Current News

/

ArcaMax

Testing poop to predict Miami's COVID-19 trends? It's been a messy process so far

By Adriana Brasileiro, Miami Herald on

Published in News & Features

MIAMI — In March, when it became clear that the U.S. was facing an unprecedented pandemic, Miami-Dade County began sampling its sewage as a potential tool for measuring the extent of COVID-19 infections. The hope was that testing the county's poop for the coronavirus could serve as an early warning indicator of a dreaded second wave expected in the fall.

Now, with cases rising all over the country and Florida experiencing a steady increase to levels not seen since August, what is the sewage saying?

The short answer: Not much, at least not yet. The process got off to a slow, messy start but they've cleaned up the data and there is still some promise it will work.

For months, the county has been paying $3,600 a week to see if testing sewage could help estimate coronavirus infection trends in the population. Miami-Dade's Water and Sewer Department sends samples from its three wastewater plants to a specialty lab in Boston called Biobot.

The results were meant to help former Mayor Carlos Gimenez and public health experts get ahead of the virus and potentially save lives by better preparing hospital infrastructure for surges. After all, everybody poops, and wastewater reflects all individuals in a community, whether they have COVID-19 symptoms and access to testing or not.

Because the coronavirus shows up in human waste just a day or two after infection, mining wastewater for COVID-19 data emerged as another COVID-19 gauge. Used in conjunction with other data, it could help provide a sense of where the COVID curve might be headed.

But while cities like Paris and Boston have integrated wastewater testing results in their strategies to fight COVID-19, WASD says the turnaround time on test results is not yet fast enough to make it a planning tool in Miami-Dade.

"We need to have more confidence in the data. The results at this point are mimicking testing data that's coming to us from the state," said WASD director Kevin Lynskey.

Miami-Dade Water and Sewer takes weekly samples from its wastewater treatment plant in Virginia Key as part of a COVID-19 testing program.

Lynskey said that the Biobot data are still mostly a look back at the COVID curve, rather than forward for indication of trend. For example, recent Biobot analysis of Dade's sewage shows case number estimates started to trend up around mid-October, but there is no data available yet after Oct. 27 because of sampling delays by the county, Lynskey said.

Problems have come at both ends of the process — the collections of samples from the county's three major sewage plants and the analysis of it by Biobot.

When sampling started in March, the results were unreliable, with spikes and drops that made no sense when analyzed against the clinical and testing data of COVID-19 patients tracked by the Florida Department of Health, Lynskey said. Biobot admits that its methods needed improvement after it launched the testing program in March.

Until August, Biobot's algorithm was estimating the prevalence of the virus, or how widespread the disease was in the population. Because the lab was still fine-tuning its technology, results were taking as many as two weeks to get back, and the testing program was seen as a work in progress, Lynskey said.

Then in August, Biobot revised its model. It included more information in its analysis, including new research on how the virus behaves in wastewater, and started producing results that showed the occurrence of new cases, which could point to trends, rather than just the proportion of cases in the population at a certain moment. And the company also began processing data much faster, turning tests around in just a few days.

"We've been working with hundreds of communities over several months, so we're in a very different place today in terms of our understanding of lab methodologies, data analysis pipelines of the virus just purely because of the data set that we have amassed," said Newsha Ghaeli, Biobot's president and co-founder. "Our sensitivity levels have changed and improved."

Back in March and April, Biobot would provide data back to communities two weeks after sampling, a result that wasn't actionable at all, she said. Now Biobot can test samples in a day and send back results the next day.

 

So why isn't Miami-Dade taking advantage of this faster turnaround? There have been delays with its own sampling over the past three weeks, so there was no testing through Biobot or data after October 27. WASD said there were delays receiving kits from Biobot and it's working on getting the testing program back on track.

In Boston, meanwhile, the water and sewer company is making the data available and encouraging officials and the public to use it in conjunction with other information to make decisions about the pandemic.

In July, the Massachusetts Water Resources Authority started posting on its website the results of COVID testing at its Deer Island treatment plant. The results are shared with staff from the Department of Public Health, the Department of Environmental Protection, the Executive Office of Health and Human Services and the Executive Office of Energy and Environmental Affairs, according to MWRA.

"It is important to note that this is a pilot of an evolving science," the website said. "The results from this study will be used by public health officials as an additional tool for the Commonwealth to track how the pandemic is trending in Massachusetts, along with data from clinical tests, hospitalizations, etc."

Some wastewater sampling programs also are being implemented by other states. Since July, the Wisconsin State Laboratory of Hygiene in partnership with the University of Wisconsin, Milwaukee and the state Department of Natural Resources has been testing samples from wastewater treatment plants once a week in about 20 counties with 75% of the population. Universities such as Yale and the Massachusetts Institute of Technology are also testing their sewage to track the COVID pandemic in their communities. The University of Miami is also starting its own program, sampling sewage from different collection points at its campuses.

It's not clear if Florida is working on a state-funded program because neither the Department of Health nor Gov. Ron DeSantis' office replied to requests for comment.

At the national level, the Centers for Disease Control and Prevention and the U.S. Department of Health and Human Services announced last month it is starting the National Wastewater Surveillance System to generate data to "help public health officials to better understand the extent of COVID-19 infections in communities."

CDC is currently developing a portal for state, tribal, local, and territorial health departments to submit wastewater testing data into a national database, to be used for public health action.

Data from wastewater testing are not meant to replace existing COVID-19 surveillance systems, but are meant to complement them by providing data for communities where timely COVID-19 clinical testing is underutilized or unavailable, the CDC said.

In the meantime, Biobot is using poop data from cities to do all kinds of epidemiological studies. For example, the lab discovered that people with COVID shed a larger viral load in their poop in the first few days of infection. So if sewage samples show a significant increase in viral load, it means that a spike in case numbers could happen in about a week or so, Ghaeli said.

Statistically speaking, when people are contaminated with the virus, they will likely show symptoms on day four or five. They will probably want to get tested, and if the system is not overwhelmed, test results would probably be available in two or three days. So the test result would trail the start of infection by about a week or more.

If the poop testing is done efficiently, it could provide public health officials with some advance warning about trends.

Lynskey said the county may decide to do more targeted sewage testing to identify hot spots and obtain a more granular view of infection trends.

"We could test individual grid points in our system. We can do more localized samples rather than sampling at the treatment plant," he said. "We could do them from more focused basins and start honing down on where the infections are more problematic."

(c)2020 Miami Herald Distributed by Tribune Content Agency, LLC