Credulous reporting on placebo effects strikes again

Let’s face it. The vast majority of “complementary and alternative medicine” (CAM) or “integrative medicine” (IM) therapies are nothing more than placebo medicine. This should be so abundantly clear to readers who have followed this blog, Science-Based Medicine, and/or Neurologica Blog more than a few weeks that I shouldn’t have to repeat it yet again, but I feel that it bears repeating today as an introduction to today’s subject matter. CAM/IM is almost all placebo medicine, and placebos effects are poorly understood (or even misunderstood), even among physicians. It is this misunderstanding that has provided the opening for CAM apologists, faced with the failure of their favorite woo to demonstrate efficacy over and above that of placebo time and time again in well-designed randomized clinical trials, to “rebrand” their methods as “harnessing the power of the placebo.” In doing so, they like to ascribe magical powers to placebos, implying that it can do more than just decrease the perception of pain or other subjective symptoms but in fact can lead to objective improvements in a whole host of diseases and conditions. Some even go so far as to claim that there can be placebo effects without deception, citing a paper in which the investigators–you guessed it!–used deception to convince their patients that their placebos would relieve their symptoms.

I was reminded of just how badly placebos are misunderstood by a recent article in the Wall Street Journal by Shirley S. Wang entitled Why Placebos Work Wonders: From Weight Loss To Fertility, New Legitimacy For ‘Fake’ Treatments. If you want a primer on how not to write about placebos as a journalist, I’d be hard pressed to find a better example for you to study than this article. Wang falls for all the CAM/IM tropes that they like to use to demonstrate that their methods are anything more than ineffective methods that can provoke a placebo response. In fact, jumping ahead in her article a bit, I know she has no understanding of the issues involved–or even an active misunderstanding–when she writes a passage like this:

Ted Kaptchuk, director of Harvard’s Program in Placebo Studies and the Therapeutic Encounter, and colleagues demonstrated that deception isn’t necessary for the placebo effect to work. Eighty patients with irritable bowel syndrome, a chronic gastrointestinal disorder, were assigned either a placebo or no treatment. Patients in the placebo group got pills described to them as being made with an inert substance and showing in studies to improve symptoms via “mind-body self-healing processes.” Participants were told they didn’t have to believe in the placebo effect but should take the pills anyway, Dr. Kaptchuk says. After three weeks, placebo-group patients reported feelings of relief, significant reduction in some symptoms and some improvement in quality of life.

Wang so completely fell for the spin that Kaptchuk put on this study and so obviously doesn’t understand what she’s writing about that it makes me instantly question the rest of her article, particularly the parts where she cites studies. As an aside, let me just complain, as well, about how Wang and the WSJ do not post links to the actual studies mentioned in the article. I can understand why studies aren’t listed by name in dead tree editions of newspapers and magazines designed for a lay audience, but there is no excuse whatsoever anymore for journalists, newspapers, and magazines not to include PubMed links to the actual studies in online versions of articles on medical and scientific topics that mention scientific studies. This would allow readers to look at at least the abstracts and make it possible bloggers like me, who have access to university accounts, to look up the actual articles themselves if we’re so inclined without having to do PubMed searches ourselves and guess which articles are being referred to. In fact, in this very article in a couple of cases I had a hard time figuring out exactly which articles Wang was referencing.

Fortunately, in this case, I had already blogged about the actual primary article in depth and explained exactly why Kaptchuk’s spin that the study indicated that it was possible to induce placebo responses without deception. The long version is in this link. The short version is the observation that subjects were recruited for this study by ads touting a “novel mind-body management study of IBS [irritable bowel syndrome], which introduced selection bias for people prone to be interested in “mind-body” interactions. Moreover, while it is true that Kaptchuk and his team told subjects that they would be receiving placebos, but they also told subjects that the sugar pills used “have been shown in rigorous clinical testing to produce significant mind-body self-healing processes,” which is, to put it kindly, an exaggeration. Add to all this the way outcomes were measured were custom-designed to exaggerate. The no-treatment arm demonstrated an IBS Global Improvement Score of 4 (no change) compared to the Open Placebo arm, which averaged 5 (slightly improved). This is highly unlikely to be clinically significant. Despite all these problems, this study was widely touted as somehow being slam-dunk evidence that placebo effects can be invoked without deception when it is anything but. The best possible spin that could be put on this study is that it is consistent with previous work that expectation effects are important in placebo effects. In other words, if you expect an effect, even if you know you’re taking a placebo, you’re more likely to feel better.

