Guiding community-based antimicrobial resistance control in low-resource settings: attribution of household sources of resistant bacteria and the impact of improved antibiotic use

In low-resource settings, bacterial antimicrobial resistance (AMR) is exacerbated by limited access to safe water, sanitation and hygiene, resulting in frequent community-acquired colonization of gut and skin with AMR bacteria, e.g. extended-spectrum ß-lactamase (ESBL) Enterobacteriaceae.Community-based interventions should target proven sources of transmission, and improve antibiotic use, to limit selective pressure.Adapting an existing source attribution model, I will attribute risk of colonization with AMR bacteria to different sources in/around households, by comparing resistance genes from sequenced isolates (and pooled metagenomes) in human stool, with those in environment sources in a population cohort in Burkina Faso. A mediation analysis will examine the role of behavior, age, and socio-economics, as mediators of exposure to bacterial sources, on the AMR bacterial colonization rate.To estimate the impact of community-based antibiotic use interventions on the AMR burden, first, I analyze the effect of individual and combined intervention components on antibiotic/patient care outcomes, exploring effectiveness measures, using data of an intervention bundle in Burkina Faso and DR Congo. Then, intervention-dependent community antibiotic use patterns can be fitted in a bacterial transmission model, to infer the impact of an intervention, changing antibiotic use or care, through changes in AMR colonization on the incidence of AMR bloodstream infections.