I thought I might start developing chest pain when I read it, but to my shock NCCAM has actually funded some worthwhile research! Even more amazingly, NCCAM described it in a press release!
Too bad it supports the contention that acupuncture is nothing more than placebo and that the attention given by the practitioner is what really accounts for much of the perceived therapeutic effect that patients attribute to it.
The press release to which I refer leads to a rather interesting study that examines the components of the placebo effect. The article, published online yesterday in BMJ, was entitled Components of placebo effect: randomised controlled trial in patients with irritable bowel syndrome and came out of a research group at the Osher Research Center, an “alternative medicine” institute run out of Harvard University and, as you might recall, the only one in my Academic Woo Aggregator that at least talked the talk in a semi-convincing way about applying science and evidence to so-called “CAM” therapies. Also involved in the study was the University of North Carolina Chapel Hill and the University of Hull in the U.K.
The study looked at the placebo effect in patients with irritable bowel syndrome. IBS was chosen because the waxing and waning of symptoms are often quite subjective and usually not clearly related to any objective measures of pathology. Indeed, IBS is generally a diagnosis of exclusion, to be made after other causes for a patient’s symptoms are ruled out, which makes it a particularly good syndrome in which to study placebo effects. It’s also fairly poorly understood and frustrating for physicians who treat it and even more frustrating for patients–again, characteristics that often drive patients to try unconventional therapies for it. Its experimental design was relatively simple. Patients suffering from IBS were recruited and randomized to one of three groups. The first group of patients (group 1, the observation group) was simply observed. They were evaluated for symptoms at the beginning and then again after three and six weeks. Their evaluation visits were the only contact the subjects in this group had with the researchers.
The second group of subjects received placebo acupuncture alone. This was done using well-validated sham needles described in previous studies described right here on this blog. This group (group 2) was called the “limited interaction” group because that’s all the subjects received. Treatments consisted of twice a week sessions, and the acupuncturists were trained not to interact with the patients. Indeed, they were instructed to tell the patients that, because this was a research study, they were not allowed to discuss the treatment or even to converse with them. All they did was to place the sham acupuncture needles and leave the room for 20 minutes, after which they came back to remove the needles with as little interaction with the subjects as possible.
The final group of subjects (group 3, augmented interaction) underwent the same acupuncture treatments as patients in group 2 under the same conditions and in the same room, except that the practitioners interacted with them extensively. The difference was that subjects experienced an “augmented” patient interaction (45 minutes) at the first visit in which detailed questions about their symptoms were asked, including how their IBS symptoms affected their lifestyles and relationships. The interviewers asked if subjects understood the “cause” and “meaning” of their conditions and at each visit incorporated at least five primary behaviors, including a warm, friendly manner; active listening (repeating the patient’s words and asking for clarifications); empathy; 20 seconds of thoughtful silence while feelign the patient’s pulse or contemplating a treatment plan; and communication of confidence and positive expectation. At each visit, the acupuncturist would place the placebo needles and then leave the subject in a quiet room for 20 minutes. When the practitioner came back to remove the needles, he or she also made sure to exchange some words of encouragement.
Every subject in every group underwent three weeks of therapy. After three weeks, subjects in groups 2 and 3 were re-randomized to either group 2 or 3 again without their knowledge, and the experiment continued for another three weeks. At three weeks and six weeks, scales measuring global improvement, adequate relief of symptoms, symptom severity, and quality of life were measured. The overall rationale for this experimental design, of course, is to test three postulated parts of the placebo effect, the response to treatment and assessment (Hawthorne effect), the patient’s response to a therapeutic ritual (placebo acupuncture), and the patient’s response to the patient-practitioner interaction. The investigators wanted to isolate these aspects of the placebo effect as much as possible, using a model of sham acupuncture for IBS as their model.
The results, as you might imagine, were not surprising. Global improvement scores were lowest in the observation group, increased in the limited group, and increased still more in the augmented group. The same was true of the symptom severity scores, only in reverse. This score decreased from observation to limited to augmented, indicating improvement of IBS symptoms, results that persisted to six weeks. This was also likely to be significant clinically, because a decrease in the symptom severity score of 50 reliably indicates improvement of symptoms and approximately 60% of patients reached this level of improvement. The results showed a statistically significant trend, and suggested that a ritualized treatment provided some placebo effect but that the practitioner-patient relationship was capable of producing the most powerful placebo effect. The investigators tested for successful blinding, and similar numbers of patients in groups 2 and 3 (around 80%) believed that they were getting effective acupuncture.
Remember, now, that none of these patients received “real” acupuncture, only sham acupuncture. Basically, the main variables being studied and measured were the effects of ritualized treatment and the practitioner-patient relationship on patient perception of symptomatic relief. Strengths of the study included its design and the use of a validated sham acupuncture. One weakness that bothered me was that the statistical analysis looked at the trend; I’m not sure that was the best way to analyze the data, and the overlap in the error bars between the observation group and the augmented groups in some of the graphs and what, to the naked eye at least, looks to be a lack of statistical significance when the observation group is compared to the limited group. One other concern is that there was, by necessity, some deception involved. Potential subjects were not fully informed of the true purpose of the study because doing so might influence the results. This sort of deception is always at least somewhat dicey from an ethics standpoint. Indeed, the design of the trial incorporated a nested substudy of acupuncture that allowed the investigators to tell subjects truthfully that, if they were randomized to an acupuncture group, they had a 50% chance of receiving “true” acupuncture. Presumably the results of that substudy will be published in a different manuscript. It should be pointed out that my concern about the deception involved was somewhat alleviated by the fact that subjects received letters after the study was over telling them the full purpose of the study and offering them the opportunity to withdraw consent to use data derived from their participation.
Concerns aside, I tend to believe the study, as its strengths sufficiently outweigh its weaknesses, and it adds to a long line of evidence suggesting that the practitioner-patient relationship is very important to the placebo effect. It also adds yet another bit of evidence consistent with the contention that whatever effects perceived to be due to acupuncture are placebo effects. What’s odd about this study, though, is how NCCAM spins it:
Simply providing placebo acupuncture yielded modest improvement of IBS symptoms over assessment and observation alone. Thus, the three components of a medical encounter can be progressively added to produce incremental symptom improvement.
Which is all that virtually all of so-called “complementary and alternative medicine” is: placebo. That’s the irony. NCCAM is touting this study, and it’s one of the only decent studies I’ve ever seen funded by NCCAM. Unfortunately for NCCAM, though, it’s one fo the best indications yet of why “CAM” appears to work for so many patients: The placebo effect.
Kaptchuk, T.J., Kelly, J.M., Conboy, L.A., Davis, R.B., Kerr, C.E., Jacobsen, E.E., Kirsch, I., Schyner, R.N., Nam, B.H., et al, . (2008). Components of placebo effect: randomised controlled trial in patients with irritable bowel syndrome. BMJ, online first(April 7), 1-8.