Note: I originally wrote this for Science-Based Medicine, but Dr. David Gorski has posted a far more detailed review of the study, which I encourage everyone to read, so I am just posting my thoughts on the subject here.
The placebo effect is a controversial, and in my opinion deeply misunderstood, phenomenon. At Science-Based Medicine, there has been plenty of discussion and debate on the subject before, and different authors have slightly different perspectives (for example, Drs. Novella, Gorski,Ramey and pharmacist Scott Gavura). A recent study in the New England Journal of Medicine (NEJM) serves to illustrate nicely my own view of the nature of this phenomenon, so I thought I’d have a go at characterizing it.
The placebo effect is often claimed to be a real improvement in a patient’s disease that is caused by an inert treatment, purportedly through the power of the mind to heal the body when the patient believes they are receiving an effective therapy. This is certainly how alternative medicine advocates characterize it when scientific studies identify their interventions as no more effective than placebos. However, in my opinion this is not an accurate characterization of the placebo effect.
There are many non-specific treatment effects seen in clinical trials, including the placebo, and all create the perception and appearance of improvement in patients given a fake therapy which is actually doing nothing to alter the underlying medical condition causing the patients’ symptoms. Some such effects are artifacts of the research setting, like the Hawthorne Effect. In this phenomenon, people take better care of themselves when they are participating in a research study and being closely monitored, and so they tend to improve regardless of whether the treatment being studies is helping them or not.
Other non-specific treatment effects, like the placebo, are illusions created by the belief and expectations of the patient or investigators. They do not represent an unlocking of deep and powerful forces of mind over body. They simply trick us into feeling or looking better when our condition is really unchanged.
One can argue, of course, that if a patient perceives inherently subjective symptoms, such as pain or nausea, to be improved, then they must really be improved. There is some truth to this, but the argument can easily lead to exaggerating the benefits of fake treatments which don’t actually affect the patient’s physical health. Homeopathy may make a person experience his or her pain less intensely, but a real medicine will achieve this, usually to a greater extent than a placebo, by actually reducing the pain generated in the body as well as the perception of pain in the mind. Taking advantage of placebo affects that accompany all real medical treatments is fair enough, but substituting alternative therapies that are nothing but placebo for real medicine is not in the best interest of patients.
And this is doubly true for veterinary patients. Lacking beliefs and expectations about their health, my patients cannot experience the placebo effect per se. Some improvement in symptoms may occur as a result of conditioning and the human contact associated with therapy, but in general animals don’t enjoy the “benefit” of being fooled by their own minds into thinking an inert treatment has made them better. Unfortunately, animal owners and veterinarians are subject to being misled by non-specific treatment effects, and the “placebo-by-proxy” effect is quite potent. It creates an especially dangerous situation in which an animal’s caretakers believe they are helping the patient and the combination of cognitive errors and cognitive dissonance prevents them from appreciating that the animal’s condition is actually not improved.
The new NEJM paper demonstrates both how powerfully our own beliefs and expectations can fool us into feeling better and how careful we must be to distinguish this from actual improvement in our condition. The full article is behind a paywall, but a short summary is available.
Wechsler, ME. Kelley, JM. Ph.D. Boyd, IOE. Dutile,S. Marigowda, G. Kirsch, I. Israel, E. Kaptchuk, TJ. Active albuterol or placebo, sham acupuncture, or no intervention in asthma. N Engl J Med 2011; 365:119-126
Forty-six patients with asthma were randomly assigned to receive no treatment, or to treatment with an inhaled bronchodilator (albuterol, which opens up airways constricted by asthma), a placebo identical to the bronchodilator treatment, and sham acupuncture presented as “real” acupuncture (whatever that is). All patients were rotated through all the treatments and assessed after each, and the patients and investigators were blinded to the treatments insofar as possible. Obviously, patients and researchers could tell the difference between no treatment, an inhaler, and acupuncture. And as is always the case in studies of acupuncture, the acupuncturist undoubtedly knew that they weren’t providing what they would call genuine acupuncture treatment.
The patients were evaluated both in terms of their own perceptions of any improvement in their symptoms as well as an objective measure of their respiratory function. When given no treatment at all, patients reported an improvement in their symptoms of 21%. This likely represents a variety of non-specific treatment effects, such as the natural history of the disease, the aforementioned Hawthorne Effect, and so on. It also illustrates the value of a no-treatment group in clinical trials, when it is ethically possible to have one. This is an element of clinical trial design that sometimes gets inadequate attention in discussions of the pitfalls of medical research, though there is evidence that when such groups are included, the purportedly powerful effects of the placebo phenomenon become less impressive.
There was no difference in the improvement perceived by patients regardless of which of the other treatments they received. Between 45-50% improvement was reported after treatment with the albuterol inhaler, placebo inhaler, and sham acupuncture. So patients experienced nearly twice as much benefit from some treatment as from no treatment, but it made no difference whether the treatment was real or inert. This would seem to suggest that the placebo is a potent therapy and as good as a physiologically active medicine. After all, asthma symptoms are quite recognizable and uncomfortable, so it seems unlikely that such high rates of improvement would be only a function of the patients’ imagination. Perhaps there is something to this mind-body medicine idea, and alternative therapies really can make us better through placebo effects even if they don’t have any other effects?
