Wastewater monitoring as a supplementary surveillance tool for capturing SARS-COV-2 community spread. A case study in two Greek municipalities

Abstract

A pilot study was conducted from late October 2020 until mid-April 2021, aiming to examine the association between SARS-CoV-2 RNA concentrations in untreated wastewater and recorded COVID-19 cases in two Greek municipalities. A population of Random Forest and Linear Regression Machine Learning models was trained and evaluated incorporating the concentrations of SARS-CoV-2 RNA in 111 wastewater samples collected from the inlets of two Wastewater Treatment Plants, along with physicochemical parameters of the wastewater influent. The model's predictions were adequately associated with the 7-day cumulative cases with the correlation coefficients (after 5-fold cross validation) ranging from 0.754 to 0.960 while the mean relative errors ranged from 30.42% to 59.46%. Our results provide indications that wastewater-based predictions can be applied in diverse settings and in prolonged time periods, although the accuracy of these predictions may be mitigated. Wastewater-based epidemiology can support and strengthen epidemiological surveillance.

Keywords: COVID-19; Machine learning; RNA; RT-PCR; SARS-CoV-2; Wastewater-based epidemiology (WBE).