Predicting COVID-19 Incidence Using Wastewater Surveillance Data, Denmark, October 2021-June 2022

Abstract

Analysis of wastewater is used in many settings for surveillance of SARS-CoV-2, but it remains unclear how well wastewater testing results reflect incidence. Denmark has had an extensive wastewater analysis system that conducts 3 weekly tests in ≈200 sites and has 85% population coverage; the country also offers free SARS-CoV-2 PCR tests to all residents. Using time series analysis for modeling, we found that wastewater data, combined with information on circulating variants and the number of human tests performed, closely fitted the incidence curve of persons testing positive. The results were consistent at a regional level and among a subpopulation of frequently tested healthcare personnel. We used wastewater analysis data to estimate incidence after testing was reduced to a minimum after March 2022. These results imply that data from a large-scale wastewater surveillance system can serve as a good proxy for COVID-19 incidence and for epidemic control.

Keywords: COVID-19; Denmark; SARS-CoV-2; coronavirus disease; epidemiological methods; epidemiological models; epidemiological monitoring; epidemiology; respiratory infections; severe acute respiratory syndrome coronavirus 2; time series; viruses; wastewater surveillance; zoonoses.

Figure 4

Equation for the average number of SARS-CoV-2 RNA copies shed per person living in the WWTP catchment area during a 24-hour sampling period.

Figure 1

Comparison of results of COVID-19 wastewater surveillance with incidence and testing rate (both national and for care personnel) over time, Denmark, illustrating the agreement between wastewater concentrations and incidence. Wastewater concentrations are based on 18,565 individual samples. Testing rates were high during November 2021–February 2022 and decreased after that time.

Figure 2

Model fit and forecasts for the national model (A) and the care personnel model (B) used in study of predicting COVID-19 incidence using wastewater surveillance data, Denmark. The fitted values (orange) are seen to follow the observed incidence (black) during the training period. The forecasts in the test period (purple) are also shown against the observed incidence (black).

Figure 3

Predicted COVID-19 incidence at a constant testing rate (purple) based on the national model, compared with observed incidence (black), in study of wastewater surveillance data as a predictor of COVID-19 incidence, Denmark. The prediction is an estimate of the true incidence. The proportion of estimated true cases captured decreased from >80% to ≈20% during 2022.