08.04.2006
A placebo can help alleviate some kinds of medical problem, even though its ingredients are designed to have no affect (pure placebo), other than possibly mimicking the side-effects of a real medicine (active placebo). Sugar pills and saline injections are common placebos. In rare cases, a dummy treatment can even involve cutting a person open in a surgery room. (Yikes!) Why are placebos important for medical research? Certain ailments are influenced by the power of suggestion: the simple act of administering care, if done credibly by a doctor or other “healer”, can make people better … or at least feel better. That is the placebo response (or placebo effect). If you want to know if a particular treatment works or doesn’t work, then you also have to factor out the placebo response. Otherwise, it is impossible to tell if a patient recovered because of the substance of a treatment or just the ritual of treatment. Indeed, many sham medical remedies appear to work but are nothing better than placebos. (Try telling that to the patient who now feels better because of a belief in the efficacy of a bit of quackery; something to remember next time you see a testimonial.)
The larger implication is that the ancient distinction between the mind and the body is too simplistic. The extent of the mind’s healing power over the body, however, remains a matter of controversy. The controversy is intense because “cured through belief” and “healing power of the mind” are phrases that smack of hucksterism and New Age mumbo-jumbo. There has to be some sort of scientific foundation to all of this. Right? Yes there is, albeit a very incomplete one. Dylan Evans has taken stock of the scientific evidence in his book Placebo. Evans knows he’s straddling the line between two warring camps. Scientifically-minded, sceptical people are right to be suspicious of claims made about the physical power of belief. The other camp, advocates of complementary and alternative medicine, don’t want their treatments talked about as mere placebos because they consider them to be more potent than that. Evans sides with the scientific camp. But he wants scientists to take more seriously the power of the mind to heal certain ailments … or, to rephrase, the influence of the mind over the body’s innate healing power … or, to rephrase again, the role that the brain plays in the body’s immune system (a field of study called psychoneuroimmunology).
As that last sentence suggests, the words chosen to describe the placebo response have implications for the way it is perceived. There is a tension here. On the one hand, common words make the placebo easier to understand but also come with a lot of baggage and imprecision. On the other hand, modern scientific nomenclature is not as easy to understand and (thus far) lacks adequate terms to describe the relationship between the mind and the body. Some scholars manage the tension by coining new terms, few of which have caught on. Evans manages the tension by judiciously reintroducing a few common words (like belief) into the debate. By being careful in his choice of words, Evans reduces his exposure to accusations of muddying the waters of science. I wouldn’t call his book a light read but Evans deserves credit for being plainspoken and understandable.
Evans reports two disconcerting findings early on. First, less than four percent of clinical trials—testing a drug or medical procedure using an experimental control group—involve both a placebo control-group and an untreated control-group. Both are necessary if medical researchers are to isolate the placebo response from instances of natural recovery, such as spontaneous remission and the normal fluctuation of symptoms. It is also necessary to identify the side-effects of a placebo. This lack of research is partly owing to ethical concerns about the withholding of treatment and the application of fake treatment. To compound the problem, the widespread use of clinical trials of any kind is relatively recent (1970s). Second, because of this lack of research, the evidence of the placebo effect is surprisingly sparse. The study that popularized the notion of the placebo response (Beecher, 1955) is down-right dodgy: extensive misquotation of research findings, grossly premature conclusions, and lack of non-treatment control-group studies. It is amazing that so many scientifically minded doctors accepted the existence of the placebo response long before strong evidence became available. Even these days, Evans has to cope with a lack of sufficient evidence in many areas of medical research. This makes it difficult to determine what ailments are responsive to placebos and what are not.
