Operationalizing wastewater-based surveillance of multidrug-resistant bacteria

Multidrug-resistant organisms (MDRO) pose a significant risk to public health. Infections with MDRO are associated with high mortality rates and healthcare costs, particularly related to hospital-acquired pneumonia. Current approaches to control and prevent transmission of these pathogens focus primarily on clinical testing of infectious patient isolates. This is costly, labor-intensive, and fails to account for asymptomatic carriage. Wastewater testing can overcome many of the limitations posed by patient-based surveillance by enabling cost-effective population-level data acquisition, which can subsequently be used to model and forecast infectious outbreaks. To date, wastewater-based testing has been successfully used for surveillance of pathogenic viruses, but barriers remain in applying this approach to MDRO. While pathogenic bacteria and antibiotic resistance genes (ARGs) have been detected in wastewater treatment plants, several factors currently limit the utility and accuracy of wastewater as a marker for overall burden and diversity of antibiotic resistance. Here, we aim to better operationalize metagenomic wastewater-based epidemiology by understanding the dynamics of multidrug-resistant bacteria during wastewater flow, as well as the relationship between wastewater and clinical detection of MDRO. First, we will design wastewater MDRO model systems by constructing plug-flow reactors and testing the effects of flow parameters such as hydraulic retention time, pH, and temperature, as well as antibiotic pressure, on the prevalence and diversity of MDRO and ARG genotypes. This will account for dynamics in growth rates and potential ARG exchange across species along the wastewater flow, which could significantly affect the accuracy of wastewater-based surveillance models. These bioreactor model systems will enable future experiments testing conditions relevant to specific MDRO species or wastewater streams. In Aim 2, we will take advantage of our ongoing longitudinal wastewater sampling at a major hospital center and the surrounding community to correlate MDRO in wastewater with clinical MDRO and existing patient surveillance cohorts. Through chromatin-linked metagenomics and long- read sequencing we will elucidate phylogenetic links between MDRO in hospital and community wastewater with infectious patient isolates, and potential differences in evolutionary patterns of MDRO in patient versus wastewater collections. Lastly, in Aim 3 we will interrogate different approaches to wastewater-based epidemiological modeling to estimate MDRO burden in a given community. We will contrast linear and nonlinear additive regression models with dynamic mathematical modeling approaches. We will incorporate wastewater flow parameters and community sociodemographics as well as molecular biomarker data, as normalization factors to improve model accuracy. Risk assessment techniques will be applied to these wastewater models to inform development of future public health decision making tools. If successful, the results of this study would enable wastewater surveillance as a tool to inform targeted mitigation strategies to prevent the spread of antibiotic multidrug-resistance.