Exploring clinician preferences for visualizing wearable health data: insights from UK anesthesiologists and Mexican general practitioners

Background Wearable Health Monitoring Systems (WHMS) have the potential to transform clinical decision-making by providing real-time physiological, behavioral, and environmental data. However, the complexity and volume of these multi-source, multi-variable datasets present challenges for clinical interpretation. Objective This study examined clinicians' perspectives on how multi-source wearable health data could be summarized and visualized to support clinical interpretation, drawing on the views of anesthesiologists in an acute-care setting and general practitioners involved in primary care. Methods A custom WHMS prototype, PatientSense, was used to collect data from participants over multiple days and nights. We developed and presented scatterplots, histograms, heatmaps, and icon-based behavioral summaries derived from the prototype developed and evaluated them through one-on-one semi-structured interviews with four anesthesiologists in the UK and focus groups with twenty-three general practitioners (GPs) in Mexico. Results Clinicians expressed a preference for a hierarchical visualization strategy that begins with summarized trends and supports drill-down into finer temporal detail. Preferences for granularity differed between the two clinician groups: anesthesiologists favored higher-frequency views for interpreting rapid physiological changes, whereas GPs preferred broader summaries for longitudinal monitoring. Clinicians also highlighted the need for validation across diverse populations and clinical settings. Conclusion Clinician feedback indicated that context-aware, adaptable visualization approaches were perceived as useful for interpreting multimodal wearable health data. Integrating behavioral and physiological signals and providing flexible temporal granularity were identified as important design requirements for supporting clinical interpretation. Further evaluation in real clinical settings is required to determine their effects on clinical decision-making and patient care.