Sewage sleuths helped an Arizona town beat back Covid-19. For wastewater epidemiology, that’s just the start

GUADALUPE, Ariz. — Valerie Molina anxiously searched the desert sky, scanning the horizon for any clouds dark with rain. But all was blue and bright. Today, the weather wasn’t going to be on her side.

It was March 20, 2020, and Guadalupe, over which Molina presides as mayor, was preparing for what should have been the third of six Friday ceremonies in the run-up to Easter. Normally, it’s the time of year when the town of 6,700 doubles in size, as spectators from across the state descend on its white adobe church to witness young men in wooden masks, a sacred deer antler headdress, and ankle rattles made from the cocoons of butterflies dance to beating drums beneath ribbons of flowers. Guadalupe was founded by Pascua Yaqui Indians who were forced from their homelands in Sonora, Mexico, and settled in the Salt River Valley in the early 1900s, bringing with them their unique religion — a blend of spiritual animism and Catholic beliefs picked up from Jesuit missionaries.

A dusty, single square mile of low-slung, sun-baked homes wedged between Phoenix and Tempe, Guadalupe is culturally and economically a world apart from the golf course-studded subdivisions just beyond its borders. A third of its residents, most of whom identify as Yaqui, Hispanic, or a combination of the two, live below the poverty line, often with multiple generations sharing the same home. It’s the kind of place where you can’t walk too far without running into a relative. (Molina’s vice mayor, Ricardo Vital, is also her first cousin.) Though insular, the close-knit community isn’t isolated.

Most of Guadalupe’s working population traveled each day to essential service jobs at schools, restaurants, and landscaping businesses in surrounding areas. Molina wasn’t sure how her resource-strapped town — without a health department and with an annual budget of just $12 million — would be able to defend itself if the coronavirus gained a foothold there.

Only days before, the World Health Organization had declared Covid-19 a global pandemic, and Arizona Gov. Doug Ducey had banned any events with more than 50 people. Crucially, his order exempted religious gatherings. “That was the one thing,” said Molina. “And that was going to kill us.”

Much to her relief, spring rainstorms had cancelled the first Friday’s festivities. And the second Friday too. That had bought her some time to meet with the town’s cultural society leaders and come up with a plan for how they might keep the sacred ceremonies from becoming coronavirus superspreading events.

The leaders had agreed to take some precautions on that Friday the 20th and at all the remaining Lenten festivities — like barricading off the church to onlookers and canceling food vendors and other carnival activities. But the measures and the virus didn’t prove to be the deterrent Molina had hoped they would be. People continued to pack into the church plaza week after week.

Looking across the crowd that gathered on Easter Sunday, she felt sick to her stomach. “All I could think about,” she recalled, “was, ‘What’s it going to look like in 14 days?’”

Fourteen days later, one of Rolf Halden’s students wheeled a scratched-up cooler into his lab on the second floor of the Biodesign Institute — a copper-paneled complex of buildings on the Arizona State University campus. With in-person classes canceled and most research on hold, the cavernous bay of work benches was eerily empty except for Halden’s small team extracting the contents of the cooler: plastic jugs of murky gray liquid that only hours before had been drawn from Tempe’s sewers.

In some ways, this was an old routine for the scientists. Halden is director of the Center for Environmental Health Engineering at the Biodesign Institute. His lab had been analyzing the city’s wastewater for nearly two years. In 2018, they started fishing out molecular fingerprints of different kinds of opioids consumed across the city. The next year, they got a National Institutes of Health grant to use that monitoring network to track influenza. But when SARS-CoV-2 emerged that winter, they quickly pivoted to testing for the new coronavirus instead.

It took his team a few months to hone their methods but by April they were bringing more of the city’s seven sewersheds online each week. They filtered and spun down the samples, then loaded them into a desktop polymerase chain reaction, or PCR, machine. Inside, short strings of synthetic DNA swirled, grabbing onto any matching bits of SARS-CoV-2 they encountered, and giving off a fluorescent glow when they did. As the machine recorded any glimmers, it beamed the readout to a nearby computer. More light meant more virus, represented on the computer screen as a series of curves, one for each sewershed where samples had been collected: hundreds of thousands of toilet flushes captured in a spray of spidery arcs.

