Sewage May Hold the Key to Tracking Opioid Abuse

The city of Cary, North Carolina, is a charming town of 130,000 in the state’s Research Triangle. With its tree-lined streets and parks, job opportunities, well-ranked schools—not to mention its residents' relatively high levels of education and its low crime rate—it's not surprising that it's ranked high on lists of best cities to live in the U.S. for years. (In 2004, it even ranked as the number-one town in America, according to Money magazine.)

So Cary is the last place that you would expect the opioid crisis to hit home.

Yet last year, 60 Cary residents overdosed on opioids—a 70 percent spike from the year before—and 11 people died. The city decided it was time to take action.

"As Mayor [Harold] Weinbrecht says, while we're not necessarily in a crisis, what city is better positioned to do something proactive about this than Cary?" says deputy town manager Mike Bajorek.

That "something" is a project that will monitor and track the use of opioids, neighborhood by neighborhood, by analyzing the town's sewage.

Funded by a Bloomberg Philanthropies Mayors Challenge grant, Cary has partnered with the startup Biobot Analytics to pilot a project where portable sampling stations suck up sewage from pipes. In the lab, analysts scan it for 16 different opioid metabolites—substances produced by the body once a person ingests or injects an opioid, ranging from legally prescribed drugs to heroin. The city hopes that the data will help pinpoint where opioid abuse is happening so it can better deploy education and resources.

Each sampling station is lowered into a pipe through a manhole, and scans between 4,000 to 15,000 people's worth of sewage. The resulting data will help the city get a neighborhood-level view of opioid abuse. Combined with other demographic data or data from the state's Controlled Substances Reporting System, which tracks when and where prescriptions are dispensed, could help the city further drill down into how drug abuse is taking place.

Mariana Matus, who cofounded the Cambridge, Massachusetts-based Biobot with architect Newsha Ghaeli, says that the advantage of her company's system is that each sampler is portable—the size of a small carry-on bag, and easily carried by one person—and costs less than competitors, which run up to $10,000 each. (Biobot charges a subscription model, with fees based on the size and density of the city and the nature of the wastewater network, although Ghaeli and Matus declined to give specifics.) The lower cost makes it possible to deploy the "biobots" all over a city, rather than in one or two places at a time. Cary's pilot, fully funded by the Bloomberg grant, includes sampling at ten sites, though Bajorek says he hopes to someday expand the program "one hundredfold."

"We are collecting sewage within the city—not just at the treatment plant, but in areas within the city," Matus says.

Mariana Matus, CEO and Co Founder Biobot Analytics

Watch on

Wastewater epidemiology is not an entirely new field. As Matus implies, others have been sampling sewage for clues into public health issues for at least a decade, mostly in Europe.

Rolf Halden, director of the Center for Environmental Health Engineering at the Biodesign Institute at Arizona State University, says that the U.S. is "playing catchup with Europe on the drug front." His own lab studies chemicals, such as antibiotics, in wastewater, and is partnering with the city of Tempe to monitor for chemicals—including opioids—there.

Halden says that sampling wastewater at the mouth of the treatment plant is "the most convenient" way to do it. "You just go see the personnel and say, 'Give me a little.'" But time is not kind to the compounds wastewater epidemiologists want to sample. "The chemicals have traveled for a long time in the sewer pipe. Sometimes it takes eight hours [to go from toilet to treatment plant]. A lot of things can happen in eight hours...In terms of data quality, it is often more superior to move up the pipe [to] pump stations...or manhole covers," he says.

Matus says that the metabolites Biobot is measuring degrade in a matter of hours, which is another advantage of its samplers—the closer to the source they can be placed, the better data they can get.

Bajorek and Jason Wittes, pharmacy director for Wake County, emphasize that the Biobot pilot is in the proof-of-concept stage for now, meaning Biobot and city employees are learning how to best analyze, validate and present the data collected from sewage. It won't be until later that public health workers actually use the information to make decisions.

"As we find out things from law enforcement, social media, as there's different kinds of medication being used or abused, they can test for it, and to have near real-time data on that is pretty remarkable," Wittes says. "Usually you see a report and [the data in it] is already a year old." He mentions, for example, that recently, some drug abusers have begun combining opioids with Neurontin, a medication used to treat nerve pain, which enhances the high one gets from opioids. "To be able to test for that, it will tell us in real time where it's being used. That helps us because we can change resources and outreach programs to areas we might not have focused on in the past." The data could help the county decide where to deploy doses of Narcan, the drug that helps reverse an opioid overdose, or which neighborhoods public health educators and peer support specialists should focus on.

Even better, Wittes says, is the potential that unlocks if and when Biobot or a competitor starts working in multiple cities—and Biobot says it will announce a second partner city this year, chosen from "a large list of cities that have expressed interest to work with us," Matus says. Wittes believes Biobot would be able to "help put us in contact with another health department that has a similar patient population using similar medications...[we can share] tactics of combating the issue in near-real time." Instead of waiting until a public health department presents at a conference, for example, two cities with similar demographics could connect and share ideas right away.

Of course, Biobot still has to prove its stuff. The startup, which spun out of a project Matus was involved with at MIT, has been testing its technology in Cambridge, and Matus thinks they've got their technique down. "We're doing much better [analysis] than has been done before," she says.

Halden is not entirely convinced. "I applaud their ability to muster support and enthusiasm for doing these measurements, to convince communities to work with them," he says. "But this is not easy. It is relatively easy to find wastewater—it is not so easy to get permission to analyze it and it is really difficult to reliably analyze wastewater for public health indicators, then take into account all the factors that can compromise the quality of the data." Such factors include population density (the number of people who are excreting into that specific pipe), the volume of other wastewater (such as from laundry or showers) that flows through that pipe, how much the chemical of interest has degraded before it was measured, and so on. "Measuring opioids in wastewater is exceedingly difficult. It's not like you can just buy an instrument and stick it on a desk."

Biobot, for its part, says that its team has, collectively, decades of experience. "We're not new to the field," Matus, who earned her PhD in computational biology at MIT, says.

There's another reason Wittes and others are hoping that wastewater drug monitoring takes off in a big way, and that's the potential of removing the stigma from drug abuse.

Halden says he's run into this issue in the past. If a city spends public money on drug monitoring, that information becomes public information (or can be obtained using public records requests). That can be a thorny ethical issue if only one city is monitoring for a certain drug.

"If you have only one entity, then everyone's eye is trained on them, and you become the capital of whatever," he says. If only one U.S. city tracked traffic deaths, "nobody would visit that city—but if you have data [from multiple cities], you know there's a risk and you deal with it." Halden has gotten around this issue so far by pooling data from cities in a given region, which doesn't give any one city specific information, but helps the region coordinate on responses. This is, of course, very different from the Biobot approach, which aims to release very targeted data on specific locations.

Wittes thinks that as long as Cary and Wake County's public health department share data carefully, the pilot project won't lead to stigma. "The data is helping us reduce stigma," Wittes says. "[Opioid abuse] is a scary and taboo thing...but it's happening everywhere."

Bajorek adds that already the project has given the city "an opportunity to talk to people about how to protect their families." It’s started a conversation.

Wittes says, "This is just day one, really."

Rachel Kaufman | READ MORE

Rachel Kaufman writes about games, cities and science.