Sentiment and Thematic Analysis of User Reviews for FDA-Cleared Prescription Digital Therapeutics: A Mixed-Methods Real-World Evidence Study.

Background Prescription digital therapeutics (PDTs) are FDA-cleared, evidence-backed, smartphone-based interventions. While clinical trials establish efficacy under controlled conditions, the patient experience in real-world settings remains poorly characterized. App store reviews, though informal, offer a potential source of real-world evidence (RWE) reflecting user sentiment, barriers to engagement, and perceived benefit. Objective To characterize the real-world user experience of FDA-cleared PDTs through a mixed-methods analysis of publicly available app store reviews, combining quantitative sentiment classification with qualitative thematic analysis. Methods FDA-cleared PDTs via FDA's De Novo or 510(k) pathways were identified. Reviews were collected from the Apple App Store and Google Play Store using structured scraping tools. A rules-based sentiment classifier was applied based on user star ratings, and thematic content analysis was conducted using a hybrid natural language processing and manual coding approach. Reviews were grouped by app and platform, and dominant themes were extracted using unsupervised topic modeling. Results Of the 13 PDTs identified, seven had publicly accessible user reviews, yielding a dataset of 247 unique entries: AspyreRx, EndeavorRx, Regulora, Rejoyn, reSET, reSET-O, and Stanza. Sentiment classification revealed that 25.1% of reviews were positive, 12.6% neutral, and 62.3% negative. Thematic analysis identified six recurrent themes, including pediatric benefit and engagement, rewards and incentives, access and activation barriers, technical issues, emotional reactions and cost sensitivity, and boredom and frustration. While negative sentiment was more prevalent overall, several products, particularly Rejoyn and EndeavorRx, received meaningful reports of perceived benefit and therapeutic value. Company transitions were noted to have disrupted the continuity of historical review data of some PDTs. Conclusions App store reviews provide a rich, user-centered source of RWE and insights for PDTs. They highlight dimensions of access, usability, and trust not typically captured in clinical trials. Integrating such feedback into post-market evaluation may inform future design, regulation, and reimbursement of software-based therapeutics.