School of Civil and Environmental Engineering

Ph.D. Thesis Defense Announcement

Quantifying the Impact of an Urban, On-site, Shared Sanitation Intervention on

Child Health in Maputo, Mozambique: the MapSan Trial

By

Jacqueline Knee

Advisor:

Dr. Joseph Brown

Committee Members:

Dr. Kostas Konstantinidis (CEE), Dr. Neha Kumar (CC), Dr. Jill Stewart (ESE, UNC-CH), Dr. Jennifer Murphy (WDPB, CDC)

Date & Time: Tuesday, July 16, 2019 at 8:00 AM

Location: SEB 122

Globally, over 50% of the world's urban population, or approximately 2 billion people, lack access to safely managed sanitation. Inequities in access increase in low-resource regions like Sub-Saharan Africa, where over 80% of urban residents lack access to safely managed sanitation. Because development of advanced wastewater infrastructure may be decades away for rapidly growing cities of developing nations, there is increasing interest in decentralized sanitation technologies that serve urban clusters. Though various methods for improved on-site urban sanitation have been piloted, the impacts of such strategies on reducing exposure to enteric (gut) pathogens and subsequent effects on enteric infections have not been characterized.  Here we present results from the Maputo Sanitation (MapSan) trial, the largest controlled health impact study to date of decentralized urban sanitation in the developing world. In this study, we examined the expansion of improved on-site sanitation serving household clusters (compounds) in low-income, densely populated urban neighborhoods of Maputo, Mozambique, where the burden of sanitation-related disease is high. We measured objective health outcomes in children <6 years old before and after the implementation of new sanitation systems which included pour-flush toilets to septic tanks with soak-away pits to discharge aqueous effluent. In parallel, we measured the same outcomes in children enrolled in control compounds that did not receive the new latrines. We used a difference-in-difference (DID) analytical approach, coupled with our controlled before-and-after study design, to estimate the effects of the intervention on three metrics of child health: enteric infection, environmental enteric dysfunction (EED), and reported diarrhea. We collected stool from enrolled children at three time-points during the study: pre-intervention (n=757), 12-months post intervention (n=803), and 24-months post-intervention (n=923). We tested stools for enteric pathogens using a qualitative multiplex molecular method detecting 15 enteric bacteria, protozoa, and viruses identified as important causative agents of diarrhea and four biomarkers of EED using enzyme-linked immunosorbent assays. Enteric infection was common at baseline with over 80% of children in both intervention and control arms positive for one or more pathogen. The sanitation improvements had no effect on enteric infections when we analyzed results from the 12-month and 24-month time-points separately or combined into a single “follow-up” phase. We also found no effect on caregiver-reported diarrhea over the course of the study. Sanitation increased concentration of EED biomarker neopterin, an indicator of immune system activation and inflammation, by 0.17 log10 nmol/L (95% CI: 0.07 - 0.27) at the 12-month time-point and 0.10 log10 nmol/L (95% CI: -0.01, 0.21) at 24-months. Results from sub-group analyses of children with data collected at multiple time-points and children born into the study post-intervention were largely similar to the main analyses. We observed a reduction in risk of Shigella infection by almost 50% (adjusted risk ratio: 0.53, 95% CI: 0.29-0.95) in children born into intervention compounds by the 24-month time-point.  There are many potential reasons the intervention had a limited effect on child health in this setting. It may not have interrupted all transmission pathways of import or limited fecal contamination of the living environment to the extent necessary to observe real changes in exposure. In these densely populated, low-income urban areas, sources of contamination are ubiquitous and our intervention sites may have been impacted by poor WASH conditions in surrounding areas. Delivery of comprehensive WASH interventions at a community level may be necessary to realize real health gains in similar settings.