Two of them caught Halden’s eye. Today was the first time they were analyzing samples from Area 3. Located in the southwest corner of Tempe, Area 3 is shaped a bit like a giant amoeba eating a milk jug. That jug is the town of Guadalupe, and all of its sewage flowed into Area 3. Tempe utility workers were also sampling at the points where the pipes changed municipalities, so they could isolate the contribution of Guadalupe’s sewage from Area 3’s.

When Halden looked at the curve for Area 3, it showed a gradual incline, barely enough virus to be detectable. In contrast, the curve for Guadalupe shot straight up into the millions of copies per liter. Those numbers meant Covid-19 was already spreading through the community. If they’d only looked at Area 3, they would have missed it completely. “Area 3 is much bigger, so this significant signal from this small community was very quickly being drowned out,” said Halden.

He alerted Rosa Inchausti, Tempe’s then-director of strategic management and diversity, who was coordinating the wastewater testing pilots. In early May, she called Molina with a warning and an offer to share the data Halden’s team had collected.

“Just by chance, it was the little nugget we were looking for,” said Molina.

With Covid testing then limited, she was blinded to the virus spreading through her community and she had struggled to get health officials in the region to pay attention to her pleas for information. But armed with the wastewater data Tempe was offering, Molina was finally able to marshal the resources her town needed to fend off the virus.

For Halden, who for decades had been a lonely evangelist for the emerging field of wastewater-based epidemiology, the story of Guadalupe illustrates the potential he believes it has to transform public health in the United States — by creating a near-real-time monitoring network that could identify not just Covid-19 and drug use within a community, but also other pathogens and even chronic conditions like cancer, Alzheimer’s, and mental health disorders. “We’ve seen from our data how important it is to have a fine enough resolution that you can catch clusters of infections in specific areas, like Guadalupe,” said Halden. “We could only have observed that by having this neighborhood-by-neighborhood network.”

In the last year, the nation’s enthusiasm has finally caught up to his own. As the pandemic spread, hundreds of U.S. cities, states, prisons, universities, and private businesses leaped, sometimes clumsily, into wastewater surveillance. Federal investments in validating the science and building out a standardized national system followed. With vaccines driving SARS-CoV-2 underground, the question now is, what will governments, schools, and businesses do with all that surveillance infrastructure?

Tempe, operating at an unprecedented level of granularity, represents the leading edge of what’s possible. But while the technology becomes more powerful for answering scientific questions and crafting public health interventions the more finely you slice and dice the sewershed, it also raises more privacy concerns. The American wastewater surveillance revolution might have started with SARS-CoV-2, but there’s no telling yet where it ultimately will lead.

Halden, now in his mid-50s, wasn’t always fascinated by feces. He grew up near Braunschweig, Germany, about 15 miles from the floodlit, mine-laden complex of interlocking fences and walls known as the “death strip” that divided his home country into West and East. In the 1960s, as the Cold War turned the severed state against itself, the Iron Curtain wasn’t exactly a safe place for a child to go wandering.

But for Halden, it was less scary than staying home. His father was bipolar, and his mercurial moods often convulsed into violence. Being outside brought Halden relief. He’d let his legs carry him through fog-covered fields, seeking out the few streams and stands of trees that had managed to resist post-war agricultural land reforms. He found friends in tadpoles and newborn fawns. He discovered that this, observing life, had a name: biology.

In college, he took up the study formally. But when he graduated a few years later, he discovered there wasn’t much demand for biologists. He opted to stay in school and pursue another degree in something more practical: sanitary engineering. Then his adviser got an offer to become a professor in America and he asked Halden to join him. It would mean switching focus yet again. And moving to a city he’d never heard of — Minneapolis. But the important thing was that it was far from Braunschweig.

After knocking out an environmental engineering master’s and a Ph.D. at the University of Minnesota, Halden was eager to leave the cold and snow, but not ready yet to return home. So when Lawrence Livermore National Laboratory beckoned in 1997 with a postdoctorate offer that included a green card and mild Bay Area climate, he jumped at the opportunity. He stayed on another four years as a staff engineer, working in the rain-starved hills above the Altamont Speedway, where in December of 1969, four Rolling Stones concert-goers had died, bringing with them the end of the era of free love and radical idealism.

