An Ingenious Approach to Drug Compliance

Some ideas are just so great I am left in awe.

INSERT DESCRIPTIONPhoto: Christopher Harting

As Emily Singer writes in Technology Review that drug-resistant tuberculosis is an important problem, especially in poor countries. After you get TB, you are supposed to take antibiotics for six months to prevent drug-resistant strains of TB from arising. The problem, however, is that the antibiotics have side effects, and there is little private benefit to the person taking the drugs. So there is little incentive to take the full treatment of medication.

So how do you solve this problem? To an economist, the obvious approach is to give the patient strong incentives to take the drugs even after they feel better. The difficulty lies in monitoring whether the patient is actually taking the drugs.

That is where the brilliance of science takes over. Jose Gomez-Marquez, program director for the Innovations in International Health program at M.I.T., along with his colleagues there, came up with an ingenious solution.

They “figured out a simple paper-based test that detects metabolites of the TB drug in urine.” So if you take the drug and pee on a special piece of paper, a secret message appears. If you don’t take the drug, you can pee on it all you want, but it will not reveal the secret message. Every time the drug taker texts the secret message to the people in charge, he earns a prize, like cell phone minutes or cash.

It’s the perfect marriage of science and economics.

J Chang

A great solution, although I foresee a future where those on TB medications start selling urine to those who choose to be non-compliant but want the rewards anyway.


Yes, what J Chang said. One pot of urine and all the people in the village dip in instead of taking those suspicious foreign drugs.

Josh S

When do patients receive the pieces of paper? If they receive them when they get the drugs can't they just pee on all the pieces of paper the first day and then just send the text messages later?


You'd think the key would be to set the incentives so they're worth doing, but not worth breaking your back (and buying contraband urine) for.

Dan G

I have one major concern with the incentives in this program: what stops a TB patient from taking the drug for two months, then giving her medication to her sibling/parent/child/neighbor? So long as somebody is taking the drug and urinating on the strips, they can continue to collect rewards, but this defeats the entire purpose of the program. This would be an even greater incentive in poor countries where TB is prevalent, but families cannot afford the treatment; the precise location that this program is constructed to target.


J Chang/1, that seems like a lot of effort to try to win a prize, when the alternative approach is to simply take a pill.

The problem isn't that people abhor taking the antibiotics (unless they're the unlucky ones to have strong side effects), it's that they don't feel like it.


I don't understand. Isn't the prospect of living with drug-resistant (and deadly) TB scary enough to enforce compliance?

Richard E

Hmmm...a market in pee. Perhaps not a healthcare professional's ideal outcome...

In fact would this not just encourage those who don't mind the side effects to take the antibiotics for a fee? As and when they need the cash and so never completing a course of treatment, thus increasing the risk of anitbiotic resistance for both TB and other diseases?


Or they could save the papers, take the drug once a week, then do all of their business at once.

The better solution is to pay them to take the pills in front of a witness - just pay someone to walk around and hand out a daily dose... with a bounty if the monitor misses their daily appointment.

Imad Qureshi

I think this is not practical. How much cash can you offer to the patient. And who would offer this reward? A patient's close relative? This doesn't seem practical at all. Good idea? I don't think so.


Traditionally, we used to have people come in, watch them take their pill, and possibly give them a small reward when they come in. Directly-observed therapy works well, it's just labor-intensive on the part of the observer. But do these poorer nations really have a shortage of people able to watch someone take a pill?


OK, April 1st, anyone?


@Sriram: The problem is that you need to keep taking the pills (and enduring the side effects) for several months after you're actually cured. This is a problem not just with TB, but with any disease that's being fought with antibiotics.

Personally, I'm pretty amazed by how naively Steven believes in the limitless power of artificial incentives, given how time and time again they prove to be useless or harmful (see current economy). As this example proves nicely, the problem is that most of the time you can "incentivize" only an indirect effect of the actual thing you want to encourage - and people end up maximizing that, often bypassing the primary goal.


and how to motivate the people in charge?


How about making the antibiotics mildly pleasant and addictive like caffeine? Then the person taking the pill gets the benefit buzz and also gets the headache if they don't.


When I got to college, I tested positive for TB, but never had any symptoms. I may have "contracted" it years earlier. The doctor recommended I take this pill that would have side effects on my liver... and I couldn't drink for 6 months.

I don't know what you would have had to pay me to follow this regiment, but it would certainly have been upwards of ten thousand dollars... certainly cost prohibitive on a national scale.

I don't know how it works, but I guess I still have a dormant TB in me... I haven't gotten the test since (more than ten years later).

I guess real statistics on drugs like this would be nice before a patient would go through the trouble. But the doctors never have them.


Just pay them.


There are two things you better understand....

1. People in South America are less likely to have the Human Leukocyte Antigen genes that are able to deal with TB.
2. And all the drugs in the world will not help these people get rid of TB.

You can try to poison TB but it will act just like a cancer does. If you don't kill the individual cell, the body's immune system won't seek the TB contaminated cells out and kill them.

Do you know how populations adapt when they don't have the genes which will help them resist a disease? The ones who can't resist, die and the one's with resistance live and fill in the empty spaces left by the ones that died.

That's why Hispanics in the US are half European genetically.

Anominus for obvious reason

One time I took antibiotics for an ear infection, it cleared the ear infection but produced bleeding hemriods. based on that , ii think it would have to be one heck of a prize for someone to continue for 6 months.

billy bob

I can give you what I think are the real statistics. About 1 percent of people who test positive for TB come down with TB. About 1 percent of people who take TB drugs develop severe side effects. So for the individual it is a complete wash. Your risk of saving yourself from TB is about the same as your risk of severe side effects usually liver damage. I hope that the people who come up with these crazy ideas actually look at some of the readers' responses. It's pretty clear that most of those African villagers who are suspicious of "genocidal" whitepeople and their drugs won't take the TB meds, but they will have a community pisspot where everyone will dip their papers to get access to the secret message.