Innovative Strategies for the Implementation of Sewage Surveillance in Dhaka's Converging Informal Sewage Network

Abstract

Introduction: During the COVID-19 pandemic, sewage surveillance (SS) was pivotal in monitoring disease transmission on a community level and complemented clinical surveillance, especially when case data was limited or lacking altogether. High-income countries possess resources and formal sewage networks to facilitate robust SS. In contrast, low and middle-income countries settings like Dhaka Bangladesh, lack formal sewage systems and the means to implement SS. Since June 2019, we devised strategies to bridge resource gaps to implement SS for infectious diseases in Dhaka. Methods: Mirpur wards 8, 9 and 10, an 8.67 km2 area, were selected to implement SS for enteric infectious diseases including polio, cholera, typhoid, rotavirus, and shigellosis. Ground surveys were conducted to identify all sewage network lines. Digitized maps, population density, and digital elevation models were integrated to estimate watershed areas and catchments. Household demographic surveillance system was established to identify children <5 years old for a supplementary immunization activity (SIA) with Sabin bivalent oral polio vaccine (bOPV). Twelve SS sites with catchment population ranging from 1000-600,000 were selected based on accessibility, flow, catchment, and physicochemical properties of the sewage. We evaluated site performance through testing if Sabin poliovirus was detected post-SIA. Six-liter grab samples were collected pre- and post-SIA, followed by molecular detection of Sabin virus 1 and 3. In March 2020 when the COVID-19 pandemic hit, we expanded SS to 51 sites, and quickly adapted existing methods to detect SARS-CoV-2. Results: Mapping the sewage network and implementing HDSS to estimate the associated population took two months. In July 2019, 92% of children under 5 were vaccinated over a 10-day period. Although 11/12 sites detected Sabin viruses, nine sites accurately captured the spike of Sabin viruses after the SIA. In univariable analyses, total dissolved solids (mg/L), sewage temperature, and the 21-day period following the bOPV campaign showed a positive association with enterovirus detection with a mean pH of 7.2. In the SS for COVID-19, Pearson correlation analysis revealed a strong association between sewage data and geolocated clinical cases (p<0.0001) which indicated that SS acted as an early warning system and tracked SARS-CoV-2 trends 1-2 weeks before community infection waves. Discussion: To implement successful SS accurate sewage mapping and understanding population catchment were critical. Other important “good” site characteristics were the year-round flow, high TDS, neutral pH, and sites not impacted by flooding. Conclusion: Despite the challenges, we successfully implemented SS for polio and SARS-CoV-2 in a low resourced area with informal sewage systems. While SS has proven effective in detecting polio virus and SARS-CoV-2, further work is required to see whether SS is equally effective for detecting other infectious diseases such as influenza, respiratory syncytial virus, and cholera, and its correlation with community-level transmission of the diseases.