Uncertainty displays in pandemic dashboards: a review of COVID-19 use cases

Pandemic dashboards are a key resource that public health professionals and the public use to access timely, dynamic, and multivariate public health data. Epidemiological data inherently carry uncertainty due to factors related to location, time, and the movement of people and the spread of disease. Uncertainty information has been neglected in dashboard design due to some technical challenges, including the complexity of communication, and to avoid potential misinterpretation. However, including uncertainty information may improve transparency and dashboard actionability in public health. This study scanned Canadian COVID-19 dashboards to assess how uncertainty was communicated during the pandemic. A total of 69 dashboards were analyzed in 2023. The findings show that most COVID-19 dashboards did not meet best practices for improving actionability and usability, nor did they include uncertainty data. Only 1 of 69 dashboards included a visualization of uncertainty, highlighting a substantial gap between established uncertainty visualization research and real-world public health dashboard design. Additionally, most dashboards lacked a clear target user and did not provide sufficient resources to support public use. Future research can investigate how dashboard users engage in public health decision-making and change their health behaviours using uncertainty information.