Research Matters: Anup Malani on “Advertisements Impact the Physiological Efficacy of a Branded Drug”
Research Matters is a biweekly feature in which a member of the faculty talks about some of his or her latest work and its impact and relevance to law and society.
Anup Malani, Lee and Brena Freeman Professor of Law and Professor at the Pritzker School of Medicine, worked with economist Emir Kamenica of Chicago Booth and Dr. Robert Naclerio, University of Chicago Professor of Surgery and Section Chief, Otolaryngology-Head and Neck Surgery, to test the effect of direct-to-consumer drug advertisements on health outcomes among patients using the allergy drug Claritin. Their work, "Advertisements Impact the Physiological Efficacy of a Branded Drug," was published in the Proceedings of the National Academy of Sciences of the United States of America.
Q. You’re a law professor and an economist. How did you end up studying the effects of drugs?
A. My dissertation was on placebo effects. I have my PhD from the economics department here. I got the idea while driving from Charlottesville, Virginia, to Washington, D.C., after I finished clerking and was about to go on my honeymoon. During the car ride it occurred to me that clinical trials are interesting for a reason that people hadn’t thought about, which is that when you don’t tell people what they’re going to get, when you blind them, but you tell them what the probability is that they’re going to be treated, you manipulate their expectation about whether they’re going to be treated. So if you compare, for example, a trial that has three-quarters of a chance that you’re going to get the treatment, versus a trial for the same drug where there’s only a 50 percent chance that you’re going to get treated, people will have different expectations. If you want to understand how expectations affect health outcomes, you can use compare treatment arms across clinical trials to do that. That was the focus of my dissertation, published in 2006. Since then, I’ve been doing a lot of work on placebo effects.
It’s not law; it is a research area that mainly doctors think about. I think, as an economist, I have a slightly different perspective that helped me make some methodological progress on questions such as, do placebo effects exist, what is the nature of such effects, how do they work, and what causes them. Also, what should policymakers do about them? Should the FDA consider placebo effects in drug approval? Should insurance companies pay more for drugs with placebo effects? How can health policy tap placebo effects, and at what cost?
Another question I’ve been thinking about was, what triggers placebo effects? What actually changes people’s beliefs about stuff? We know drug companies spend a lot of money trying to change our minds about their products. The question is, how does that change your beliefs, and how does that change in beliefs about the quality of drugs affect the efficacy of those drugs, the performance of those drugs, once you take them? I mean that in an objective way, not what you think about them, but how do they affect your physiology? That’s what brought us to this experiment on the effects of advertising on allergy medicines.
Q. How was the experiment performed, and what did you find?
A. We gave everybody in our experiment Claritin and something called a “histamine challenge.” If you’ve ever had an allergy test, you know they put a series of pricks with some allergens on your arm, and they slap it down, and they see if you have inflammation in your skin. We did the same thing but with histamine instead of allergens. Histamine is something all human beings will have an inflammatory response to, even if they are not allergic to things like pollen or grass or cats. So we gave people a histamine challenge to measure their inflammatory response to histamine, then we gave them Claritin, which in general should reduce the size of the inflammation on their skin from histamine (that’s why it’s called an antihistamine). Then we gave them a histamine challenge again after Claritin to measure the effects of Claritin. Some of these people had standard allergies and some did not.
But here’s the interesting part of this study: when people go and do these studies, they typically sit in a lab and bide their time by watching a movie. So what we did is splice the movie – “Shakespeare in Love” – with car and cereal advertisements, just like they’d show if they were showing the movie on TV. We randomized people to see different sets of ads. So for one group, we replaced one ad in each break with an ad by the manufacturer of Claritin that talked about how good Claritin is. For the second group, we replaced one ad in each break instead with an ad for Zyrtec, which sometimes said Zyrtec is great and sometimes said that Claritin doesn’t work, specifically that it takes a long time for it to work. And for the third group, we didn’t replace any ads with drug ads. That was the control group. We wanted to see what the interaction is between the advertising and the efficacy of Claritin.
At first, we thought we’d find that Claritin advertisements improved the efficacy of Claritin, and we were not sure what we’d find with the Zyrtec group. We were surprised to find that we didn’t find that the Claritin ad improved the efficacy of Claritin, relative to the control group, but we found that the Zyrtec ad reduced the efficacy of Claritin by a half, almost, after 120 minutes, relative to the control group. This is a surprising result.
The Proceedings of the National Academy of Sciences said that the first experiment looked good, but it’s really hard to believe. Please replicate this with a larger sample size. So we did, except this time we only focused on the Claritin group and the Zyrtec group. We randomized over 300 people to either get Claritin and see a Claritin ad, or get Claritin and see a Zyrtec ad. Everybody got Claritin, but they saw different advertisements. And we found that the efficacy of Claritin was statistically significantly lower for people who saw the Zyrtec ad than the Claritin ad. But the magnitude of the treatment effect was smaller. It wasn’t a 50 percent effect, it was a 5 percent effect. But it was the same sign.
There is one important caveat, which is that we divided people into two groups based on whether they already had seasonal allergies. What we found was that the advertisements were only impactful for people who did not have seasonal allergies coming in. The reason why we think that happened, although we are not 100 percent sure, is that we think that people who have allergies have already paid attention to a lot of allergy ads. So the incremental dosage of ads that we provided in the study is not very large, relative to what they’ve already received. They’re not going to be very responsive to seeing a few more ads that are pro-Claritin or anti-Claritin. Whereas people who don’t have allergies weren’t paying attention to those ads. The advertising dosage was relatively bigger for them. It’s still very policy relevant because about 10 to 15 percent of users of Claritin each year are newly diagnosed individuals who have not been paying attention to ads.
Q. Why does this effect happen?
A. That’s the topic that we hope will be the subject of follow-up research. One possibility to investigate is that people are stressed, both by the histamine challenge and by hearing that the medicine will not work. There may be a “flight or fight” response where their stressed bodies send out hormones that actually trigger an inflammatory response. The purpose of Claritin is to reduce inflammation. So you might have an effect from the Claritin, but it’s being offset by the immune system sending out pro-inflammatory hormones.
Q. Why does this matter for medicine and policy?
A. What medicine should be interested in is the effect of different drugs on health outcomes. Medicine cares about the effect of the drug on the outcome, regardless of what’s mediating it. A lot of the time we don’t even know what mediates the effect of a drug on an outcome. We’re not sure we 100 percent understand why each drug works. Placebo effects highlight the fact that there might be a non-physiological, a psychological, channel through which the drug has an impact on your outcomes. I see no reason to discount that. It still affects outcomes. That’s why medicine should care.
It’s also a reason why policy should care about it. If you’re trying to figure out whether to approve a drug, or whether to reimburse a drug, you should care about the effect of the drug on outcomes, not necessarily whether it comes from one physiological channel or another physiological channel or a psychological channel. Right? So that’s why you should care about placebo effects.
The paper says that advertising might affect your psychology in such a way that it affects the efficacy of drugs. Not just the drug itself, but the drug, plus some advertising, can have a positive or negative effect on outcomes. And one topic of policy debate is the extent to which we should regulate advertising. That is a law question as well, and a First Amendment question. To me, that’s the policy relevance of this paper. I want to be very careful: this is not an argument for why we should ban drug advertising, or even negative drug advertising. This is a potential that we’ve discovered in a clinical trial. In the real world, people are exposed to Claritin ads and Zyrtec ads at the same time. We need to do a little more research before we jump to a policy conclusion. But at least we should keep our mind open to the possibility that this has advertising regulation related implications.