Scientists are dipping into city sewers in search of drugs. Why? The same reason people on parole take urine tests instead of surveys. You can always lie on a questionnaire; outwitting your toilet, on the other hand, can be a bit more challenging.

The technique, known as sewer epidemiology, was recently used to compare illicit drug use across 19 European cities, and could represent the future of population-wide drug testing. To learn more, we spoke with the study's lead author, along with one of the field's pioneering scientists.


Did you snort coke today? Did you smoke any weed? How much did you consume? How often do you use? Given the chance to answer these questions anonymously for a nationwide survey, would you? If so, would you answer truthfully; or โ€” perhaps a better question โ€” could you answer truthfully? How closely do you monitor your drug intake, really?

These are just a few of the pitfalls of questionnaires on illicit drug use. People avoid them. When they don't, their answers are inherently subjective. Most people probably couldn't tell you how much coke they consume in a typical line (about 100 milligrams). And, of course, people lie.

Sewers, on the other hand, positively reek with honesty. They also draw from an unprecedentedly massive sample population. "In countries where you have well-developed sewage collection networks, everyone has to go to the toilet, and most people use a toilet that's connected to a sewer system," explains Kevin Thomas, an ecotoxicologist at the Norwegian Institute for Water Research in Oslo, Norway. "At the end of the day," he tells io9, "sewage is just a really big urine sample."


Thomas is the lead author of a study, published in the August issue of Science of The Total Environment, that used sewer analysis to monitor and compare illicit drug use throughout Europe. The most ambitious study of its kind ever conducted, the researchers spent one week in March 2011 collecting waste water at the inlets of 21 sewage treatment plants distributed across 19 cities in 11 countries. These plants service an estimated 15-million European inhabitants โ€” a downright formidable sample population.

Thomas and crew were looking for molecular traces of five drugs: cocaine, amphetamine, ecstasy, methamphetamine and cannabis. The figure featured here illustrates their results: population-normalized loads of amphetamine, methamphetamine, cannabis and ecstasy, along with average estimates of cocaine consumption throughout the cities. All values are presented in units of milligrams per 1000 people per day. (Click to enlarge โ€” for more details, see Figure 2 of the research article).


Thomas says there are some key things to take away from the data: Apart from Amsterdam (almost certainly due to its liberal weed laws), numbers for cannabis consumption were pretty consistent across all the tested sites; and yet, the numbers for every other drug bespeak dramatic regional differences in consumption.

The Netherlands and Belgium have a very high instance of cocaine, says Thomas, whereas Finland, Norway and the Czech Republichave a very high instance of meth amphetamines.

"High per capita ecstasy loads were observed in Dutch cities, as well as in Antwerp and London," he explains. Perhaps not surprisingly, cocaine and ecstasy loads were significantly higher on the weekends than during the week. Some "back of the envelope calculations," as Thomas calls them, suggest that the European continent consumes between 350 and 360 kilograms of coke per day.


Thomas also points out that it's possible some of the drugs being detected in waste water are coming not from drug users, but producers. Drug labs inevitably lose some yield to sewers, while impending drug raids can pressure suppliers into flushing large quantities of product down the toilet:

"In the course of one week of collecting samples, we think we managed to detect a raid in Utrecht when high levels of ecstasy in our sewage samples were found to correspond with a drug bust performed two days prior," explains Thomas.


Cocaine lends itself to similar fine-toothed analyses. Unlike ecstasy, which the researchers identify in sewer samples by searching for the drug itself, cocaine can be detected either in its pure form or its metabolic byproduct (the compound that it becomes after passing through the human body, a chemical known as "benzoylecgonine"). By comparing the ratio of these two different but related compounds, researchers can compare how much coke is going up people's noses against how much winds up circling a shower drain. Another twist: when someone snorts coke and drinks at the same time, the result is an entirely new metabolite called cocaethylene. In theory, says Thomas, scientists could actually use sewer analysis to determine how much cocaine is being co-consumed with alcohol.

