What Do People Want in a Smoking Cessation App? An Analysis of User Reviews and App Quality

摘要

Introduction Mobile smoking cessation (mCessation) apps have the potential to complement and enhance existing interventions, but many are of low quality. Exploring app reviews can provide a broader understanding of user experiences and engagement, to enhance the quality, acceptability, and effectiveness of future developments. Methods Publicly available user reviews and ratings of smoking cessation apps were mined from Google Play and the App Store via a targeted two-stage search strategy. English language smoking cessation apps with at least 20 consumer reviews between 2011 and 2020 were included. User reviews were thematically analyzed using Braun and Clarke's framework. Apps were independently scored using the Mobile Apps Rating Scale (MARS) and compared to average user star ratings. Results Forty-eight versions of 42 apps, encompassing 1414 associated reviews, met eligibility criteria. Inductive coding of reviews produced 1084 coding references including reviews coded across multiple nodes. Themes generated included: (1) supportive characteristics/tools; (2) useability; (3) influence on smoking behavior; (4) benefits of quitting; and (5) role as a supplementary tool for quitting. The mean MARS score of 36 free and accessible apps was 3.10 (SD 0.71) with mean scores ranging from 2.00 to 4.47. An inverse relationship between MARS scores and average user star ratings was observed. Conclusions App personalization, relationality, functionality, and credibility were important to users, and should be considered as key design components for future apps. Differences between user star ratings and MARS scores may illustrate competing priorities of consumers and researchers, and the importance of a codesign development method. Implications This is the first study to use unsolicited user reviews from a large population to understand the general mCessation user experience in relation to making a quit attempt. Our findings highlight specific features favored and disliked by users, including their influence on engagement, and supports previous findings that mCessation applications need to be highly tailorable, functional, credible, and supportive. We recommend a consumer-driven, co-design approach for future mCessation app developments to optimize user acceptability and engagement.