Spatial and temporal variability and data bias in wastewater surveillance of SARS-CoV-2 in a sewer system

Abstract

The response to disease outbreaks, such as SARS-CoV-2, can be constrained by a limited ability to measure disease prevalence early at a localized level. Wastewater based epidemiology is a powerful tool identifying disease spread from pooled community sewer networks or at influent to wastewater treatment plants. However, this approach is often not applied at a granular level that permits detection of local hot spots. This study examines the spatial patterns of SARS-CoV-2 in sewage through a spatial sampling strategy across neighborhood-scale sewershed catchments. Sampling was conducted across the Reno-Sparks metropolitan area from November to mid-December of 2020. This research utilized local spatial autocorrelation tests to identify the evolution of statistically significant neighborhood hot spots in sewershed sub-catchments that were identified to lead waves of infection, with adjacent neighborhoods observed to lag with increasing viral RNA concentrations over subsequent dates. The correlations between the sub-catchments over the sampling period were also characterized using principal component analysis. Results identified distinct time series patterns, with sewersheds in the urban center, outlying suburban areas, and outlying urbanized districts generally following unique trends over the sampling period. Several demographic parameters were identified as having important gradients across these areas, namely population density, poverty levels, household income, and age. These results provide a more strategic approach to identify disease outbreaks at the neighborhood level and characterized how sampling site selection could be designed based on the spatial and demographic characteristics of neighborhoods.