Challenging assumptions: narrative analysis of interviews for digital health development

Background We are in a time of transition. Where once health technology was sequestered behind the guarded curtains of research and healthcare, it is now for the masses. Successful digital health research relies on the public to direct what is relevant, representative and useful. However, too often representative and diverse perspectives are absent from digital health development, including artificial intelligence (AI) for health. As a result, assumptions about what individuals (lay persons) want and need from digital health solutions go unchecked. As part of a larger project to develop a risk prediction and prevention app, this interview study will contribute insights about its reach and potential usability amongst lay persons and potential personalization of the app. Methods In Norway and Spain, interviews with lay persons explored perceptions, expectations and beliefs about health and health technology. Thematic analysis and Lubov’s structural model provided a scaffolding for narrative analysis, used to identify nuanced relationships between participants’ views of health and health technology. Results Twenty-one individuals participated in interviews (n = 8 in Norway, n = 13 in Spain, ages 18–60 years old, n = 6 women). Themes included: personal history , societal context , social network , health status and management , health beliefs , technology experience , and beliefs about AI and technology for health . Three main narratives described participants’ relationships with health technology, “Not for me, but good for you”, digitally dependent health users, and “Not on my radar”. Two stories are presented for each narrative to exemplify the complexity of relationships between an individual, their health and health technology. Conclusion By assessing lay persons’ relationships with their health and health technology, prior to the development of an app for the prevention of chronic illness, we were able to explore the real world potential of these technologies without a presumption of use and relevance. In doing so, we identified reasons that contributed to participants’ choice to use or not to use digital health for prevention. As part of a larger parent project, these results contributed to the personalization and usability assessment of an AI-driven app meant to predict the risk for and provide recommendations for prevention of chronic diseases.