Of course, I shouldn’t be too hard on Wang, I suppose, at least not for this. After all, it apparently fooled Edzard Ernst himself into calling it “elegant.” I’ll also mention that she also discusses (and gets mostly right) another study that I’ve blogged about just last summer. In fact, it was a study that was prominently featured in the discussion panel that I participated in at TAM last summer, along with Steve Novella, Kimball Atwood, Mark Crislip, Harriet Hall, Rachael Dunlop, and Ginger Campbell. Steve even mischievously switched back and forth between two of the graphs in the paper to make a point. Yes, I’m referring to the “placebo in asthma” study, or, as I called it, dangerous placebo medicine for asthma. Wang correctly points out that only the active treatment (albuterol) improved the underlying biology but that both groups felt better. This is more or less the very definition of placebo effects: Feeling better without any actual improvement. Yet that’s not the overall impression that her article gives, as she cites a number of studies that suggest that placebo effects are more than just an effect on “how a person experiences or reacts to an illness.” She even uses an argument from popularity, pointing out how many physicians knowingly prescribe placebos based on a study from 2008 (which, I can’t help but mention, I also blogged about when it came out), as did Abel Pharmboy, Janet Stemwedel, Jake Young, revere, and Peter Lipson, who quite aptly said about this study, “Placebo—I do not think it means what you think it means.” The reason is that the authors counted many things as placebo, including pills known to have actual pharmacologic activity and how the authors defined placebo, as Peter pointed out:

In the current study, a placebo is defined as “positive clinical outcomes caused by a treatment that is not attributable to its known physical properties or mechanism of action.” This implies that the physician either knows the treatment shouldn’t work, or doesn’t understand how it works. This isn’t just semantics; we have many treatments available whose exact mechanism of action isn’t known, but whose effectiveness has been proved. If you interpret the definition less strictly, it oxymoronically defines a placebo as something that works despite it’s lack of efficacy. If I prescribe something expecting a predictable effect, and it produces that effect, by definition it isn’t a placebo. If I prescribe something I expect to work, and it doesn’t, then it isn’t a placebo. If I prescribe something expecting failure, but it works, I’m a lucky idiot. This would seem to imply that there is no such thing as a placebo (and I might agree).

The same is true of another study cited by Wang, a study by Alia and Langer from 2007 that examined the question of whether placebo effects can augment the effects of exercise:

Hotel-room attendants who were told they were getting a good workout at their jobs showed a significant decrease in weight, blood pressure and body fat after four weeks, in a study published in Psychological Science in 2007 and conducted by Alia Crum, a Yale graduate student, and Ellen Langer, a professor in the psychology department at Harvard. Employees who did the same work but weren’t told about exercise showed no change in weight. Neither group reported changes in physical activity or diet.

This is the study in question, which was not only not randomized (the hotels from which subjects were recruited were randomized to control or “informed,” not individual subjects, but tolerated a huge difference in the ages of the informed group and the control group, with the mean age of the informed group being eight years younger than that of the controls. Moreover, the reported changes in the informed group are pretty small, so much so that one wonders whether, even if they are real, they are clinically significant. For example, the change in weight in the informed group was from 145.5 lbs to 143.72 lbs, and in body mass index from 26.05 to 25.7. These are, of course, good things, but small changes, although the ten point decline in systolic blood pressure was intriguing. Does this remind you of anything? Perhaps of the IBS study I mentioned above? It should. Based on a small study finding small differences in outcomes that, while statistically significant, are probably not clinically significant (as I discussed the other day!), the authors make an overblown conclusion, in this case, that “mind set” is a powerful mediator of exercise effects, concluding:

People have mindlessly overlooked what it means that placebos are inert. Ultimately, each individual is responsible for their effects. Recognizing this suggests that it is time for us all to explore more direct means of controlling our health, such as pursuing mindfulness (see Langer, 1989) as a tool to actively and deliberately change our mind-sets.

