The National Fight Against COVID-19 Isn’t Ready To Go To The Sewers – FiveThirtyEight

As COVID-19 testing sites close and experts warn that case numbers are capturing a small minority of infections, many public health experts are turning to a newer source that might tell us what’s going on with the virus: our poop.

In the past two years, scientists have developed systems that can detect COVID-19 in our wastewater. This is a great early warning system, since the virus can show up in people’s waste days before they begin to experience symptoms or are able to get tested. It’s also less biased than case data: Not everyone can find a COVID-19 test and not every positive result will get reported … but everybody poops.

As with so many other COVID-19 metrics, however, interpreting wastewater data is not as simple as it seems. Before COVID-19, this type of data hadn’t been used to track respiratory viruses. This means the Centers for Disease Control and Prevention has little established infrastructure to build upon. The agency is attempting to standardize reporting from researchers across the country, many of whom have different water sampling methods. Plus, the state and local health officials who cite wastewater as a potential replacement for underreported case numbers aren’t used to interpreting data from the environment, which has unique caveats and requires a learning curve for those used to looking at numbers from hospitals and health clinics.

The Documenting COVID-19 project surveyed 19 state and local health agencies, as well as scientists who work on wastewater sampling, to learn about the challenges they’re facing. We found that many states are months away, if not longer, from being able to use wastewater data to guide public health decisions, even as the rise of an omicron subvariant, BA.2, looms. Meanwhile, the CDC’s highly shared wastewater surveillance dashboard is a work in progress, and is difficult to interpret for users who might hope to follow the trends in their areas.


“People are saying, ‘We can’t trust the [PCR] testing data now,’” and shifting to rely more on data from wastewater, said Steve Balogh, research scientist at the Metropolitan Council, a local agency in the Twin Cities, Minnesota, metro area that started wastewater surveillance in late 2020.

But health departments can’t just flick a switch — or gaze into the bottom of a toilet bowl — and suddenly get comprehensive COVID-19 trends from wastewater. It takes time to set up sampling technology, understand the environment around a wastewater site and collect enough data for trends to be easily interpretable.

Some universities and their public health partners in California began investing in wastewater surveillance early in the pandemic. San Diego, for example, began wastewater sampling in fall 2020 as part of the campus reopening efforts for the University of California, San Diego, said Smruthi Karthikeyan, a postdoctoral researcher who works on the surveillance. It was later expanded to other parts of the city, including sites chosen to monitor COVID-19 at local public school districts.

Karthikeyan’s team uses machines called “autosamplers,” which are placed inside a sewage system and programmed to collect small volumes of water over time. These machines slowly collect a set amount of water over the course of 24 hours, which is a more thorough method than grabbing that volume all at once since it captures waste from the entire day. Researchers will then typically take a small portion of the wastewater, dilute it with other chemicals to preserve genetic material in the sample and run COVID-19 PCR tests to determine whether the virus is present. They might also run tests to look for specific variants, like omicron. Karthikeyan’s team uses robots to automate these analysis steps and cut down on errors; other researchers have more human-driven processes.

UCSD now processes about 200 wastewater samples daily, Karthikeyan said. When signs of COVID-19 show up in the wastewater for a particular site, another automated system alerts residents or workers of that site that they should get a PCR test. These alerts help contain outbreaks on UCSD’s campus while minimizing testing costs.

People who receive the alerts “feel like they have a reason to get tested,” Karthikeyan said. “And we get 98 percent compliance when we send out these emails to students.” Other colleges and universities have similarly used wastewater to drive targeted testing.

The San Diego wastewater surveillance network also directs COVID-19 precautions at the UCSD Health system, with tiered guidance based on wastewater data (and a couple of other metrics) telling workers when they need to mask, reduce their in-person work hours or take other precautions. In March, just one week after the health system started using this guidance, wastewater trends prompted a return to mandatory masking.

But many places don’t have the same resources to set up wastewater surveillance — or peg public health actions to data — as San Diego does. When asked if he knew of other institutions using wastewater to guide their safety measures, UCSD Health chief medical officer Dr. Christopher Longhurst said no: “I couldn’t point you to one.”

In recent months, scientists in California have expanded wastewater surveillance to more rural parts of the state, with support from the state’s Department of Public Health. This poses new challenges: Public health officials often aren’t used to looking at data from outside hospitals or health clinics, and there are many logistical hurdles to setting up sampling in new locations, combined with the complications of interpreting data from less populous areas, where wastewater surveillance is highly sensitive to changes in COVID-19 spread.