Back then, the U.S. Department of Energy used its nearby research facility, dubbed Site 300, to test new explosives and nuclear weapons. As a result, by the time Halden arrived a few decades later, Site 300 was one of the most polluted places in the U.S. It was his job to help figure out how to clean it up. His team discovered that some of the microbes that lived there were as efficient as any machines they had to extract the contaminants.

In 2001, he left Livermore to join the Center for Water and Health at Johns Hopkins University. While there, he got interested in another kind of toxic sludge, the kind that is left over after the contents of a city’s worth of toilet bowl flushes has gone through an industrial treatment plant. This semi-solid material is composed of bacteria and protozoa and everything the microbes can’t break down, including organic solids, minerals like phosphorus and nitrogen, and heavy metals. “Whatever ends up in the sludge is the stuff that biology doesn’t know what to do with,” said Halden. “It becomes an inventory of the toxic chemicals people are exposed to.”

The carbon in sludge acts like a sponge, soaking up and concentrating all the water-fearing chemicals. They might be rare in the environment — a few nanograms floating in the air or buried in the soil. But in sludge they build up many hundreds of thousands of times. And since many cities and counties convert their sludge into dry fertilizer and sell it to farmers and urban gardeners, essentially recycling and redistributing any pollutants inside it, Halden thought someone should be looking at sludge more carefully.

But the towering German, despite his infectious enthusiasm, couldn’t exactly drive around to Baltimore’s wastewater plants with canteen in-hand asking for handouts. “If you show up with a truck you could get a load for free and nobody would ask any questions,” he said. “But if you come with a jar and you look like a scientist, well, then you have a hard time, because they don’t want you to find something that will potentially create problems for them.”

Frustrated, he called up the Environmental Protection Agency, which regulates wastewater. In 2001, the agency had collected about 100 samples of sludge from across the country as part of a nationwide survey. As it happened, they’d been sitting in a freezer for five years and were about to be thrown out. All Halden had to do was drive to a nearby contractor’s lab and pick them up. A few years later, when the agency did another national survey, Halden, who had relocated to Arizona State, raised his hand again. He was also spending hours on the phone with wastewater treatment plant operators, begging for access to samples.

Sludge repository amassed, Halden’s team dove in, pulling the samples apart molecule by molecule using a gas chromatography-mass spectrometer. Among the molecules they discovered were significant amounts of triclocarban and triclosan, two microbe-slaying chemicals widely used in soaps, detergents, and hand sanitizers. When broken down in the human body, they derail hormone signaling, and they have been linked to allergies in children and low sperm counts in men.

Other researchers had found the same chemicals in the bodies of earthworms and bottlenose dolphins, in the eggs of seabirds and in human breast milk. But Halden’s research was the first to find them in sewage sludge. He began to sound the alarm, even testifying before Congress on the environmental health risks of the increasingly ubiquitous pollutants. In 2016, the Food and Drug Administration banned triclocarban and triclosan, along with 17 other antimicrobials from hand soaps and body washes, in part based on data from Halden’s wastewater surveillance.

Around that time, Halden had recently returned from a sabbatical to Europe, where a different form of wastewater monitoring, one that worked upstream of treatment plants, had begun to take off. Over there, countries incinerated their sludge, they didn’t spread it on food-growing fields. So there wasn’t much reason to study it. But starting in 2000, researchers in Milan showed that they could use samples taken from the Po River and from inside sewers in different Italian cities to measure the amounts of different prescription medications the local populace was excreting. A few years later, they did the same thing with cocaine. In 2010, researchers established a Europe-wide network of sewage samplers to probe for illicit drug use across the continent. By the time Halden arrived to tour the project, it had grown to an annual survey of 56 cities in 19 countries.

“The concept was to analyze in as many places as possible and you do it right and you do it every year and you look at trends,” he said. Such data serves its purpose well — illuminating for European regulators the true size of the drug market — but it has its limitations. Halden, who had been bitten by the public health bug during his stint at Johns Hopkins, saw the potential to generate data that could have a more direct impact. “I thought, why don’t we do this much more frequently and share the data with the community in real time so it actually becomes informative for taking public health actions?”

To test the idea, he applied for a number of grants, but only got a pile of rejections. Then, in 2018, he learned that the Tempe City Council was looking for projects to help reduce opioid-related overdoses and deaths.

