Data Challenge: A Statistically-Driven Interactive Dashboard for Geo-Temporal Surveillance and Empirical Antibiotic Guidance

Antimicrobial resistance (AMR) surveillance data is often underutilized for real-time clinical and public health decision-making. To bridge this gap, we propose developing an interactive, statistically informed dashboard called AMR Navigator to visualize resistance trends over time and geography and to guide empirical treatment recommendations at the regional level.
Using the ATLAS dataset, which includes over 900,000 global bacterial isolates with detailed metadata across 20 years, our project will model temporal trends in pathogen–drug resistance using linear regression and trend tests. We will apply geospatial comparisons and significance testing to identify regions with persistently high or low resistance levels. These findings will power a recommendation engine that suggests empirical treatment options by pathogen and location, based on statistically significant differences in susceptibility rates.
The dashboard will be built using R’s Shiny framework and will include features such as dynamic filtering (by region, drug, pathogen, and year), interactive time-series plots, resistance heatmaps, and alert zones where resistance has risen rapidly. In addition to visual analytics, the platform will embed statistical summaries such as trend significance and confidence intervals, making the insights both interpretable and actionable.
This project will equip stakeholders, especially in low- and middle-income settings, with timely and evidence-based insights. It is designed to be lightweight, scalable, and clinically relevant, with potential applications in antimicrobial stewardship, policy planning, and outbreak response. This approach will translate resistance data into accessible and evidence-backed guidance, which fills a crucial gap between surveillance and intervention.