One document from Stanford University’s Sewer Coronavirus Alert Network (SCAN) describes how the genetic material shed by someone with COVID-19 may change: over the course of their infection, from person to person and depending on measurement techniques, the weather, an influx of spring breakers, or even local business practices. For example, Modesto — a city in California’s Central Valley — had a lot of fruit cannery waste in its sewage. This industrial flow may have blocked the signal of the coronavirus’s genetic material, impacting scientists’ ability to isolate it in PCR testing, said Colleen Naughton, an environmental engineering professor at the University of California, Merced, who works on wastewater monitoring in this region.

The Maine Center for Disease Control and Prevention has faced similar issues as it expands monitoring from Portland, the state’s major urban center, to more rural communities, said Michael Abbott, who leads wastewater screening at the agency.

One of Portland’s wastewater treatment plants has been a long-running collection site for Biobot, a wastewater monitoring company based in Cambridge, Massachusetts. In Portland, a city of almost 70,000, it takes a “really significant increase” in COVID-19 prevalence for the virus levels in sewage to begin ticking up, Abbott said. So when the wastewater goes up, the trend is easy to interpret.

But in rural parts of the state, some of which started monitoring during the omicron surge, “the data tends to bounce up and down more rather than following a fairly smooth curve,” Abbott said. Small numbers of cases may have an outsized impact on wastewater levels, particularly when those cases represent outside tourists going into tiny towns, as is common in Maine.

As public health officials learn to navigate wastewater data, several states’ public health agencies told us they don’t yet consider the system a reliable source for making policy decisions — at least not in isolation. A recent report from The Rockefeller Foundation found that many local agencies, especially those serving rural areas, don’t have the internal capacity to go all-in on wastewater data as they deal with budget cuts and burnout.

The Minnesota Department of Health is working to expand wastewater sampling, agency spokesperson Garry Bowman said in an email. But this expansion would take weeks or months to turn into useful data, long after Minnesota’s PCR testing numbers have become wholly unreliable. And even if Minnesota does expand its wastewater monitoring to cover the whole state, the agency isn’t sure if the federal government will keep up its funding, said Bowman.

It would be hard to tell these local wastewater challenges exist, though, looking at the CDC’s wastewater dashboard. Highly publicized upon its addition to the agency’s COVID Data Tracker in February, the dashboard only showed one metric until early April: colored dots representing an increase or decrease in coronavirus levels detected at each site over the past two weeks. There was no context about the virus’s actual prevalence or how recent trends compare to longer time frames.

If a site measures “virus not detected” (meaning no COVID-19) for three weeks in a row, and then measures a fairly low level of COVID-19, the CDC dashboard would show a 100 percent increase, said Zuzana Bohrerova, an environmental scientist at Ohio State University who works on Ohio’s monitoring program. A red dot on the dashboard might be a serious warning, or it might be unimportant — the CDC’s original presentation made it difficult to tell.

“I think they were trying to be simpler,” Naughton said of the CDC dashboard. “They didn’t want to release all the concentration data, since they thought that’s difficult for people to understand.”

On April 8, the agency updated this dashboard, adding new metrics and the option to click into a specific site for a chart showing wastewater trends at that location. The updates were intended to help users better understand changes in virus levels at different sites, CDC spokesperson Brian Katzowitz said. Wastewater experts who talked to me on Twitter said they were glad to see the CDC adjusting its data visualization. Still, there’s room for other improvements that would make the data more easily understandable, Katzowitz said.

Even if thousands of new wastewater collection sites are added to the U.S.’s network in the coming weeks, it will take serious investment in data analysis and communication for the country to actually use these numbers in predicting new surges. Expanding wastewater sampling can take weeks or months before the effort begins to produce usable data and it costs a lot — money agencies aren’t sure the federal government will continue to provide.

Despite its challenges, experts say that wastewater monitoring has potential beyond COVID-19 — for other respiratory viruses, antimicrobial resistance, and even identifying entirely new viruses. But without continued federal support, we won’t even be able to get useful COVID-19 surveillance at a time that it’s deeply needed.


This story was published in partnership with the Documenting COVID-19 project, which is supported by Columbia University’s Brown Institute for Media Innovation and MuckRock. The project collects and shares government documents related to the COVID-19 pandemic and works on investigative journalism projects with partner newsrooms.