What’s standing in the way of wastewater data becoming a more mainstream public health tool

When case counts suggest Covid-19 is in check, the truth is in the toilets.

The pandemic has long been in its fatigue phase, making it hard for public health departments to have the return on investment to keep testing clinics open — and making it all but impossible to know, in real time, how many people in a certain area are sick. And while hospitalizations are fairly reliable, that indicator lags behind cases. So why not switch to the more-reliable, lower-effort, lower-cost wastewater testing to keep tabs on population-level cases?

“When we think about where this could be in three to five years, we would love to see wastewater be part of kind of a health weather alert,” said Amy Kirby, leader of the Centers for Disease Control and Prevention’s National Wastewater Surveillance System (NWSS) program. “You wake up and check your health weather alert, and say, ‘In my community, are we seeing increases in Covid or is there a norovirus outbreak going on? Is there something else that I need to be more careful about?’”

While the national wastewater surveillance infrastructure built to detect SARS-CoV-2 has made this kind of application more realistic, there are still several hurdles to clear before a future “health weather alert” system is possible.

Public health officials remain cautious

Every toilet flush contains information about a community’s health, whether that’s influenza spread, opioid use, or the prevalence of antibiotic-resistant bacteria. But the idea didn’t go mainstream in the U.S. until 2020, when the CDC established NWSS (pronounced “news”). NWSS aggregates wastewater data from federal contractors, academic labs, state labs, and state-contracted commercial labs, displaying the results on its publicly available wastewater metric chart.

Mariana Matus, co-founder and CEO of wastewater analysis company Biobot, said that Biobot’s local and state government partners like wastewater testing because it is reliable; how many infected people the testing encompasses doesn’t depend on people getting tested for Covid.

But public health officials are cautious about working with new kinds of epidemiologic data. This creates a “chicken and egg problem,” said Megan Diamond, who leads wastewater surveillance efforts as the director for global health partnerships at the Rockefeller Foundation. “People want to see the results of the data before they’re comfortable using or investing in them,” she said, “but in order to get the data, you need to invest in [getting the data].”

NWSS has been able to increase adoption of wastewater monitoring via their contract with companies like Biobot, which manages over 400 of the roughly 1,250 NWSS reporting sites, said Kirby. “What we have found will convince people of the utility of wastewater the quickest is actually seeing it work in their community,” she said. “At no cost to your community, you can try out wastewater surveillance and see if it works for you. We’ll do all the heavy lift; we’re just going to hand you data.”

Still, there’s a question of whether communities can figure out what to do with that data. While cities like Boston use local wastewater data to decide whether to enact mitigations like school masking recommendations, Kirby said that some municipalities end up discounting wastewater data because it’s hard to reconcile with more traditional Covid metrics.

Wastewater data is different than conventional public health data

The area served by a given sewer system cuts across ZIP codes, census tracts, and counties, making the data incompatible with metrics like case counts, which are often reported according to those more common geographic categories. Because sewer system maps aren’t readily available, databases with the information needed to translate sewer system maps to categories like ZIP codes don’t exist. One of NWSS’ goals is to compile that data — which has to be done by contacting each wastewater treatment plant and asking about their historical system map — to make that data translation easier, said Kirby.

The fact that not everyone is on a sewer system further complicates the effort. In rural areas, more people have their own septic tanks and would be thus excluded from wastewater data, though it’s unlikely this creates a bigger data gap than current clinical testing oversights. However, Aparna Keshaviah, a principal researcher at Mathematica who directs the organization’s wastewater research, countered that some people who are on septic may be accounted for in their workplaces’ data, if that building is serviced by sewers.

In fact, rural and smaller sewer systems that serve tens of thousands of people have advantages over large, metropolitan systems that may serve a million or more people, said Keshaviah. The smaller the population using the sewer that leads to the collection site, the more granular the data. Multiple universities monitored wastewater for individual dormitories to mitigate Covid spread earlier in the pandemic.

Wastewater data is also hard to interpret without other context. Just because one town has more viral copies per liter of wastewater than another doesn’t mean it has more cases.  “It turns out the way the wastewater system is built and operated has a huge impact on the number that you get at the plant,” Kirby said. To try to make comparisons across the country when the number of viral RNA copies isn’t comparable across different sites, the NWSS dashboard instead reports how many sites have seen a certain amount of change in virus levels in the past 15 days. That metric can be difficult to make useful.

The wastewater metric NWSS wants to be able to provide is an estimate of case level from sewage data: how much viral RNA corresponds to one person with Covid. The current models for that number are so imprecise that the range is too huge to be useful, said Kirby. Producing a more precise estimate will require researchers to get a handle on three factors that are difficult to pin down, she said. “For each person that’s infected with SARS-CoV-2, how much virus do we expect them to contribute to the wastewater, and how long will that last, and how much of it will be detected at the treatment plant?” she said.

Another flaw in the current NWSS system is that the numbers will fluctuate depending on who handles the samples and analyzes the data, said Matus. It’s hard for the NWSS system to compare trends across different sites because many different vendors send data into the database. However, samples analyzed and processed by the same lab with the same methods have good internal consistency, which is why comparisons like Biobot’s regional trend comparisons are possible, she said.

However, the noise in the data is still too high to be a reliable leading indicator for Covid surges. That’s why Keshaviah and Diamond’s teams at Mathematica and the Rockefeller Foundation developed an alert algorithm that evaluates several different kinds of wastewater metrics to determine when a true surge is happening and a framework for how communities can integrate traditional metrics like hospitalizations and clinical case counts with wastewater data.

Long-term investment could prove difficult to get

As with any pandemic preparedness tool, it is difficult to take the momentum that Covid-19 gave wastewater monitoring and sustain investment in its development.

“Communities and societies are much more likely to take corrective action than preventive action,” said Keshaviah.

In theory, it shouldn’t be a hard sell. While clinical testing is still important for individual-level decisions, on a population level, wastewater monitoring is much more affordable and comprehensive.

Anne Zink, president of the Association of State and Territorial Health Officials and Alaska’s chief medical officer, told the New Yorker that the state is investing in wastewater testing because it “doesn’t have the money or political backing to set up daily nasal-swab testing sites” and the practice is just more sustainable. Technicians use the same PCR machines and reagents for both wastewater and clinical testing, but wastewater testing can cover thousands of people in a single sample, which can cost a community anywhere from $300 to $1,200, according to Keshaviah.

Wastewater surveillance also has uses beyond monitoring for infectious pathogens like SARS-CoV-2, mpox, or polio, and the Covid-expanded infrastructure might bring on funding for other wastewater monitoring programs.

Before the pandemic, several researchers were testing wastewater data as a way to monitor for opioid use. But experts said they’re worried that without a streamlined effort to bolster research using wastewater analysis more broadly, all the progress that’s been made during the pandemic will end up wasted. Researchers are watching multiple departments from the same area ask understaffed wastewater treatment plants to run separate programs for infectious disease and for opioid use, which is twice as much work for data that could be collected from a single sample if there were more coordination.

Tracking cryptic Covid strains individual to one person or cataloging SARS-CoV-2 variants through airplane wastewater show the creative use cases for the method, but wastewater surveillance is an important tool for keeping schools, communities, and populations healthy, said Diamond.

“It allows us to do public health with our blinders off and really is the future of public health surveillance,” said Diamond. “[I] just hope to see more investment from government and industry alike into this field.”

Get your daily dose of health and medicine every weekday with STAT’s free newsletter Morning Rounds. Sign up here.