Data Challenge: Analysis of antimicrobial resistance trends and associated factors

Student team, led by Nabuuma Ruth Nambi of Kampala International University, Uganda with Kyamazima Ibrahim, Kyahurwa Nancy, Chelimo Specioza and Tabaro Joshua of Kampala International University Uganda. Title; AI-Enabled Multi-Dataset Surveillance Framework for Antimicrobial Resistance Pattern Detection and Predictive Modeling Antimicrobial resistance (AMR) is a complex evolutionary and epidemiological process driven by antimicrobial selection pressure, microbial adaptation, and heterogeneous healthcare exposure. Effective characterization of these dynamics requires integration of multi-source surveillance data capturing variation across pathogens, antimicrobial classes, geography, clinical context, and time. This study proposes an AI-enabled framework integrating SPIDAAR, ATLAS, GASAR, and PLEA datasets to enable AMR pattern detection, resistance trajectory modeling, and predictive clinical decision making. Current AMR surveillance systems are largely retrospective, fragmented, and descriptive. ATLAS provides large-scale standardized susceptibility data, SPIDAAR contributes high-resolution data from underrepresented settings, reducing geographic bias, GASAR enhances coverage of Gram-negative pathogens and resistance phenotypes and PLEA introduces clinical and contextual variables. However, these datasets remain underutilized in unified predictive systems. This study addresses this gap through a multi-dataset fusion that transforms isolated surveillance into an integrated system. Novelty lies in a Trajectory Modeling AI dual-layer system; an epidemiological forecasting layer using GEO-temporal models and a clinical decision layer generating effective antibiotic susceptibility scores for empiric therapy. Integration with DHIS2, a widely adopted digital health platform by the World Health Organization that has been successfully utilized in infectious disease surveillance, including monitoring and reporting outbreaks such as EBOLA virus disease, will enable real time data collection, case tracking and rapid dissemination of information to the clinical health workers. Overall, this reduces treatment failure, strengthens antimicrobial stewardship by reducing inappropriate antibiotic use, supporting evidence-based prescribing, improving therapeutic precision and strengthening global health resilience against antimicrobial resistance.