Monitoring COVID-19 in sewage

In October 2019, Rolf Halden and his coworkers at Arizona State University were gearing up to see whether they could monitor the seasonal flu in wastewater. Their plan: look for nucleic acid sequences specific to the flu in sewage in Tempe, Arizona, and see whether the viral levels they measured lined up with the infection rates being reported by the city’s health authorities.

Little did they know that their preparation for the flu was just a warm-up for a pandemic caused by a novel coronavirus.

Halden is a pioneer in wastewater-based epidemiology. In this field, measurements of compounds in municipal sewage—usually by mass spectrometry—provide information about the health of a city, a neighborhood, or even a school. The field got its start monitoring drugs such as opioids to determine how widespread use is and to gauge the effects on public health.

The field has been relatively small, but the pandemic is changing that. “All of a sudden, we find ourselves in the focal attention of the public, and the discipline has exploded,” Halden says. A year ago, he says, the general public was unaware of what could be learned by analyzing wastewater. “Today, most likely they’ve heard about the possibilities,” he says.

Like Halden, many in the field have turned to monitoring SARS-CoV-2, the virus that causes COVID-19, in wastewater. Researchers who previously focused on measuring viruses in lakes, rivers, and oceans have found themselves drawn into the field. And researchers who analyzed wastewater to improve treatment plants are shifting to study the virus. Their measurements are helping guide and assess public responses to the pandemic.

Moving into viruses is “a logical step,” says Adam Gushgari, CEO of AquaVitas, a company spun out of Halden’s group at Arizona State. “One of the real beauties of the field and the reason we were able to stand up these monitoring campaigns so quickly is a lot of the techniques and strategies used for chemical analysis or other viral analysis were easily translated to SARS-CoV-2.”

Early in the pandemic, scientists realized that individuals infected with SARS-CoV-2 shed the virus in feces (Sci. Total Environ. 2020, DOI: 10.1016/j.scitotenv.2020.141364). By February, they’d determined that fecal shedding can start before people show symptoms—and even if they never show symptoms.

People quickly recognized the potential of wastewater analysis to track the virus’s spread in the community. Ana Maria de Roda Husman and Willemijn Lodder of the National Institute for Public Health and the Environment in the Netherlands detected viral RNA in wastewater collected at Amsterdam’s Schiphol Airport just 4 days after the country’s first confirmed cases of COVID-19 (Lancet Gastroenterol. Hepatol. 2020, DOI: 10.1016/S2468-1253(20)30087-X).

Drugs and other compounds in wastewater are usually measured by mass spectrometry. Although viruses can also be detected by mass spec, researchers achieve lower detection limits with nucleic acid testing, one of the types of clinical diagnostics being widely deployed in the pandemic. These tests capture virus-specific RNA sequences, convert them to DNA, and amplify and detect the DNA using quantitative polymerase chain reaction.

Importantly, what the researchers are measuring is viral RNA, not the active virus.

When people hear that viral RNA is in wastewater, they may think it’s loaded with SARS-CoV-2, but it isn’t. Instead, it contains genetic material from the virus that can’t make people sick, says Kyle Bibby, an environmental engineer at the University of Notre Dame who, along with Halden and colleagues at Howard University and Stanford University, is leading a global collaborative on SARS-CoV-2 surveillance. Water treatment approaches remove much hardier viruses than SARS-CoV-2, he says. Researchers need to be careful not to overstate the danger of sewage itself.

Researchers think wastewater-based epidemiology could answer three key questions about the pandemic. At this stage of development, the method is better able to answer two of those three questions.

The most basic question is simple: Is the virus present?

Wastewater monitoring can easily give that a yes or no answer. Results can usually be reported within a day, which can help public health officials determine where to direct resources. But interpreting that answer is complicated, Bibby says. There’s variability in the amounts of virus that people shed. Some people don’t shed the virus in feces, and others might continue to shed after the virus is no longer replicating.

The second question people would like to answer is whether wastewater monitoring can accurately track trends in the amount of virus in wastewater, says Mariana Matus, the CEO of Biobot Analytics, a company focused on wastewater-based epidemiology. “There’s an increasing consensus that wastewater can give you an early warning for spikes,” she says, “because people will start shedding the virus right away as they get infected, but they develop symptoms later, and they get the test even later.”

Such trend data can also reveal whether restrictions are effective. For example, on Tempe’s COVID-19 dashboard, which Halden is involved in, the viral level dropped below method detection limits for weeks in a row in April and May.

“That’s the lockdown,” Halden says. “It worked like dynamite. And as soon as it was lifted, the numbers crept up again.”

Because of how quickly the virus shows up in wastewater, people hope testing it can provide an earlier warning of outbreaks than is available with clinical testing, which often happens only after people are symptomatic. A heads-up of even a couple of days could be helpful for mobilizing a response.

In Queensland, Australia, “we use it to identify hot spots as an early-warning system,” says Kevin Thomas, director of the Queensland Alliance for Environmental Health Sciences. “We’re in a lucky situation that we have very few cases.” In that situation, any measurements above the detection limit are of concern.

For areas with more cases, wastewater measurements may not be able to predict further spikes in cases, according to Davey Jones, an environmental scientist at Bangor University who is involved with government programs to monitor the virus in Wales and part of England. Spikes can be difficult to distinguish from the normal variation in signals. “I think we missed the boat there,” he says.

And it’s still not clear, Bibby says, how the numbers should guide actions. “If you’re measuring water out of a nursing home and you get a positive, you don’t really care about some of the nuances. You’re probably going to test all the residents and staff,” he says. “But if you’re in a community and one day the concentration is 100 genome copies per liter and the next day the concentration is 300 genome copies per liter, it’s not clear that’s a significant difference or that we need to implement a lockdown.”

