Some placebos are more equal than others

ResearchBlogging.orgPerusing the news early this morning, I noticed an article on ABC News about placebos. One thing I found interesting about it was that it was a story about a research letter to JAMA, not a full study. Heck, there isn't even an abstract.

Even so, the study was rather interesting and described thusly:

The more expensive your pain medications are, the better the relief you get from taking them -- even if they're fake.

That's according to a study published this week in the Journal of the American Medical Association, which suggests that sugar pills labeled as expensive drugs relieve pain better than sugar pills labeled as discounted drugs.

Of course, this is entirely consistent with what is known about the placebo effect, namely that more invasive and elaborate procedures or treatments tend to have a seemingly greater effect on the relief of subjective symptoms. Indeed, this is very likely a major reason why homeopathy appears to work for so many people, even though most homeopathic remedies have been diluted to nothing but water. It's a reason why surgical procedures themselves can produce a powerful placebo effect, even when they do nothing.

So what did this little study show?

In the study, 82 volunteers were subjected to a series of electric shocks -- a standard research protocol for measuring pain thresholds. They were then given a placebo pill alongside a fake drug company brochure for the fictitious drug "Veladone-Rx," -- ostensibly a new fast-acting painkiller made in China.

The only catch was that half of the test subjects received brochures showing that the drug had been marked down from the original price of $2.50 a pill to 10 cents a pill. These modified brochures also included circled fine print which suggested that the pills were manufactured in China.

After participants went through the shocks again, 85 percent in the full-price group reported pain relief from their sugar pill, while only 61 percent in the discount group reported pain relief.

This describes the brief report pretty accurately. The p-value for the above finding wasn't the kind that would knock me out (0.03), but it was statistically significant. The authors also divided up the results according to all shocks and the highest intensity shocks only, with results as follows:


What's also remarkable is that not only did more patients receiving the more "expensive" placebo report a reduction in pain but more patients receiving the "less expensive" placebo reported an increase in pain than those receiving the "more expensive" placebo (43.9% versus 19.5%). The results were even more striking in that increased pain relief was observed for the "expensive" placebo at all voltages used. The authors conclude:

These results are consistent with described phenomena of commercial variables affecting quality expectations1 and expectations influencing therapeutic efficacy.4 Placebo responses to commercial features have many potential clinical implications. For example, they may help explain the popularity of high-cost medical therapies (eg, cyclooxygenase 2 inhibitors) over inexpensive, widely available alternatives (eg, over-the-counter nonsteroidal anti-inflammatory drugs) and why patients switching from branded medications may report that their generic equivalents are less effective. Studies of real-world effectiveness may be more generalizable if they reflect how medications are sold in addition to how they are formulated. Furthermore, clinicians may be able to harness quality cues in beneficial ways,6 for example, by de-emphasizing potentially deleterious commercial factors (eg, low-priced, generic).

This study is, of course, rather small, but, given its consistency with much of what is known about the placebo effect, I tend to think it's likely to be correct. Obviously, replication and extension are required. I'm not sure I agree with the authors on how this effect can be harnessed, though:

For Hadler, the study might convince doctors to develop their own positive marketing for a treatment.

In his own studies, Hadler has found that the way a physician describes a drug can change how much a patient will follow through with a treatment regimen.

"Compliance goes down when you go through all the side effects listed for the drug," said Hadler. "But if you say, 'This is the best thing ever, side effects are rare,' people will respond positively."

The problem with this approach is that it can easily flirt with being unethical and with being dishonest to patients. If there are known complications, it's also a lovely way to leave oneself open to a malpractice suit for failure to disclose risks of treatment. On the other hand, changes in packaging to resemble brand-name drugs and other methods of making generics look less "generic" may well be helpful.

This study also suggests to me why some CAM therapies works so well. Not only do patients tend to invest a lot of effort and time into them (think acupuncture, for example, which frequently requires multiple sessions a week lasting 30 minutes or more), but these therapies can be fairly expensive. It is as though CAM practitioners instinctively understand how best to maximize the placebo effect.