Well, not so fast. In addition to asking the patients how they felt after each treatment, the investigators also measured their lung function, using an instrument that records, among other data, how much air the patients could force out of their lungs in a given period of time. It turns out that this objective measure showed a 20% improvement with the bronchodilator inhaler, but a significantly lower 7% improvement with the inert therapies or no treatment at all. So while the patients couldn’t tell the difference between real and fake therapies, their lungs certainly could.
This illustrates what I think is the key point to understand about the placebo effect: It can make us feel better without actually affecting our physical health. In any discussion of the placebo effect, we must bear this fact in mind. There is no magical power of mind over body involved. Placebo treatments affect our perceptions of our symptoms, but they do not improve our underlying medical condition.
It is a desirable and appropriate for doctors to make their patients feel better, but it is not enough to do so only in ways that leave the physical health of the patient fundamentally unchanged. We owe it to our patients to apply rigorous scientific methods to investigating our therapies and to developing interventions which actually treat the cause of the symptoms, not just the patients’ perceptions of them.
There is nothing wrong with taking advantage of placebo effects that inevitably accompany therapeutic interactions. But I believe it is unethical to employ therapies which are demonstrated to have only placebo effects, particularly when interventions that actually treat the underlying medical condition are available. And as a veterinarian, I am strongly opposed to using treatments that clearly work in humans only through belief and expectation, since these are very likely to mislead owners into believing the therapy is working without actually benefitting the patient.
This study illustrates both the power of the placebo effect to make us experience marked improvements in seemingly undeniable disease symptoms and also the powerlessness of the placebo effect to improve our physical health. It shows how easy it is for us to be fooled concerning the effectiveness of a medical therapy by our personal experiences as patients and doctors, as well as how easily a clinical study can generate the false impression of a beneficial therapeutic effect if placebo and no-treatment controls and objective outcome measures are not employed. Finally, it provides a point of evidence against the currently popular contention that even though some alternative therapies cannot be shown to have effects greater than placebo, they can be viewed as having real benefits by means of the placebo effect and the power of mind over body. At best, placebo effects can alter our perceptions of our symptoms to reduce our discomfort. But they cannot fundamentally improve our physical health, and there is real danger in believing they can if it leads us to rely on therapies that are only placebos.
Regarding the Hawthorne Effect. Do you know of anywhere online that has a comprehensive summary of all these sorts of effect (placebo, confirmation bias, regression to mean etc.) that may explain perceived improvements? (Or have you made such a list yourself anywhere here?) It would be useful to have in discussions.
Keep up the good work!
I am not aware of a comprehensive list, but here are a couple of useful sites:
Why Bogus Therapies Seem to Work
The Skeptic’s Dictionary
Cognitive BIases
Let me offer you a quote of mine from a thread on SBM:
“………the concept of treating different patients with individualized therapies within the same clinical trial already exists. In gene therapy studies, subjects are given remedies based on their own personal genetic makeup. The studies, for the most part, are testing the platform for delivery and the technology behind the identification, isolation and development of each biological drug rather than the specific drug itself. In homeopathy, much like gene therapy, treatments are given based on the disease/disorder as well as the patient’s individual needs instead of the traditional “one application one drug” pharmaceutical approach. So, for there to be an effective clinical trial, homeopathy would need to be tested in a similar fashion as DNA/genetic biologics since a system is being tested, not a specific formulation of a specific drug.”
Gene therapy is an interesting and promising area of research, but not a well-developed therapetic approach with a wide range of applications like traditional pharmaceuticals. The idea that we could individualize therapy in a meaningful sense is a good one, but we are a long way from achieving it, and in the meantime, the equivalency you seem to be trying to establish between this approach and homeopathy is false for a variety of reasons.
For one thing, the individualization of therapy in clasical homeopathy is based on criteria which are pure fantasy, so the fact that a unique fantasy diagnostic and treatment plan is created for each patient is irrelevant. I can individualize a dream interpretation or horoscope for everyone I meet too, but that doesn’t mean what I’m providing for them has any reality behind it. Gene therapy is at least based on plausible scientific principles, and if it shows anything it is how difficult and labor/time intensive it is to approach medicine this way and how much rigorous investigation is needed before one can be effective. Homeopathy is based on a fanciful notion and a series of “experiments” correlating symptoms witht therapies that are unsupported by any plausible scientific principles, so individualized nonsense is still nonsense.
Secondly, clinical trials have been done in which homeopaths are allowed to follow the rules of classical homeopathy and prescribe their variously labeled water-soaked sugar pills based on their imaginary diagnoses. These have not shown any meaningful superiority for this approach over any other, so there is little evidence to suggest the individualization process accomplishes anything.
Finally, the reality that this line of thinking ignores is how stunningly successful the scientific approach to limiting variables has been in developing effective therapies. If one focuses solely on the limitations of medicine and what one cannot achieve or what has not yet been achieved, one can try and argue that the fundamental approach is flawed. However, when one looks at the totality of healthcare interventions and how they have developed since the scientific approach began to be employed, the unprecedented success is obvious, and one has to admit that the appproach is the best game in town. Perhaps it will eventually be replaced by something better that is fundamentally different, rather than simply the result of incremental improvements. But that remains to be seen, and it is irrelevant to the manifest failure of homeopathy as a therapetic approach.
Pingback: The Harm Complementary and Alternative Medicine Can Do | The SkeptVet
Pingback: Acupuncture Works! (Just as Well as Fake Acupuncture Anyway) | The SkeptVet