“We were trying to find where the opioids were within certain areas of our city,” said Tempe City Councilmember Joel Navarro, a firefighter who is also chair of the Regional Opioid Action Plan Committee. Halden told them he could provide a neighborhood-by-neighborhood look into the Tempe population’s drug use by panning for chemical remnants left behind in the city’s sewage. After all, everybody poops.

After considerable public conversation, the council voted to try it out. Tempe invested $35,000 to build out the sampling network, and ASU kicked in matching funds for Halden’s lab to run the analysis. One of the first issues was figuring out where to do the sampling. In Tempe, as in most cities, that system relies on gravity. When a resident there flushes a toilet, the contents go out into a sewer line angled so that everything inside it flows downhill until it reaches a bigger pipe that feeds into a regional treatment facility. At the last downstream point before those bigger pipes dump into the plant, the city has established monitoring sites to measure the rate and volume of flow. The plant then uses those measurements to charge the city for processing the wastewater. From an efficiency point of view, it made sense to just repurpose those. So the city installed composite samplers at each one — essentially an automated pump programmed to draw a small amount of wastewater out of the sewer at regular intervals over a 24-hour period.

Starting that summer, municipal utility crews would stop by each station weekly to gather the goods. Then one of Halden’s students would pick the samples up and bring them back to the lab, where they’d be filtered and run through the spectrometer, which would tell them exactly which opioids were in each one, and how much. They reported these levels back to the city, which posted them to a public dashboard.

The information helped Tempe’s Fire and Medical Rescue Department respond when synthetic opioids like fentanyl and carfentanyl began flooding into certain corners of the city at the end of 2018. “We saw this drug show up as an extremely high spike before we started getting the calls,” said Nick Ells, medical services deputy chief for Tempe. That spurred them to stock up their trucks and ambulances with more of the overdose medication Narcan. With heroin, it usually takes 1 or 2 milligrams of Narcan to bring somebody into a state where they’re breathing. For fentanyl and carfentanyl, his team has found it often takes 5 milligrams or more. “These synthetic narcotics are just so much more potent than anything else you’ll find on the streets,” said Ellis.

Around that time, Halden nabbed a $1.5 million NIH grant to turn the network they’d established for monitoring drugs into an early warning system for flu outbreaks. His lab was gearing up to start that work in December 2019. And then SARS-CoV-2 emerged in Wuhan, China. “By January, it was clear the coronavirus would be coming here too,” said Halden. “So we started to explore the possibility of pivoting to measure an RNA virus instead.”

Inside the Centers for Disease Control and Prevention, molecular epidemiologists like Amy Kirby were batting around the same idea. She had studied using wastewater to track the movement of antibiotic-resistant bacteria in the environment. And she knew that countries like Israel had used the technology for decades to monitor for the virus that causes polio. There was also increasing evidence that people infected with SARS-CoV-2 were shedding viral RNA into their feces. “The problem was that prior to 2020, there was no infrastructure for wastewater surveillance in the U.S.,” she said.

How much information you can extract from sewage depends on what you’re looking for. Because SARS-CoV-2 was so new, scientists hadn’t yet figured out how to translate viral RNA concentrations into numbers of infected people. (In fact, that’s still an open question.) And lots of local variables — temperature, how far wastewater travels, how long it sits in the sewer — can all impact the outcome. Without preexisting infrastructure, no one had set standards or best practices for how to collect and analyze samples.

“We didn’t want to just muddy the water with more data that’s not interpretable, so the first question we tried to answer was can we understand whether this data is useful for public health?” Kirby said. To do so, the agency leaned on a few early adopters, like Halden and the city of Tempe, who were willing to volunteer their data. It showed that the most meaningful information came from looking at changing viral concentrations over time. “We saw repeatedly across multiple communities that wastewater trends were leading case trends,” said Kirby. “So when cases went up, the first signal we’d get would come from the wastewater, about four to six days before confirmed cases.”

Last August, the CDC announced the creation of a National Wastewater Surveillance System, or NWWS (pronounced “news”), with Kirby at its head. One of its aims is to create a data portal to collate local findings. She’d always wanted to build something like that, but the return on investment for monitoring antibiotic resistance hadn’t been high enough. “Covid completely changed the calculus on that,” she said. What would normally have taken four to five years to establish, her team was trying to do in four to five months. “Now, not only is it worth establishing, but it’s worth establishing as fast as you can with whatever data you can pull together.”