All these examples point to the tremendous utility of sewer epidemiology, but they also raise some pressing questions. How reliable is this technique, how might it be implemented in the future, and โ€” the question on many people's minds โ€” could it be used to track drug loads in my neighborhood, my street, or even my home?

For insights, we spoke with Roberto Fanelli, head of the Environmental Health Department at Mario Negri Institute for Pharmacological Research. Fanelli helped pioneer the practice of probing the environment for signs of illicit drug use in a 2005 study of water from Italy's Po river. Since then, he's been at the forefront of sewer analysis, and pushing for its widespread implementation.


Fanelli is quick to praise the technique's reliability, but he says this wasn't always the case. "At first, I was surprised by the fact that we could use this methodology and always get results," he tells io9. In time, however, Fanelli believes sewer analysis has proven to be very reliable.

Both Fanelli and Thomas note that, since 2005, sewer analysis studies have consistently turned out results that are in-line with those of traditional, survey-based drug statistics. They also tend to match well with Europe's ranking of national drug problems. What's more, they do so with unprecedented speed. Data is collected and processed in almost real-time. "In a few days," Fanelli explains, "you can determine how much of a drug has been used in a given town." A typical survey-based technique, he says, might take up to a year to distribute, collect and analyze. By then it's old data, and limited, at that. In addition to the issues we mentioned earlier regarding subjectivity and honesty, surveys lack the temporal resolution afforded by sewer analysis.


The fact that you can trace a population's drug use over the course of any given day, says Fanelli, means that this technique could be useful in a variety of potential applications. Many people will be relieved to hear that monitoring individual houses, streets, or even neighborhoods is not one of them. According to Fanelli, such an analysis is possible in principle, but it's also tremendously impractical.

Timing, for instance, would be a logistical nightmare. "Let's say you have two people using illicit drugs in a neighborhood of 1,000 people," Fanelli explains. In order to detect their drug use, the timing of your sewer samples would have to coincide very closely with their bathroom visits โ€” events he describes as being "very diluted in time."

"This is quite difficult," Fanelli says, repeating himself for the sake of emphasis; "quite difficult."


Thomas agrees:

You could go to a neighborhood and conduct an analysis, but that's not what we're looking at, or what anyone else who's working with the technique is doing, as far as I can gather. It's easier to determine population level statistics, and not trying to track down which neighborhood has a bigger drug problem.

As a point of reference, Thomas notes that the smallest city his study looked at had a population of 112,000 people. "There's no way whatsoever we can connect what we measure to any single individual or even group of individuals," he explains. "It's totally anonymous."


On a larger scale, the applications of sewer analysis are manifold. Among the most interesting that Fanelli mentioned in our interview is its potential for gauging the economics of international drug trade:

One thing that people are not aware of is that this is a very good methodology, we think, to calculate the amount of money [tied up in] illegal traffic. One can use this methodology to determine the kilograms of cocaine or other drugs, which are used every day in a certain place or city, and easily calculate how much illegal money is entering the economy.


But sewer epidemiology has some major hurdles to overcome before achieving its full potential. Thomas, for example, warns against extrapolating the data from his team's study (which examined individual cities) to entire countries. Also significant, he notes, is the fact that his study, while unprecedented in its scope, still only considers measurements made over the course of a single week. "That's just a snapshot," Thomas explains. To realize their full potential, sewer analyses would need to be conducted on a much larger scale, and over longer periods of time.

For now, at least, the technique appears to be making significant headway. Thomas tells io9 that a followup study is already in the works, and that it will incorporate a U.S. city. The inclusion of America in a comparative investigation, notes Fanelli, is an important milestone in and of itself. Americans, he says, are a very private people, and that can incumber progress.

"In Europe we do not have this problem. We are not afraid, in general, of privacy or something like that, but in the U.S. it is very different."


I ask him if he thinks the U.S. will resist the adoption of sewer analysis.

"Yes," he responds with a laugh. "You have a lot of lawyers."

The results of Thomas and his colleagues are published (free of charge) in the August issue of Science of the Total Environment.


European Illicit Sewer Tour poster by io9's Stephanie Fox; all other images via Shutterstock