiGAs with non-prescription mHealth apps in Germany, analyzing both average app store ratings and written reviews. Our study pioneers the use of BERTopic for sentiment analysis and topic modeling in the mHealth research domain. The dataset comprises 15 DiGAs and 50 comparable apps, totaling 17,588 German-language reviews. Results reveal that DiGAs receive higher contemporary ratings than non-regulated apps (Android: 3.82 vs. 3.77; iOS: 3.78 vs. 3.53; p < 0.01; non-parametric Mann-Whitney-Wilcoxon test). Key factors contributing to positive user experience with DiGAs are customer service and personalization (15%) and ease of use (13%). However, challenges for DiGAs include software bugs (24%) and a cumbersome registration process (20%). Negative user reviews highlight concerns about therapy effectiveness (11%). Excessive pricing is the main concern for the non-regulated group (27%). Data privacy and security receive limited attention from users (DiGAs: 0.5%; comparators: 2%). In conclusion, DiGAs are generally perceived positively based on ratings and sentiment analysis of reviews. However, addressing pricing concerns in the non-regulated mHealth sector is crucial. Integrating user experience evaluation into the review process could improve adherence and health outcomes.