Separating signal from noise in wastewater data: An algorithm to identify community-level COVID-19 surges in real time

Abstract

Wastewater monitoring has provided health officials with early warnings for new COVID-19 outbreaks, but to date, no approach has been validated to distinguish signal (sustained surges) from noise (background variability) in wastewater data to alert officials to the need for heightened public health response. We analyzed 62 wk of data from 19 sites participating in the North Carolina Wastewater Monitoring Network to characterize wastewater metrics around the Delta and Omicron surges. We found that wastewater data identified outbreaks 4 to 5 d before case data (reported on the earlier of the symptom start date or test collection date), on average. At most sites, correlations between wastewater and case data were similar regardless of how wastewater concentrations were normalized and whether calculated with county-level or sewershed-level cases, suggesting that officials may not need to geospatially align case data with sewershed boundaries to gain insights into disease transmission. Although wastewater trend lines captured clear differences in the Delta versus Omicron surge trajectories, no single wastewater metric (detectability, percent change, or flow-population normalized viral concentrations) reliably signaled when these surges started. After iteratively examining different combinations of these three metrics, we developed the Covid-SURGE (Signaling Unprecedented Rises in Groupwide Exposure) algorithm, which identifies unprecedented signals in the wastewater data. With a true positive rate of 82%, a false positive rate of 7%, and strong performance during both surges and in small and large sites, our algorithm provides public health officials with an automated way to flag community-level COVID-19 surges in real time.

Estimating Rates of Change to Interpret Quantitative Wastewater Surveillance of Disease Trends

“…Greater transparency would be afforded by explicitly specifying the threshold above which rates of change would be considered meaningful, as when Keshaviah et al specified the doubling of wastewater viral quantities as a component of an algorithm to detect infection surges. 36 As of December 2023, CDC NWSS similarly published fixed categories defined by the percent change in viral load over the previous 15 days to classify trends, with changes  100% considered "large" increases. 52,53 Both quantitative and classification accuracy differed between approaches, but the differences were generally observed only for smoother trends.…”

“…First Derivatives of Smooth Functions of Time Substantial fluctuations over short timescales in both measured wastewater viral loads and reported infections have motivated the use of a variety of smoothing approaches to better characterize infectious disease trends from noisy surveillance data. 20,26,36 Many common smoothing techniques, including simple moving averages and locally weighted scatterplot smoothing (LOESS), use the values of neighboring observations within a user-defined window to estimate smoothed values. 17,18,20,37 Such techniques are entirely data-dependent and do not have simple mathematical representations like those from the methods previously presented.…”

From pandemic to endemic: Divergence of COVID-19 positive-tests and hospitalization numbers from SARS-CoV-2 RNA levels in wastewater of Rochester, Minnesota

“…Due to the limiting factors of our dataset, it was not possible to determine the exact length of the anticipation afforded by the wastewater data. However, other studies have reported that wastewater data precede clinical case detection by 0–14 days [ 39 , [67] , [68] , [69] , [70] , [71] ]. This broad range suggests that the anticipation period is dynamic over time and space [ 46 , 70 ] and is affected by multiple factors including the strains in circulation and the vaccine coverage [ 71 ].…”

Public health policy impact evaluation: A potential use case for longitudinal monitoring of viruses in wastewater at small geographic scales

“…Studies suggest that virus shedding patterns differ among SARS-CoV-2 variants [76][77][78][79], so changes in SARS-CoV-2 variants over time could be another reason for the change in correlation over time. We also did not consider lead-lag time effects between wastewater monitoring and case surveillance data as done in other studies [16,80], so future work could investigate how lead-lag time effects between wastewater monitoring and case surveillance data have changed over the course of the pandemic. Nonetheless, wastewater monitoring data are independent of test-seeking behaviors or test reporting patterns so may be a less biased tool for monitoring public health, particularly in periods characterized by low test-seeking and reporting rates.…”