Global dynamics of antibiotic resistance: a modeling study of the spread of endemic and emerging resistance in E. coli

BackgroundAntibiotic resistance (ABR) poses a growing threat at the global level. Emergence and worldwide dissemination of multi-resistant clones are frequently reported. However, associated drivers are still largely unknown and are a crucial knowledge to mitigate ABR risk in humans. MethodsWe used mechanistic modeling to analyze annual trends of extended-spectrum {beta}-lactamase-producing Escherichia coli (ESBL-EC) in 39 countries over 2006-2019. The model formalizes ESBL-EC transmission, colonization and infection at the country and global levels, accounting for heterogeneous antibiotic prescriptions and human mobility. The fitted model is used to explore hypothetical scenarios of emerging antibiotic-resistant clones like carbapenem-resistant E. coli (CR-EC) and evaluate their global dissemination risk. FindingsInternational travels alone do not explain heterogeneous ESBL-EC dynamics. Reproducing observed trends requires accounting for antibiotics impact on ESBL-EC acquisition and heterogeneous within-country transmission rates. Strong differences in transmission rates are inferred between Europe and South-East Asia, ranging on average from 7.05x10-3 day-1 in Spain to 16.21x10-3 day-1 in Thailand. For CR-EC, spatiotemporal dynamics depends on the explored emergence scenario. We show that mobility patterns drive 5-years resistance dynamics, while 20-years dynamics are mostly predicted by within-country transmission and antibiotic use. InterpretationThis spatiotemporal analysis highlights the weight of international travel in the early global dissemination of antibiotic-resistant clones but suggests that reducing transmission and selection pressure is fundamental to avoid fixation. There is a need to strengthen surveillance and data collection of community colonization and infection to refine modeling and support decision and ABR control at the global scale. FundingPfizer independent medical grant (number 57504809)