Tracking drug addicts to identify, avoid hot spots

Let's face it: maps make scientists drool. And increasingly easy-to-use yet complex maps of anything from the spread of diseases to cloud formations kept all the geo geeks giddy at the Association of American Geographers' annual meeting in D.C. last week. One of those maps could reveal new clues about when and where drug addicts are at their most (and least) vulnerable.

When it comes to drug addiction, the tendency is to think in terms of a disease of the brain, says National Institute on Drug Abuse researcher David Epstein, whose presentation at the annual meeting included a map tracking the whereabouts of two heroin addicts in Baltimore over several months.

But what if addiction is also an environmental disease, the prevention of which requires avoiding mapped hot spots of temptation? As Epstein sees it, "One of the major cognitive theories of relapse involves seemingly irrelevant decisions--like a recovering alcoholic takes the scenic route, which happens to include a bar, and they relapse. You can have an intervention that on-the-spot warns people about where they are going based on data about neighborhoods in general and their behavior specifically."

Epstein and colleagues are now studying 25 heroin addicts, all found via a methadone treatment research clinic and involved in "methadone maintenance." Epstein says they all participate in studies that can last up to nine months, and sign consent forms from the start. Through this clinic, he hopes to track 150 Baltimore addicts in the coming months.

So how does his team get people willing to be followed as they engage in illicit behavior? "We explain in the consent form some rationale behind the study, with a sentence that the purpose of the GPS is not to keep an eye on you but to study how neighborhoods affect behavior," Epstein says. "People seem to get that."

Each patient is given a GPS unit the size of a thick domino that tracks motion each time a distance of 25 meters is covered, or clocks in the same location point every 25 minutes when each patient is still.

Each patient is also given a PDA, which beeps at four random times every 24 hours with a series of multiple-choice questions about drug use, stress, who's around, and what is transpiring. Participants are asked to record when they use drugs, and random urine tests gathered three times a week help determine how honestly and accurately people are self-reporting their drug intake.

These patients check in at the research clinic every day, at which point the previous day's GPS and PDA data is uploaded. The resulting map of the first two subjects, which I cannot publish due to privacy concerns, uses squiggly lines to show where each person went and black dots to show where each person reported using.

The team had already gathered all sorts of data to measure such things as wealth, liquor store prevalence, violence and crime rates, employment, and race throughout Baltimore. With their first participant, they found that while the man spent the majority of his time in lower-crime, higher-income neighborhoods, most of his "black dot" time--aka when he got high--was spent in the lowest-income, highest-crime neighborhoods.

Epstein says that ultimately these maps can help researchers not only gather better descriptions of actual real-time behavior--as opposed to retrospective data, typically gathered several weeks after events and more likely to involve biased memories--but also to better map out hot spots within cities that addicts could more actively try to avoid.

"Maybe we can broaden the focus to see that addiction is a brain disease, but it is also an environmental disease, and maybe you can intervene on an environmental level."

In other words, there may soon be an app for that.