Global Prediction of Antimicrobial Resistance Trends Using Statistical and Machine Learning Models: Evaluating National Action Plan Policy Impacts Through Interrupted Time Series Analysis

The study presents a data-driven global analysis of antimicrobial resistance (AMR), using statistical and machine learning models to forecast AMR trends and evaluate the effects of national action plans on resistance patterns. Clinical isolate data from 65 countries were analysed, revealing that AMR burdens are highest in low- and middle-income regions, particularly in Southeast Asia and Africa, with older adults and males showing disproportionately higher resistance levels. Forecasting models (LSTM, SARIMA, and hybrid) project continued rises in resistance through 2030, while interrupted time series analysis suggests that existing national action plans have not yet produced significant global reductions in resistance. The findings underscore the need for tailored, region-specific strategies and robust predictive frameworks to inform future AMR interventions