The increasing resistance to antibiotics is proved to be a public health threat, as antibiotics are the only widely accessible and effective treatment against bacterial infections. While their extensive use is undoubtedly the main driver of this phenomenon, it does not adequately explain the antimicrobial resistance levels in different populations. Exploration of additional factors is required, but it is challenging to acquire experimental or empirical data of relevance. Simulation studies can thus facilitate the delimitation of critical drivers to guide future research.
The majority of previous modelling studies of antimicrobial resistance are limited to one level of system complexity, and often lack the integration of multiple drivers. Our expanded modelling approach combines within-host, between-hosts and between-population dynamics. As such, it allows us to explore how the prevalence of resistance is affected by the combination of three factors, namely antibiotic use, cost of resistance and bacterial transmission.
The model is built upon the competitive Lotka-Volterra equations. Each host may carry two strains (susceptible and resistant) that compete in a batch culture experiment. The strains mimic the host’s commensal microbiome and are not the target of the antimicrobial treatment. The model simulates a population of hosts, who are treated periodically with antibiotics and transmit bacteria to each other.
We propose that the ever-changing environment of periodical antibiotic treatment promotes the coexistence of susceptible and resistant strains (temporal storage effect). This coexistence is reinforced by bacterial transmission, which usually increases resistance levels in a host population. We suggest that transmission can also decrease resistance when the resistance cost is low, through the replenishment of susceptible populations (bacterial metapopulation rescue). Transmission between host populations leads to them having more similar resistance levels, surprisingly, by aiding the population that receives more frequent antimicrobials to reduce its resistance.
Overall, through the simulations, we make testable experimental predictions and identify key areas for future antimicrobial resistance research.
Amalia Bogri’s presentation will be uploaded after the manuscript is accepted for publishing.