Placebo - The Power behind the Throne?
Part 1
By John Reggars

Ever wondered why some chiropractors and osteopaths, who appear to employ the same treatment protocols, seem to have bigger practices and appear to have higher success rates with their patients’ than their peers? Ever wondered why each of the many and varied chiropractic and osteopathic techniques appear to have at least some success in treating all sorts of conditions? These questions are even further compounded when one considers that many of these techniques bare no resemblance to each other.

This article will briefly explore the therapeutic benefit of both the placebo and non-specific effects which may influence health outcomes in manual therapy. Part 1 will focus on what are termed non-specific effects of any therapy, while Part 2 will focus on the true placebo effect and how it can be harnessed and enhanced.

It could be argued that, based on the fact that some of these techniques have some scientific evidence to support their therapeutic benefit, that one practitioner has more skill than another in applying the same technique. But what about the techniques that have little or no scientific basis to their efficacy or have been proven to be ineffective? They must work in some way, otherwise their promoters would not get an audience and the patients of the practitioners who use them would seek alternative therapies.

The real answer may have little to do with the technique per se. It may have more to do with where and who is administering the technique and what happens during the therapeutic encounter rather than what is actually done to the patient. Could it be just a placebo response? Placebos are in essence a form of a treatment but without its substance. Something that looks like the real therapy but in itself is inert, having no therapeutic benefit.

Placebos have been used therapeutically by all sorts of healers for thousands of years. They are certainly still used today by both orthodox medicine as well as other health disciplines, either wittingly or unwittingly. However, it was not until Beecher’s seminal research in 1955 that this incidental phenomenon was taken seriously. It soon became an integral part of the benchmark of medical research, the Randomised Controlled Trial (RCT). Briefly, Beecher reviewed fifteen clinical research trials, which included a placebo treatment group, on a variety of diseases and found on average 35% of the symptoms of these diseases were satisfactorily relieved by placebos alone.

However, the often cited placebo effect of 35% may, in some cases, be substantially underestimated or overestimated. Beecher’s percentage was based on an average over the 15 trials examined. But some of his and other studies have reported placebo responses as high as 70-100%. A recent study on an ionized bracelet and a placebo of identical appearance found that on average all participants experienced a 29% decrease in pain regardless of whether they were wearing the ionized bracelet or the placebo.

It is also important to identify the “true placebo effect”, as other non-specific effects may significantly influence the course of any disease. These non-specific effects include such entities as natural history, spontaneous improvement, fluctuation of symptoms, regression to the mean, observer bias, answers of politeness, experimental subordination, conditional answers and questionable response variables.

The natural history of certain diseases may have a huge impact on the results of any clinical trial, followed over a critical period of time. Non-specific low back pain for example is known to have a limited natural history, with up to 90% of patients being significantly better within 6 weeks. It could be argued that, regardless of the therapy, if you can keep your low back pain patients in your care for at least this period of time then your success rate in the treatment of this condition should be close to 90%.

Some of the non-musculoskeletal conditions that chiropractors and osteopaths claim success with, such as infantile colic and otitis media, also have a limited natural history. Could it be that any improvement seen in these patients is due to natural history alone?

“Spontaneous improvement”, “regression to the mean” and fluctuation of symptoms” are also factors that can dramatically influence the course of many diseases. “Sick people often get better”. People tend to consult their health professionals when their symptoms are at their peak. In many diseases, this period of severity is short lived and improves even without treatment. Again, low back pain and otitis media are typical examples of conditions which are characterized by periods of remission and exacerbation.

“Answers of politeness” and “experimental subordination” also influence health outcomes. A polite answer refers to patients who report improvement to their therapist just to please them, even when no change has occurred. In fact the word placebo literally means “to please”. It is for this reason that well controlled studies usually have someone other than the researcher collect any data from the research subjects. It is also why treatment outcome measures are an invaluable tool in accurately accessing a patient’s response to the treatment provided.

Conditioned answers or effects must also be considered in this context. Like Pavolv’s dogs, patients may be conditioned for a specific response. The “chiro junkie” who comes in for treatment once or more per week for their “fix”, regardless of how they feel. Or the patient who is not satisfied until they hear the sound of a joint cavitation. These could both be examples of this effect.

Questionable response variables may also have a profound effect on the interpretation of the results obtained from both placebo and real therapeutic interventions. This effect refers to response variables that are irrelevant to the condition in question. For example one study on multiple sclerosis claimed a 73% placebo effect, but on later analysis of the data there was no objective change in the patient’s neurological condition. The 73% improvement related to purely subjective criteria such as increased euphoria, strength and agility. Could this be the same reason why manual therapists and patients alike claim success in the treatment of asthma? Several research trials, involving the treatment of bronchial asthma with spinal manipulation, have not shown any significant objective improvement in lung function. Yet many subjects report they require less medication, can breathe more easily and had a significant improvement in feelings of well being, etc.

Without the recognition of and provision for, these non-specific effects, the validity of any clinical research trial must be questioned. In 1997 when Kiene Kienle re-analysed Beecher’s classic paper, taking into account these non-specific effects and looking more closely at the methodology, data etc., they demonstrated that, with the exception of bronchial asthma, all the trials in which Beecher claimed a significant placebo effect could be explained by either the presence of these non-specific effects, misquotations or poor methodology.

References available from the editor on request.