Data challenge: Forecasting Pediatric Antimicrobial Resistance: A Spatiotemporal Analysis of Global Surveillance Data

Despite having distinct resistance profiles and exposure risks, children are often overlooked in AMR forecasting. Our proposal aims to fill this gap by developing an early warning system tailored to pediatric antimicrobial resistance (AMR). We can empower decision-makers with predictive insights by modeling how resistance to both individual antibiotics and combinations is likely to evolve over time, across age groups, and within regions.
We propose adapting the Spatiotemporal Antibiogram Pattern Prediction (STAPP) approach to forecasting AMR trends in pediatric populations using global surveillance data, such as the Merck/ATLAS dataset. Our focus will be on addressing a critical yet underexplored area of understanding and predicting resistance patterns among children.
Our approach begins with extracting pediatric antibiogram records, stratified by age group (0–5, 6–12, 13–18), region/country, and year. This data will inform the construction of co-resistance networks, where nodes represent antibiotics and edges indicate observed co-resistance relationships in pediatric isolates.
Using spatiotemporal machine learning techniques, we will analyze how resistance combinations evolve geographically and temporally. These models will predict the emergence of new multidrug resistance (MDR) patterns and visualize results through dynamic maps and network graphs. These visualizations will identify high-risk resistance clusters in specific pediatric populations, enabling real-time interpretation of threats.
This work has the potential to transform pediatric antibiotic stewardship. Clinicians and public health authorities could proactively adjust treatment guidelines and develop targeted stewardship strategies based on anticipated resistance trends. Our project directly supports ViVli’s goals by innovatively applying data science to a pressing global health challenge and equipping stakeholders with tools to protect vulnerable pediatric populations from rising AMR threats.