“Mindlessly” overlooked? Who edited this paper? That’s not exactly scientific writing there. Maybe that sort of language is considered acceptable in the social sciences. I can say from my experience that such language is not acceptable in formal scientific writing for the medical sciences. Be that as it may, This study doesn’t exactly mean what its authors think it means, and they seem confused about placebos as well, defining them as “any effect that is not attributed to an actual pharmaceutical drug or remedy, but rather is attributed to the individual’s mind-set (mindless beliefs and expectations”). Oh wait. perhaps “mindless” is meant to mean “unconscious.” That would explain a lot. Clearly the editors were unconscious when this paper came through.

All too often, “placebo” seems to mean exactly what people choose it to mean, no more, no less (apologies to Lewis Carroll). In reality, a placebo is nothing more than “a substance or procedure a patient accepts as medicine or therapy, but which has no specific therapeutic activity” or, as Wikipedia now defines it, “simulated or otherwise medically ineffectual treatment for a disease or other medical condition intended to deceive the recipient.” This strikes me as a better definition than the old definition because it emphasizes that placebos are medically ineffectual and that they involve deceiving the patient. Indeed, the necessity of deception is, Kaptchuk’s claims otherwise, part and parcel of placebo use, which is one key reason why using placebos has fallen out of favor. Using placebos outside of a clinical trial is now generally considered at best paternalistic and at worst downright unethical, because it violates informed consent and patient autonomy. Sixty or seventy years ago, it was considered acceptable for physicians to deceive patients that way. in 2012, not so much.

In any case, what we call the “placebo effect” is not a single effect, and it has many components. Placebos area actually best viewed as a rather artificial tool used in clinical trials to control for nonspecific effects. There are expectation effects, in which patients experience what they are led to expect to experience. There are effects due to observation. Patients in clinical trials almost always do better than those not in clinical trials, thanks to the closer attention and more rigorous treatment protocols. This has sometimes been referred to as the “clinical trial effect.” Then there are effects due to reporting, which can introduce bias. There are also a whole host of other factors that determine how strong a placebo effect will occur in any situation. Surgery and invasive procedures are more powerful placebos than pills, for instance. More expensive placebos tend to produce stronger apparent effects. Indeed, there’s a whole hierarchy of placebos, and placebo effects can be enhanced by things like empathy and the doctor-patient relationship. None of these things require the introduction of pseudoscience such as CAM to achieve.

In the end, I’m coming to agree more with Mark Crislip than I used to in that I’m starting to question whether placebos are nearly as powerful as they are commonly advertised, although I don’t think I go so far as to call them a myth. Placebo effects, more than anything else, appear to involve changes in how pain or subjective symptoms are perceived, not any physiological change that concretely affects the course of a disease. Consistent with this concept I have yet to come across a study that provides serious objective evidence that placebos change “hard” objective outcomes, such as survival in cancer. What placebo is frequently claimed to be by advocates like Kaptchuk but is almost certainly not is “mind over matter” or thoughts and mind controlling health. Unfortunately, Kaptchuk and his ilk frequently find willing mouthpieces like Wang to spread this message because it’s such a seductively appealing message. After all, who doesn’t want to believe that we can control our health with our minds? Who doesn’t want to feel that powerful, particularly when disease strips us of control?

More importantly, for people like Kaptchuk who believe in CAM, placebo effects provide a new rationale to use CAM even though the vast majority of it is placebo medicine. By that I mean physiologically inert but represented by practitioners to patients as real medicine. Clinical trials, as ill-advised as many of them are, continue to reinforce that conclusion. In medicine, when a treatment performs no better than placebo, it is interpreted, and correctly so, as meaning that treatment doesn’t work. Thanks to the magic of “mind-body” placebos, propagandists like Kaptchuk have found a new rationale to use the ineffective treatments that make up so much of CAM.

And credulous journalists like Wang are more than happy to help him promote that rationale.