Developing a framework using causal modelling to inform the design and evaluation of antibiotic stewardship and infection control interventions in resource-limited hospital settings.

In many middle-income countries, antibiotics are widely overused in hospitals in part due to limited evidence to inform the design of antimicrobial stewardship (AMS) programmes. Moreover, there is often underinvestment in infection prevention and control (IPC) programmes which have the potential to substantially reduce the burden of antimicrobial resistance (AMR). Whole genome sequencing (WGS) methods have been used extensively in high-income countries to support outbreak investigations and inform infection control interventions. However, corresponding evidence to inform AMS and IPC programmes in middle-income countries, where WGS is becoming affordable, remains scarce. Having conducted a prospective surveillance study on hospital-acquired infections in Thailand, I have firsthand experience witnessing the challenges faced by IPC teams in hospitals. To offer tailored solutions, I will generate multi-species bacterial genomic data and link the data to clinical, microbial phenotype, treatment, and outcome data on hospital-acquired infections. I will then apply existing analysis methods and state-of-art causal inference approaches to the high-dimensional data to develop a new framework to inform AMS and IPC guidelines in middle-income settings with high incidence of antibiotic resistance. Moreover, combining probabilistic bacterial transmission and causal models and using passive surveillance data, I will quantify the expected impacts of implementing different guidelines.