To show that the placebo response is plausible, Evans has to come up with a substantive biological explanation of how placebos work. At risk of over-simplification, it goes something like this. The placebo response is a chain of physiological reactions involving the immune system. It starts in the brain, an organ that can secrete chemicals (“chemical messengers”) into the bloodstream. The chemicals suppress inflammation (pain, swelling, fever, and redness) and related signs of sickness (lethargy). There is no consensus about what those chemicals are, but one theory holds that endorphins (the body’s natural “painkiller”) are involved. But wait! Isn’t inflammation a good thing because it encourages an unwell person to protect a wound from further damage and preserve energy so that the body can repair itself? Yes, but there’s more to it than that. There are different types of inflammation that are driven by two parts of the immune system. The innate immune system is a rapid response to bodily injury. It produces an inflammation called the acute phase response shortly after the injury begins. The acquired immune system eventually takes over to produce a different type of inflammation, one that plays a more decisive role in repairing the body. The release of chemicals from the brain is suspected of suppressing the first form of inflammation (the acute phase response), possibly because: (a.) the inflammation itself may be causing harm in some cases; or (b.) the innate immune system may be getting in the way of the activation of the acquired immune system. This chain of reactions helps explain why placebos only seem to work on injuries effected by the innate immune system.
It therefore stands to reason that placebos have a limited effect. Strong evidence confirms that a placebo can relieve pain, although it is not clear whether all kinds of pain are placebo responsive. What else? Well, that’s where it all gets rather sketchy. It is possible that placebos have an effect on ulcers and depression. There isn’t enough research to indicate whether other mental illnesses are effected. And that’s about it.
Evans’ more detailed explanation of how the placebo response works is a fairly shaky house of cards given the paucity of concrete evidence.
- Some of the foundational cards are deductions from evolutionary medicine, a nascent theory compared to evolutionary theory proper. Its use here may be an overstretch because the placebo response involves a sophisticated social activity (medical treatment administered by another person) that is almost certainly a recent development in human history—thus, too new to have evolved. It might be an evolutionary by-product of something else, but that “something else” is proving just as difficult to pin down. There are times when Evans admits to being “deep in the realm of speculation” on this count (pp. 104, 128).
- Some of the middle cards are inductions made from the ailments known to respond to placebos. The problem here is that, in some cases, ailments are deemed to be placebo-responsive based on only a single study or a couple of studies (some of which have serious methodological shortcomings). Some of the claims made about commonalities among ailments are also tenuous. Evans is scrupulous in the way he reports findings, yet he relies heavily on wily phrases while drawing tentative conclusions: “tantalizing hint”, “has some evidential value”, and “while by no means conclusive—do provide some basis for suspecting…” Phrases like these do pile up after only a couple of chapters.
- The cards at the top include conjecture about the workings of the brain which, despite many advancements in neuroscience, remains poorly understood. The scholarly disciplines from which Evans draws his ideas do not even share a vocabulary to describe brain activities. Our limited knowledge of cellular functioning poses a similar problem given that so much of the placebo response hinges on the operation of the immune system.
All told, Evans does a good job of putting together a coherent explanation from scant evidence. I suspect, however, that this explanation will not last long once future research inevitably pulls a card out of the house. Alas, such is the nature of scientific explanation in complex, immature fields of study. The vulnerability of this explanation does affect the plausibility of down-stream arguments made in the book. How do you apply such a shaky explanation with any degree of confidence? With lots of disclaimers, apparently.
One set of down-steam arguments is about whether the placebo response can explain the effects of alternative medicine, such as homeopathy, spiritual healing, chiropractic medicine, and the like. That is difficult to determine for several reasons. First, the alternative medicine movement (or is it industry?) has a pitiful track record of providing evidence of success using methodologically rigorous studies. If it were otherwise, these treatments wouldn’t be considered “alternatives” to scientific medicine, but scientific too. Second, some alternative medicines are not amenable to placebo and non-treatment trials. For example, how can a placebo version of acupuncture be administered? Give the patient a series of bum pinches and nipple twists? I think not. To make a long story short, there is only a small amount of dubious evidence suggesting that these treatments do better than placebos.
