Alerting Type 2 Diabetics of Nocturnal Hypoglycemia Risk - Integrating user-centered design in a self-management app to present AI-based health predictions
Term
4. Term
Publication year
2022
Submitted on
2022-05-30
Pages
29
Abstract
People with diabetes who are insulin treated are at risk of experiencing hypoglycemia – a condition of potentially dangerously low blood glucose levels. Predictive algorithms can aid these people in preventing hypoglycemia. In this two-part study we applied user-centered design to explore how a personalized nocturnal hypoglycemia risk prediction can be communicated in a mobile self-management app (SMA) for people with type 2 diabetes. In a preliminary understanding study, we conducted semi-structured interviews with potential users (N=10) and one endocrinologist to acquire domain knowledge, establish user needs, and identify preferences for visual presentation of hypoglycemia risk. Based on our initial findings, we developed a mobile application prototype which was evaluated in a think-aloud evaluation with prospective users (N=5) to validate user needs and verify design and functionality. Our findings generally align with previous research in emphasizing the significance of using color cues and simple visualizations supplemented with concise text or numbers. We also found that the SMA should provide flexibility as users need different levels of detail and that the user should be able to personalize the prediction configuration. We further identified barriers and enablers for adoption of mobile diabetes SMA and highlight the importance of collaboration between users, health professionals, and researchers. Based on our findings we draw recommendations for future work in this field.
Keywords
Documents