As Kirby was leading the charge to develop standards and recommendations, wastewater surveillance was exploding, driven by local government officials and business leaders scrambling against the spread of the virus and desperate for any scrap of data that could help them make better decisions. Hundreds of towns and cities participated in a free testing program offered by Biobot Analytics, a startup that spun out from the Massachusetts Institute of Technology in 2017. Dozens of colleges and universities launched their own pilot projects.

In the fall, the Department of Health and Human Services signed a $1.5 million contract with Aquavitas, a Scottsdale, Ariz.-based startup run by one of Halden’s former graduate students, to study how well its wastewater epidemiology technology tracked SARS-CoV-2 across 100 treatment plants serving 10% of the U.S. population. That project is now moving into phase two — sampling wastewater from 340 plants across 42 states, albeit with Biobot now running the testing.

Adam Gushgari, Aquavitas’ CEO, has been somewhat shocked at the sudden interest. He recalled attending a conference back in 2018, where Halden, his adviser at the time, gave the keynote address. Afterward, they gathered for beers with all the other wastewater epidemiology researchers in the U.S. There weren’t more than 15 of them. “Now, if you got everybody in the U.S. who is doing this, we would fill a stadium!” he said. “All of a sudden we went from being a fringe science that nobody really cared about to being in the forefront of the public eye.”

For some legal scholars and privacy advocates, that sudden shift signals trouble down the road. “Covid-19 provided the trigger for lots of colleges and municipalities to invest in this surveillance infrastructure,” said Natalie Ram, a law professor at the University of Maryland who studies bioethics and genetic privacy. “They’re not just going to disassemble it when Covid-19 is no longer a public health emergency. Which raises the question, what else is this infrastructure going to be used for?”

Companies like Aquavitas can provide at least a glimpse into the possibilities. Gushgari told STAT that while Covid-19 is still his biggest business driver, he’s starting to see the emergence of other markets. He said he’s in talks with one city about monitoring for molecules that people with Alzheimer’s disease shed into their blood and urine. Another one is interested in cancer. One client, a private high school with a strict vegetarian diet policy, contracted Aquavitas to look for signs their students were sneaking animal products on the sly. He says there’s no ceiling on what the technology can be turned to. “If you take it or you’re exposed to it and it comes out of your urine or feces, we can track it,” he said.

Tempe, for its part, is starting to explore other applications of its surveillance apparatus. On June 10, the city adopted a budget that includes an additional $950,000 to begin monitoring for at least a dozen new substances, including molecular traces of other viruses such as measles, parasites, alcohol, asthma medications, and stress hormones, and biomarkers of exposure to toxic fumes like benzene. The results will be publicly reported on an online dashboard. “I think there is potential to start really morphing into a whole gamut of areas to make the city healthier,” said Councilmember Navarro. The funds will also expand the city’s wastewater sampling infrastructure to 11 sewersheds over the next two years.

When such monitoring is happening on the scale of an entire neighborhood or university campus, Ram said there’s not much to worry about in terms of one’s privacy. Wastewater, by its very nature, is an admixture of material from lots of different households. And there’s no way to connect which toilets different chemicals came out of. But the more granular things get — say a city sampling every block — then there’s more cause for concern.

There’s also the scenario where it’s not governments, but private entities, like schools or employers, using wastewater as a way to find out something about the health of its workforce that it otherwise wouldn’t be privy to. As new as the science of wastewater epidemiology is, the legal and ethical frameworks around it are even less established.

“That’s what gives me pause,” said Ram. “Like all surveillance tools, it always starts with the least objectionable use — the most heinous crimes, the most palatable public health crisis. But it almost never stays with that narrow use. So that’s not to say we shouldn’t embrace the technology during a hopefully once-in-a-century public health emergency, but we also need to think about the ethical, legal, and social implications of doing so.”

If you ask Molina, she’d say that embracing the technology made all the difference for Guadalupe. Shortly after last year’s Easter celebrations, an official from the Maricopa County Public Health Department let her know that someone living in the ZIP code that encompasses Guadalupe and neighboring areas of Tempe had tested positive for Covid. She asked for more details. Guadalupe accounted for only a fraction of the nearly 46,000 people living in that ZIP code. Was it actually in her town? But, she said, Maricopa County officials offered no further information, not until she sent them the wastewater data Halden’s lab had generated.

