Novel Wastewater Toolkit Flags Covid-19 Surges Using Multiple Data Sources

Wastewater monitoring has proven to be an effective tool in tracking Covid-19 trends and anticipating impending outbreaks. However, scientists have grappled with challenges in how to interpret and translate wastewater surveillance data to inform public health responses, restricting policymaker response. To address this need, researchers from Mathematica, funded by The Rockefeller Foundation, developed a promising new toolkit called Covid-SURGE which supports early warnings of Covid-19 surges at community level.

Designed for local public health officials and wastewater enthusiasts, The Covid-SURGE (Signaling Unprecedented Rises in Group Wide Exposure) toolkits has two components: an algorithm that can flag community Covid-19 surges and a risk tool estimator that synthesizes wastewater data with traditional public health surveillance metrics into a composite risk score. The evidence to support the tool development was recently published in Proceedings of the National Academy of Sciences (PNAS) journal by Keshaviah et al., after an in-depth analysis of 1,800 wastewater samples collected over 62 weeks from across North Carolina, and further validation using data from 7 other states.

Covid-SURGE Algorithm

The Covid-SURGE algorithm addresses a major question in the field – which metric should policymakers utilize to identify an impending surge? While there are many approaches used by different entities – the proportion of samples with detectable levels of the virus, the percent change in viral concentrations, and the absolute magnitude of viral concentrations – Mathematica’s analysis indicates that the best value is when they are used together. By using a composite metric that looks at recent rises in the context of broader historical trends, the algorithm can spot significant shifts in community risk profiles–for instance, a substantial and sustained increase in viral concentrations in wastewater that could indicate an imminent outbreak.

Crucially, the Covid-SURGE algorithm activates the early warning potential of wastewater data, showing in real-time when wastewater data signal a need for heightened public health response. Evidence from deploying the algorithm in North Carolina shows that it can identify a surge 4-5 days before clinical counts begin to rise. This capability is especially valuable during periods when traditional testing may be less frequent, but the risk of transmission could be higher.

Dr. Petros Chigwechokha of the Malawi University of Science and Technology with Dr. Rochelle Holm of the University of Louisville working on multi-pathogen wastewater surveillance.

While the algorithm was optimized for North Carolina’s data, validation analyses indicate that it should perform well in other states and to flag Covid-19 surges caused by different variants. The algorithm is freely available as an excel calculator that enables users to adjust the algorithm’s thresholds to their local context, which may be needed in tourist areas, college towns, or other communities with high population fluxes, or when wastewater sampling occurs less frequently than twice a week.

Worker taking water from the wastewater treatment pond to check the quality of the water.

Multi-Indicator Risk Score

Another tool that closes the gap between metrics we have and metrics we need for pandemic response is the Covid-SURGE Multi-Indicator Risk Score, which addresses the well-recognized fact that the value of wastewater surveillance data is heightened when examined in the context of other data sources.

Because each type of data provides information at a different cadence, and on potentially different segments of the population, a key challenge is how to integrate the data and facilitate decision making.

By standardizing, visualizing, and synthesizing a range of Covid-19 indicators–including case counts, hospitalizations, deaths, health care utilization, and wastewater data–the estimator offers insights into how the Covid-19 risk picture varies across different tracking methods, and the added value of each dataset to provide a more complete understanding of the risk level in a community.

The risk estimator has a few distinct features. First, it provides a composite risk score at the county level constructed by weighting and summing the different indicators mentioned above. The composite score ranges from 0 to 100, with 100 representing very high risk. Second, the risk estimator displays trends over time in wastewater viral concentrations alongside trends in other measures. This context helps public health officials understand how the risk picture is changing, what the future risk might look like, and if trends are similar across different data sources. Building off this, the estimator classifies risk directionality to highlight if risk is increasing, decreasing, or stable based on recent trends in the data. Lastly, the risk estimator has an alert function that indicates when risk is becoming significantly elevated compared to historical trends, which could signal the need for additional public health interventions. Showing how trends and alerts vary across different tracking methods, and when they align, gives officials more confidence to act or to take a watch and wait approach.

These features have been developed for public health officials without deep technical expertise, making wastewater data an accessible source for local health departments that have variable capacity. The interactive dashboard has intuitive data visualization, a user-friendly interface and updates in real-time — designed for every day, ongoing use.

Future Applications

 The Covid-SURGE toolkit, which includes the algorithm and risk estimator, represents a significant step forward in public health surveillance. Through rigorous analysis, it provides mechanisms for interpretation and use of wastewater data for public health action, which has been recognized as an ongoing challenge in the field. The alert algorithm is most useful during periods of low transmission, to provide an early warning for new upticks, while the risk estimator may be most useful to assess the extent of spread during an outbreak and determine when all data sources point to waning threat levels.

The significance of these findings are not only applicable for the current Covid-19 pandemic but could also be adapted for other health threats. For example, the logical criteria used to create the Covid-SURGE algorithm could potentially be fine-tuned to develop alert algorithms for flu, polio, or other public health biomarkers being monitored in wastewater.

As the focus on the pandemic wanes with the ending of Covid-19 emergency declarations, the Covid-SURGE algorithm and risk estimator offer public health officials a valuable way to stay vigilant. They provide a timely way to assess when renewed attention and action may be needed, helping to mitigate the impact of future outbreaks, Covid-19 or otherwise.