COVID-19 wastewater-based epidemiology back calculation using hybrid modelling methods
Following the first wave of the COVID‑19 outbreak in Canada, provincial authorities are currently relaxing lockdown restrictions of various institutions and types of social gathering. Policies to mitigate the spread of COVID‑19 are based on the state and forecast of SARS-CoV-2 infection prevalence, using various models and more or less granular population data. Current models mostly exploit COVID‑19 clinical test results, hospital admissions, and deaths. However, the COVID‑19 clinical test results only react with a delay of two to five days after a patient is infected and contagious.
The proposed project aims at complementing existing COVID‑19 infection models through the use of data collected from wastewater analysis. This technique has historically been effective in detecting different kinds of pathogens and recently proven to concur with COVID‑19 population test results and precede those tests by as much as 7 days.
In order to address current challenges posed by the usage at scale and in real-time of wastewater-based epidemiology (WBE) for COVID-19, this project proposes a research agenda that will lead to progress in the interpretation of viral signal data in the context of WBE. The proposed approach, based on a hybrid between current phenomenological WBE fate models and time series Machine Learning (ML) models is aimed at exploiting both publically available highly granular clinical testing results from the Canadian health institutions and other publically available information such as sewershed configurations, meteorological information, influent wastewater flows and compositions.
We expect this project will lead to advances in WBE and the standardization of its protocol, which in turn will lead to the generation of federated models at the Canadian level, in order to provide a global, but granular forecasting model of a pandemic.