Quantified Flu: an individual-centered approach to gaining sickness-related insights from wearable data

Abstract

Background: Wearables have been used widely for monitoring health in general and recent research results show that they can be used for predicting infections based on physiological symptoms. So far the evidence has been generated in large, population-based settings. In contrast, the Quantified Self and Personal Science communities are comprised of people interested in learning about themselves individually using their own data, often gathered via wearable devices. Objective: We explore how a co-creation process involving a heterogeneous community of personal science practitioners can develop a collective self-tracking system to monitor symptoms of infection alongside wearable sensor data. Methods: We engaged into a co-creation and design process with an existing community of personal science practitioners, jointly developing a working prototype of an online tool to perform symptom tracking. In addition to the iterative creation of the prototype (started on March 16, 2020), we performed a netnographic analysis, investigating the process of how this prototype was created in a decentralized and iterative fashion. Results: The Quantified Flu prototype allows users to perform daily symptom reporting and is capable of visualizing those symptom reports on a timeline together with the resting heart rate, body temperature and respiratory rate as measured by wearable devices. We observe a high level of engagement, with over half of the 92 users that engaged in the symptom tracking becoming regular users, reporting over three months of data each. Furthermore, our netnographic analysis highlights how the current Quantified Flu prototype is a result of an interactive and continuous co-creation process in which new prototype releases spark further discussions of features and vice versa. Conclusions: As shown by the high level of user engagement and iterative development, an open co-creation process can be successfully used to develop a tool that is tailored to individual needs, decreasing dropout rates.