Wastewater testing is the future. How should you interpret the data?

Note: This story is more than a week old. Given how quickly the Covid-19 pandemic is evolving, we recommend that you read our latest coverage here.

Burlington's main wastewater treatment plant along the Lake Champlain waterfront. Ten communities in Vermont sample wastewater for Covid-19 and so far four share their data publicly, but officials are still grappling with what exactly to do with the results. File photo by Jim Welch/VTDigger

With daily PCR testing falling to about a third of its Omicron peak and more people using at-home antigen tests, wastewater testing has emerged as a tool that could give officials, and the public, advance warning of an impending coronavirus outbreak.

Ten communities in Vermont sample wastewater for Covid-19 and four currently share their data publicly. Thus far, however, it’s not clear the state knows how best to use the results.

Vermont Health Commissioner Mark Levine said Tuesday that officials are grappling with what exactly to do with wastewater data, including how to convey that data to the public. 

“We're just trying to learn how to aggregate the data appropriately and interpret it and make sure that people are appropriately led by it, as opposed to misled by it,” he said at Gov. Phil Scott’s weekly press conference. “So, more to come.”

Wastewater data can be confusing. Different testing operations use different metrics, and unlike the compiled results of nasal swab testing, the data does not represent a total number of positive cases.

Instead, communities testing wastewater for Covid regularly analyze a small sample of sewage from a treatment plant. The results can show not only whether the virus is present, but how much of the virus is present, allowing communities to track case trends across an entire population over time. 

Burlington has tested wastewater for Covid since August 2020, releasing the genomes-per-liter measurements — a metric of how much virus is in a sample — once a week on the city’s website.

In September 2020, the U.S. Centers for Disease Control and Prevention launched the National Wastewater Surveillance Service, which provides funding for municipalities to send their sewage to third-party labs for testing. Four Vermont municipalities joined the program earlier this year: Brighton, Springfield, St. Albans and Winooski. 

For privacy reasons, the CDC does not release wastewater data from communities under 3,000 people, so Brighton’s test results are not publicly available. The CDC does publish data from the other three locations, showing the percentage change of the viral load in the wastewater over a 15-day period.  

Johnson and Morrisville have begun submitting wastewater for testing but have not yet collected the more than two weeks’ worth of data the CDC uses to track trends, Department of Health spokesperson Ben Truman wrote in an email to VTDigger. Newport City, the village of Essex Junction, and the joint Troy-Jay wastewater facility have signed up for the program but have not started generating data. 

According to Truman, Bennington also wants to join the program, but “recruitment at the NWSS end is temporarily on hold as CDC transitions between contractors.”

Hiccups with testing have caused some issues with that data. As of March 31, two of the three publicly visible sites showed no data at all, and Springfield showed a 100-999% increase. 

Shifts in wastewater Covid levels must be dramatic in order to be meaningful, according to Timothy LaPara, a professor at the University of Minnesota whose research focuses on wastewater microbiology.

LaPara said most microbial data follows what is called Log-normal distribution. As a result, relatively small changes may not reveal significant trends, but changes by factors of 10 typically do.

“Two hundred genomes per liter to 300 — I may or may not worry about that at all. But if it goes from 200 to 2,000, yeah, that's real,” he said. 

Yet the CDC’s data shows only percentage changes, not the raw genomes per liter. By LaPara’s interpretation, changes of 1,000% in either direction would suggest significant changes, whereas smaller percentage changes would not necessarily convey a meaningful trend. 

The most useful wastewater data, LaPara said, comes from the same source using the same testing lab over an extended period of time. “There's lab-to-lab variations, because a lot of these methods are still research methods rather than standard methods,” he said. “So, there are quirks here and there that tend to pop up.” 

If the data is coming from the same source — say, Burlington’s wastewater facilities — one can see trends. But comparing a single day of Burlington’s data to a single day of another city’s data may not be informative.

One benefit of collecting wastewater data is that it tends to predict rising Covid cases before other forms of testing. Exactly how much warning wastewater testing provides is a topic of debate, but LaPara said the method predicts case trends a week to 10 days in the future. 

While wastewater testing may not be widely implemented and understood yet, LaPara believes it will be crucial to the future tracking of Covid and other diseases. 

“Every measurement anybody takes, ever, has some bias to it,” he said. “Wastewater is going to, in my opinion, be the least biased.”

Sign up for our guide to the global coronavirus outbreak and its impact on Vermont, with latest developments delivered to your inbox.


Ethan Weinstein

Send us your thoughts

VTDigger is now accepting letters to the editor. For information about our guidelines, and access to the letter form, please click here.


Recent Stories

Thanks for reporting an error with the story, "Wastewater testing is the future. How should you interpret the data?"
  • Hidden
  • Hidden
  • This field is for validation purposes and should be left unchanged.