What's really going on with bird flu wastewater data

During COVID, public health experts and armchair epidemiologists became obsessed with poop — or, more accurately, the secrets it held about the state of the pandemic. Wastewater, which can measure how much virus humans excrete, has become a valuable disease-tracking tool.

But using that tool to track our current viral threat, the H5N1 bird flu that has begun circulating in cows, is much trickier. Finding even high levels of bird flu in wastewater does not necessarily mean an area is experiencing an outbreak. Rather, it sets off a hunt for the source.

That’s still valuable information, particularly amid the paucity of testing that has made it nearly impossible to gauge the state of the outbreak. But it’s also leading to some confusion and even alarmist interpretations.

Although scientists had been mining wastewater as a public health tool for years, the approach really came into its own during the COVID pandemic, becoming a reliable sentinel of an outbreak. Although the Centers for Disease Control and Prevention regularly published testing and hospitalization numbers, those were lagging indicators. The wastewater numbers tended to tick up in time to serve as a warning to the wary.

When it comes to bird flu, though, the information offered by wastewater is much murkier.

The confusion began when wastewater showed an uptick of influenza A, a broad category that includes the H5N1 strain, in parts of the country at the same time bird flu came onto the public’s radar.

That stoked fears that, rather than an unseasonable bump in garden variety flu (or a quirk in the way CDC was measuring its data), H5N1 was quietly running rampant. But that confusion didn’t abate once wastewater tests were refined to specifically pluck out H5, a narrower category that includes the strain circulating in cows.

Whereas COVID was concentrated in humans, bird flu by now is circulating in an assortment of species. Researchers currently can’t decipher whether a signal is coming from an infected human, a cow, or a dead bird that fell into the sewer system.

Further complicating matters is the high level of virus in the milk supply — bits of virus have shown up inasmuch as 20% of store-bought milk. Although industrial farms aren’t typically connected to municipal sewer systems, waste from dairy processing plants does end up there. And people dumping their expired milk down the drain could be enough to trigger a positive signal in wastewater.

So when a wastewater sample comes back showing H5, it’s not a sure sign of an outbreak in that immediate area. But it’s still helpful information given the appalling lack of testing for bird flu in both humans and cows.

Verily, the life sciences arm of Alphabet, has teamed up with an effort from Stanford and Emory universities called WastewaterSCAN to test for H5 across their network of wastewater collection sites. Samples from early in the outbreak, which began in Texas, showed high levels of bird flu in Amarillo — an area with no farms, but a dairy processing plant just a few blocks from the wastewater treatment plant, says Bradley White, who helms Verily’s wastewater work.

Verily has also been able to detect H5 in states where cows later tested positive. Most recently, it picked up a signal in Iowa and Minnesota about a week before those states reported infected herds. Meanwhile, it has shone a light on the need for further investigation in sites with known outbreaks. Michigan, for example, is currently investigating the source of sky-high H5 levels in its wastewater to determine if infected herds or even infected humans are flying under the radar.

Verily and its partners, which include the CDC, can’t solve every wastewater mystery. When H5 showed up in a wastewater sample in the San Francisco Bay Area — a situation that led to some overly alarming headlines that seemed to suggest the virus was spreading undetected there — it took serious sleuthing by public health officials to determine that the likely cause was infected chickens.

But there’s hope for more clarity in the future, and we should be continuing to develop this technology to squeeze out all the information it can offer.

Verily is working on ways to disentangle useful signals from the sludge. The company is developing a method that could allow its scientists to more clearly pinpoint the source of a given viral fragment. The goal is to eventually be able to estimate that, say, 80% of the flu in a sewer system is coming from cattle and 20% is coming from birds. It could also, in the worst-case scenario, identify how much is coming from humans, White says.

Of course, that only works for the places where wastewater data are being tracked. The U.S. should be working to assess as much of the country as possible. Although imperfect, this disease snapshot is an important one to keep developing — especially when government agencies seem unable to offer a clear picture via testing of individual cows and farmworkers.

Even as we develop better ways to capture bird flu data from our sewers, we must complement that data with more testing of people and animals. That will put us in a better position if H5N1 does eventually adapt to be dangerous to humans.

Lisa Jarvis is a Bloomberg Opinion columnist covering biotech, health care and the pharmaceutical industry. Previously, she was executive editor of Chemical & Engineering News.