Assessing the global reach and value of a provider-facing healthcare app using large-scale analytics

Abstract

Background: The rapid global adoption of mobile health (mHealth) smartphone apps by healthcare providers presents challenges and opportunities in medicine. Challenges include ensuring the delivery of high-quality, up-to-date and optimised information. Opportunities include the ability to study global practice patterns, access to medical and surgical care and continuing medical education needs.

Methods: We studied users of a free anaesthesia calculator app used worldwide. We combined traditional app analytics with in-app surveys to collect user demographics and feedback.

Results: 31 173 subjects participated. Users were from 206 countries and represented a spectrum of healthcare provider roles. Low-income country users had greater rates of app use (p<0.001) and ascribed greater importance of the app to their practice (p<0.001). Physicians from low-income countries were more likely to adopt the app (p<0.001). The app was used primarily for paediatric patients. The app was used around the clock, peaking during times typical for first start cases.

Conclusions: This mHealth app is a valuable decision support tool for global healthcare providers, particularly those in more resource-limited settings and with less training. App adoption and use may provide a mechanism for measuring longitudinal changes in access to surgical care and engaging providers in resource-limited settings. In-app surveys and app analytics provide a window into healthcare provider behaviour at a breadth and level of detail previously impossible to achieve. Given the potentially immense value of crowdsourced information, healthcare providers should be encouraged to participate in these types of studies.

Keywords: analytics; anesthesiology; global health; mHealth; practice patterns.

Figure 1

Overall roadmap of the approach to analysis of the data. This indicates the dependent and independent variables that were examined (see Methods) and the number of study participants in each of these subsets. Subsets were combined to create various tables and figures, and the number of study participants with data in the combined subsets are shown.

Figure 2

Standard boxplots demonstrating frequency of app use as a function of key user characteristics (dots are outliers). N per category is shown. Negative binomial regression was used to assess the significance of the association between these factors. (A) User primary country World Bank income level and (B) provider role/provider type. NS, not significant.

Figure 3

Penetration of app into the physician workforce by country. The app adoption penetration index was calculated as the estimated number of physician app users per 1000 physicians in the country. WHO Global Health Observatory data were used to obtain the estimated total number of physicians in the country. (A) Choropleth map; no data for countries in white. (B) Standard boxplot showing the app adoption penetration index grouped by World Bank country income level. Number of countries in each category is shown. Negative binomial regression was used to test the significance of the association between country income level and the app adoption penetration index (see online supplementary appendix).