Separating Features From Functionality in Vaccination Apps: Computational Analysis
George Shaw Jr,Devaki Nadkarni,Eric Phann,Rachel Sielaty,Madeleine Ledenyi,Razaan Abnowf,Qian Xu,Paul Arredondo,Shi Chen
DOI: https://doi.org/10.2196/36818
2022-10-11
Abstract:Background: Some latest estimates show that approximately 95% of Americans own a smartphone with numerous functions such as SMS text messaging, the ability to take high-resolution pictures, and mobile software apps. Mobile health apps focusing on vaccination and immunization have proliferated in the digital health information technology market. Mobile health apps have the potential to positively affect vaccination coverage. However, their general functionality, user and disease coverage, and exchange of information have not been comprehensively studied or evaluated computationally. Objective: The primary aim of this study is to develop a computational method to explore the descriptive, usability, information exchange, and privacy features of vaccination apps, which can inform vaccination app design. Furthermore, we sought to identify potential limitations and drawbacks in the apps' design, readability, and information exchange abilities. Methods: A comprehensive codebook was developed to conduct a content analysis on vaccination apps' descriptive, usability, information exchange, and privacy features. The search and selection process for vaccination-related apps was conducted from March to May 2019. We identified a total of 211 apps across both platforms, with iOS and Android representing 62.1% (131/211) and 37.9% (80/211) of the apps, respectively. Of the 211 apps, 119 (56.4%) were included in the final study analysis, with 42 features evaluated according to the developed codebook. The apps selected were a mix of apps used in the United States and internationally. Principal component analysis was used to reduce the dimensionality of the data. Furthermore, cluster analysis was used with unsupervised machine learning to determine patterns within the data to group the apps based on preselected features. Results: The results indicated that readability and information exchange were highly correlated features based on principal component analysis. Of the 119 apps, 53 (44.5%) were iOS apps, 55 (46.2%) were for the Android operating system, and 11 (9.2%) could be found on both platforms. Cluster 1 of the k-means analysis contained 22.7% (27/119) of the apps; these were shown to have the highest percentage of features represented among the selected features. Conclusions: We conclude that our computational method was able to identify important features of vaccination apps correlating with end user experience and categorize those apps through cluster analysis. Collaborating with clinical health providers and public health officials during design and development can improve the overall functionality of the apps.