• Davidon Jani
• Alice Chimhondoro
• Senzenia Chakauya

Antimicrobial resistance (AMR) is a global health concern that requires a paradigm change from reactive surveillance to proactive, predictive forecasting. The intricate, adaptive dynamics of AMR are not captured by traditional epidemiological models, which are frequently limited to a single scale (within-host or between-host) and calibrated statically. They are therefore unable to take advantage of real-time data streams. Within a Bayesian data assimilation paradigm, we introduce a novel framework that combines structured between-host transmission dynamics with within-host pharmacokinetic-pharmacodynamic models of bacterial development. To continually calibrate the model and generate probabilistic forecasts with quantifiable uncertainty, an Ensemble Kalman Filter dynamically assimilates diverse surveillance data. Our methodology correctly predicted cases four weeks in advance (MAPE < 15%) and identified asymptomatic carriers in a high-fidelity simulation of carbapenem-resistant Enterobacteriaceae. Infection control methods are 45% more effective than stewardship, according to digital twin trials, making this strategy a potent tool for proactive AMR management.

Antimicrobial Resistance, Multi-Scale Modelling, Data Assimilation, Ensemble Kalman Filter, Bayesian Inference, Pharmacokinetic-Pharmacodynamic Modelling.

" A MULTI-SCALE, DATA-ASSIMILATIVE FORECASTING FRAMEWORK FOR ANTIMICROBIAL RESISTANCE (AMR)", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.12, Issue 11, page no.b593-b607, November-2025, Available :http://www.jetir.org/papers/JETIR2511176.pdf

" A MULTI-SCALE, DATA-ASSIMILATIVE FORECASTING FRAMEWORK FOR ANTIMICROBIAL RESISTANCE (AMR)", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.12, Issue 11, page no. ppb593-b607, November-2025, Available at : http://www.jetir.org/papers/JETIR2511176.pdf