That said, alternative treatments may be particularly good at eliciting the placebo response compared to modern medical practices, according to Evans. Alternative treatments usually involve lots of direct contact with the patient (often including much touching). Explanations are elaborate and stuffed full of soothing healing/nature/purity talk. Benefits are promised with lots of reassurances. All of these things boost patient satisfaction and, to the impressionable among us, credibility. In contrast, mainstream doctors have less face-time with patients, go out of their way to demystify procedures, and offer a more realistic assessment of probable success. All of these things undermine the placebo response. Evans’ advice is for doctors to act more like alternative therapists—perhaps without all of the witch-doctor theatrics—in order to amplify the benefits of placebos. The ethics of this approach is not clear cut, although Evans does take a stab at it.
Evans, a former psychotherapist, goes on to look at whether psychotherapy is pure placebo. Short answer: yes, or maybe not even that.
What I find lacking in Evans’ book is a more direct discussion of sceptics’ concerns. With the possible exception of relieving pain, the case for the existence of a placebo response seems extremely shaky, at least based on the evidence Evans presents. Many studies are problematic because they are based on self-reporting. Diagnostic procedures for some conditions are far from perfect, creating large error rates in studies. Some evidence is anecdotal. This would suggest the need for a far more cautious book. I imagine it wouldn’t be as thought-provoking but it certainly would be more probative. The role of scepticism is also interesting because sceptics are increasingly reliant on placebo-based explanations to dismiss chicanery. Scepticism seems to have a stake in both sides of the debate.
As mentioned earlier, a shocking episode in our understanding of placebos is the story of how Beecher’s early research became gospel despite its massive flaws. I wonder if this example highlights how the medical sciences are particularly vulnerable to the inertia of research findings. There are a few plausible causes for this inertia. The enormous flow of medical research, coupled with the expense of gaining complete access to the stockpile of findings, means that many medical practitioners are out of the loop. The busyness of these practitioners makes it unrealistic to expect full digestion of up-to-date research findings—even when meta-studies summarize a literature. The decentralized nature of the health system in almost all jurisdictions means that there is no fast and easy way to communicate critical discoveries. (Of the ten thousand or so doctors in my province of residence, a large portion don’t even have a professional e-mail address; beeper yes, e-mail address no.) Finally, there are uneven patterns of on-the-job learning among medical practitioners because, as often happens with experts, they have too much confidence in their current knowledge and past training.
I tripped over another example recently while at a conference about managing pandemic emergencies. The example revolves around the “three foot rule” for the spread of airborne-contagious diseases. Early research from the 1920s claimed that people are safe from getting such contagious diseases so long as they stayed at least three feet away from someone who is infected. This is very, very wrong for many diseases. Yet, case studies of pandemic containment show that a large number of medical professionals still operate according to this erroneous rule. At its peak, this research finding had pretty much reached the status of folk wisdom. (Double Yikes!) Apparently, medical research corrections are not highly contagious.
In the final analysis, my conclusion is that at least another ten years must pass before we get a decent overview of placebos. This shouldn’t take away from Evans’ efforts. He did an excellent job given the evidence available. I look forward to future editions—maybe not the second edition, but certainly the fourth or fifth.
Why should the ordinary punter give a monkey’s about any of this? Um … I mean, how is this relevant to anyone other than medical and public policy researchers? We are all bombarded by news reports about the promise of new medical breakthroughs and infomercials about the promise of alternative treatments. Knowing more about placebos can help us put these claims into perspective. It’s too bad that more conclusive answers will have to wait a while.
Review by Peter Stoyko
Update (20.05.09)
Harriet Hall has an excellent overview of this topic entitled “The Placebo Effect”. It’s from the newsletter of the (U.S.) Skeptics Society. You can find it here.
Update (25.08.09)
Steve Silberman has written an interesting article in Wired magazine about how recent tests are apparently showing that the placebo effect is getting more powerful. He also presents an interesting chart showing how the physical form the placebo takes can increase the effect. You can find the article here.
REFERENCES