Predicting sepsis mortality into an era of pandrug-resistant E. coli through modeling

Abstract

Background: Infections caused by antibiotic-resistant bacteria are increasingly frequent, burdening healthcare systems worldwide. As pathogens acquire resistance to all known antibiotics – i.e., become pan-resistant – treatment of the associated infections will become exceedingly difficult. We hypothesized that the emergence of pan-resistant bacterial pathogens will result in a sharp increase in human mortality. Methods: We tested this hypothesis by modeling the impact of a single hypothetical pan-resistant Escherichia coli strain on sepsis deaths in the United States. We used long-term data on sepsis incidence, mortality rates, strain dynamics, and treatment outcomes to parameterize a set of models encompassing a range of plausible future scenarios. All models accounted for historical and projected temporal changes in population size and age distribution. Results: The models suggest that sepsis deaths could increase 18- to 46-fold within 5 years of the emergence of a single pan-resistant E. coli strain. This large and rapid change contrasts sharply with the current expectation of gradual change under continuing multidrug-resistance. Conclusions: Failure to prevent the emergence of pan-resistance would have dire consequences for public health.