Bibby is alluding to an as-yet-unanswered question for the field: How many infected people does a given measurement represent? A major hindrance to making that calculation is that the amount of virus shed by individuals can differ by seven orders of magnitude, he says.

Much remains unknown about how many virus particles individuals excrete and whether and how that amount varies with the disease’s severity, Halden says. “There’s too little information to make accurate assessments, but that doesn’t render the signal useless.”

Matus thinks such estimates will be possible soon. Starting in late March, Biobot ran a 2-month pro bono campaign in the US and Canada, providing SARS-CoV-2 measurements for communities in 43 states and provinces. It analyzed samples from 360 wastewater facilities that served communities ranging from rural areas and small towns to large cities and entire metropolitan areas. The company paired those data with the number of clinically confirmed cases in those regions to look for trends.

The company is estimating the approximate number of cases from the virus measurements, Matus says. Such calculations have been controversial in the field because of the many uncertainties.

When correlating wastewater measurements to positive clinical tests, Matus and her colleagues found that trends in the wastewater values ran about 4–10 days ahead of trends in confirmed clinical tests (medRxiv 2020, DOI: 10.1101/2020.06.15.20117747). The research has been published as a preprint, which means it has not been peer-reviewed. “We believe it’s because of the lag between people starting to shed and people developing symptoms and getting a diagnostic test confirmation,” Matus says.

Communities are using Biobot’s data to help shape their response to the pandemic. For example, in New Castle County, Delaware, officials used the data to determine which areas they would prioritize for mobile testing services, Matus says.

Some universities are using wastewater analysis as part of their strategies for preventing and containing outbreaks. In the spring, the University of California San Diego was one of the first to announce that wastewater surveillance would be part of its reopening strategy.

UC San Diego’s program involves monitoring individual dormitories. “You have to go right to the source,” says Smruthi Karthikeyan, a researcher involved in the university’s program. Karthikeyan says finding the perfect spot in the sewage system to target a particular dorm is tough because samples collected too close to the source are, as she puts it, “more ‘solidy.’ ”

In the UC San Diego project, researchers try to determine the number of infected individuals by correlating the measurements from the dorms with parallel measurements of the on-campus hospital, where they know how many COVID-19 patients there are.

Universities aren’t the only schools where wastewater is being analyzed. Lian Lundy of Middlesex University is participating in a project to use wastewater to track SARS-CoV-2 levels in 70 primary and secondary schools in the UK.

“The advantage is that schools are a controlled population, so you know exactly who’s there,” Lundy says. “We know that kids tend not to get the disease as seriously, but we didn’t know anything about their role in transmission. Are kids regularly carrying the virus but not getting sick, or are they not carrying the virus at all? If you know the answer to that, then even if you have an outbreak, maybe you don’t need to lock down the schools.”

The program will run in parallel with one that performs clinical tests at the same schools. The researchers collected the first samples at 10 schools last month, Lundy says, and their goal is to collect samples every other week throughout the school year. The advantage of the wastewater testing is that it’s noninvasive and nondisruptive, Lundy says.

Lauren Stadler, a wastewater engineer at Rice University, has been working with a colleague at Rice and with the Houston Health Department to collect and analyze samples from 39 wastewater treatment sites. They analyze the samples in parallel in two, and sometimes three, labs.

“We report the results back to our health department every week. And the health department actually is using our data to inform interventions,” Stadler says. “I think the reason it’s actually being used is because it’s a project that’s really been led by the health department and the health department’s been involved every single step of the way.”

Despite adoption in many settings, wastewater analysis still has multiple sources of uncertainty. For example, samples can be collected in different ways, either as grab samples or as so-called 24 h composite samples. Grab samples are just what they sound like—stick a container in the waste stream and pull out a sample. Composite samples combine multiple smaller samples into a single sample, but the name can be misleading.

“A 24 h composite sample is not 24 h of chemistry. It actually is only about 5 min of every 60 min,” Arizona State’s Halden explains. “People obsess over the assay they use,” he says. “In the grand scheme of things, it matters very little” compared with the probability that an excreted signal actually ends up in the sample container. “The smaller the population sampled, the more important it becomes to employ specialized samplers that access wastewater continuously, not just every 15 min or so.”

Analytical methods can introduce more uncertainty. The accuracy of the measurement depends on the methods used to concentrate, extract, and quantify the virus, which can lead to different numbers. “We learned very quickly that the methods that you choose to use impact the raw concentration value that you determine,” Stadler says. But they reach similar conclusions about what those concentrations are telling them even with the different methods.

Research networks are trying to make it easier to develop and share methods. For example, Bibby started a global collaborative to coordinate sharing of methods and data (Environ. Sci. Technol. 2020, DOI: 10.1021/acs.est.0c02388). “It began as an informal network,” he says, but it quickly grew to hundreds of researchers. “The collaborative is really a platform for people to talk to each other.” The hope is that it will help new people move into the field more quickly and increase the productivity of those already in the field.

Lundy is working on a European joint initiative that has shared a common protocol for collecting and analyzing sewage samples to facilitate the comparison of results from different studies.

In the early days of the pandemic, “anybody who could was rushing out and getting samples and analyzing them with different techniques and approaches. One way to reduce this uncertainty is to try and get a common method,” Lundy says. “If we can have a common method for collecting the samples, storing the samples, analyzing the samples, and then making that data available, that would be a contribution to the research community.”

Researchers in the field are optimistic that the infrastructure they’re building will help with this pandemic and the inevitable future ones. But they also worry that the rapid development could work against the field.

“There is hard evidence that the science [of wastewater-based epidemiology] works, but I am the first to tell you that there are limitations,” Halden says. “If we don’t pay attention, it can really send the wrong message. Overpromising and underdelivering is a horrible habit.”