Waber, R.L., Shiv, B., Ariely, D. (2008). Commercial Features of Placebo
and Therapeutic Efficacy. Journal of the American Medical Association, 299(9), 1016-1017. (Link)


More like this

This may be a naive sort of question, but: Is there any correlation between a patient's a priori belief in the efficacy of a treatment? That is, is the placebo effect affected by a patient's trust or skepticism toward a particular treatment?

Yes, good stuff, though it does confirm and reinforce earlier research. I quote some good science from Ben Goldacre at

We know from research that four placebo sugar pills a day are more effective than two ... we know that salt water injections are a more effective treatment for pain than sugar pills, not because salt water injections are medically active, but because injections are a more dramatic intervention

and the more elaborate placebo ritual was more effective than the simple placebo tablet.

There's a book about people's reactions to food tastes, which plays on similar placebo effect. People think food and drink tastes better the nicer the presentation (or the more expensive the label) or the more expensive it is, in blind taste tests that serve the same thing two different ways. I think it's called "Mindless Eating."

I'd like to know if the effect of an existing medication is influenced by a patients knowledge about a newer/more expensive/"better" treatment...

I'm always stunned to see results like this *not* followed by calls for studies attempting to isolate the brain patterns leading up to the ingestion of the "expensive" medicine vs. the "cheap" medicine.

I'd really like to see research that works on drugs or therapies which replicate these "placebo effect" mechanisms in a controlled manner. Ultimately, we should have doctors prescribing *those* alongside the pain drugs.

Sure, "lie to the patient" is one such "therapy", but rather than wasting time/energy quibbling over the underlying ethics of it, shouldn't we shift focus to finding a way to accomplish the same ends ethically (and likely with superior results)?

Maybe I'm missing something - is there work being done in this area?

Yes there is research being done on the placebo effect and how to trigger it. It can be triggered pharmacologically (to some extent) via increasing the basal NO level as I discuss in my blog on it.

My interpretation of the placebo effect is that it is a "standing down" of physiology from a "fight or flight" state where resources are diverted to healing and away from being reserved for immediate consumption such as to run from a bear.

Most researchers don't understand the placebo effect and everything that is non-pharmacologic gets lumped into it, including investigator bias. That is a wrong approach.

I see the placebo effect as the neurogenic mediated physiology that the body uses to switch from the "fight or flight" state to the "rest and relaxation" state. The archetypal placebo effect is a mother's "kiss it and make it better". It tells the child they are safe now, and can calm down and start to heal. It is increasing the basal NO level neurogenically.

I think my bacteria can produce an increase in basal NO, and in effect trigger the placebo effect pharmacologically. Once the placebo effect is maximally triggered, there is nothing more that any other placebo can do. I think this is the way to deal with the CAM crapola that is only placebos. Use a better placebo, one that works pharmacologically.

Thirty-plus years ago I worked with pharmacists who told me they occasionally received prescriptions for placebos with the note "charge appropriately." The idea, in accord with this study, was to convince the patient they were receiving something valuable.

"It is as though CAM practitioners instinctively understand how best to maximize the placebo effect."

An appeal to instinct is unnecessary. Think of it as natural selection.

Another reason not to buy cheap drugs from "Canadian" internet pharmacies - they will have inferior placebo effects.

By Freddy the Pig (not verified) on 05 Mar 2008 #permalink

Does this mean that we'll see a hike in price of normal drugs too, just to improve their feel-good factor? ;)

DrFrank: This was my first thought, too. "Hey, we've got a great research-backed way of making this painkiller more effective." Or even "Do you want the pretty good painkiller for $10 or the really effective painkiller for $50?"

My mother in law was a practicing psychologist for many years, and in the last years of her practice she made extensive use of aromatherapy and similar stuff. When she mentioned this, I was very surprised, and asked her if she thought they did any good. She responded that she figured she was getting nothing but the placebo effect, but that since she was just trying to make her patients feel better/more relaxed, that was just fine.

By albatross (not verified) on 06 Mar 2008 #permalink