Once they had segregated out Covid-19 test results by address, it was obvious exactly how much of a problem Molina had on her hands. Guadalupe’s residents were contracting the coronavirus at rates nearly six times higher than the rest of the county. Given the lack of testing at the time though, the real number of infections was likely far higher.

That data also helped Molina get the ear of Peter Yucupicio, the new chairman of the Pascua Yaqui Tribal Council. As a kid, he’d spent summers with his grandfather in Guadalupe. And he’d seen what was happening in the Navajo Nation, where Covid-19 was infecting more people per capita than anywhere else in the U.S. As the death toll climbed, the loss of tribal elders — the keepers of language and stewards of sacred knowledge — was especially traumatic. He couldn’t let that happen to Guadalupe. Yucupicio gathered the council and together they decided to send $2 million to the town.

It was a generous gift, and Molina wanted to use it well. She thought back to when she was a kid, and what would happen if you started missing school. Two women, friends of her grandmother, would come around to find out what was going on. And then they’d help. Sometimes that meant doing laundry so the children had clean clothes to wear. Sometimes it meant teaching parents how to rid their kids’ hair of lice. They were like social workers, but they didn’t work for the state. These so-called promotoras were women with deep ties whom everyone trusted, especially when they gave advice about health.

“I thought, that’s what we need for Covid-19,” said Molina. “We need a promotora to go to people’s homes and educate them about masks and testing and how to quarantine.”

So she called up Veronica Perez, a recently retired tribal social worker who knew just about everyone in Guadalupe and offered her the job. Perez agreed, and she recruited Graciela Holguin, a hospice worker, to help her. Together, they began criss-crossing the town, working off a list of names and addresses provided by the county’s contact-tracing team, delivering boxes packed with food, laundry detergent, masks, thermometers, hand sanitizer, and disinfectants to households where someone had tested positive. If people wanted to talk, they’d do so from the driveway.

Early on, Perez made the mistake of wearing black sneakers — they got so hot from the summer sun that they burned weeping blisters onto her toes. Some days they started before 6 a.m., and they often finished after midnight. “Everyone was in the dark,” said Perez. “We were just trying to help them with whatever they needed.”

But they were behind the curve, and as Covid-19 engulfed the state in a post-Memorial Day surge, more and more of the promotoras’ time was spent at wakes. In La Cuarenta, the oldest part of town named for the original 40 acres on which Guadalupe was founded, the church towers peeled a mournful toll every time a Covid-19 victim passed away. Residents of that neighborhood began calling it the “summer of the bells.”

“It was one or two every week,” said Molina. “That’s when it really hit our community how serious this was.”

Around that time, the Maricopa County health department and Arizona State joined the Pascua Yaqui Tribe as partners, and in mid-June, the town launched an official Covid-19 response team. They converted an open-air market to a free testing site. The county bought up rooms in a hotel just up the road from Guadalupe and converted it to a free isolation center, providing private rooms for people with Covid-19 to recover away from vulnerable family members.

The partnership and the promotora program started to pay off. By the end of August, the amount of virus in the wastewater had dropped below detectable levels. Confirmed cases petered out too.

Throughout the fall and winter months, Guadalupe experienced a few upticks, but nothing like the devastating post-holiday surges that stole hundreds of thousands of American lives across much of the country. When the FDA authorized vaccines, the promotoras switched up their routine, educating the town’s residents about how the vaccines worked and answering their questions. By the time this year’s Lenten festivities rolled around, more than 40% of Guadalupe residents had received at least one shot. Most days the town didn’t record a single new Covid case.

On Easter Sunday this year, the plaza filled with flowers and music and the smells of foods being cooked over open flames. People packed the church to pray. And this time, Molina wasn’t hovering, worrying. She was at home, where her fully vaccinated family gathered, all together for the first time in more than a year. There were hugs, and tears, and not a Zoom room to be seen. “It felt like a different world,” Molina said.

Correction: A previous version of this story incorrectly described the funding of Halden’s wastewater sampling work for Tempe.