Factors of Full-Scale Drinking Water Treatment that Contribute to Risk of Opportunistic Infectious Disease

Each year, 16,000 hospitalizations and 3,000 deaths caused by non-tuberculous mycobacteria occur in the USA. Non-tuberculous mycobacterial infections are primarily transmitted by water. This work helps to explain why mycobacteria are present in the finished water produced by drinking water treatment plants, and identifies treatment plant design and operational considerations to minimize the risk of waterborne mycobacterial infection. It was found that viable microbial cell concentrations decreased significantly in the first ozone contact chamber of multi-chamber ozone contactors in a full-scale drinking water treatment plant. However, cell concentrations rose in subsequent chambers across the contactors. This increase resulted from detachment from biofilms on contactor surfaces and from biomass in sediments within a hydraulic dead zone. The biofilms downstream of the dead zone contained a significantly higher relative abundance of mycobacteria than biofilms in earlier chambers. Viable mycobacteria populations were detected in ozone contactor effluents and in biologically-active carbon (BAC) filters downstream. These findings present an argument for improved hydraulic efficiency in multi-chamber contactors, e.g., through decreasing chamber width, and underscore the importance of filter maintenance practices that avoid reinforcing the presence of mycobacteria. During BAC filter backwashing with water containing monochloramine, monochloramine concentrations at the top of filter beds were measured to be similar to concentrations previously shown to be sub-lethal for the mycobacterial species Mycobacterium avium. M. avium dislodged during filter backwashing upregulated mammalian cell entry gene mce1C substantially. M. avium’s response to disinfectant exposure during backwashing raises the possibility of an adaptive response that increases its resistance to disinfection. Its upregulation of mammalian cell entry genes has implications for its role as an intracellular pathogen, and is consistent with laboratory scale findings that sub-lethal monochloramine exposure induces mycobacterial virulence factors. Where possible, utilities should limit the use of backwash water with disinfectant to reduce selecting for disinfectant-resistant bacteria. Mycobacteria can resist residual disinfectant in distribution systems and, therefore, distribution systems represent a transmission route of mycobacteria in treated waters to consumers. Despite higher concentrations of mycobacteria reported in water disinfected with chloramine compared to chlorine, the type of disinfectant used to provide a residual disinfectant during distribution of water was not a significant predictor of infection in a case control study of patients at the University of Michigan’s academic medical center Michigan Medicine. However, the use of drinking water primarily sourced from a surface water body compared to a groundwater source was found to be significantly associated with infection. This result suggests that monitoring mycobacteria in source waters used for drinking water production may be a proactive strategy that can be implemented by drinking water treatment plants. The results presented in this dissertation suggest that specific drinking water treatment plant design and operational considerations minimize risk of waterborne infectious disease. This work should assist water system managers to improve the microbial quality of drinking water to promote health for everyone, including individuals sensitive to opportunistic bacterial infections.PHDEnvironmental EngineeringUniversity of Michigan, Horace H. Rackham School of Graduate Studieshttps://deepblue.lib.umich.edu/bitstream/2027.42/138669/1/kotlarz_1.pd