Some Thoughts on Efficacy Beyond the Placebo Effect

One of the most frequently asked questions concerning complementary and alternative medical interventions is whether they are anything more than placebo effects. This question could more accurately be stated as follows: are any positive effects produced by an unconventional medicine intervention greater than the effects of a sham intervention in a randomized clinical trail (RCT)?

The question whether an intervention has greater effectiveness than a placebo was asked for the first time in the medical literature before World War II. It became a common question with the systematic introduction of the randomized controlled trial after World War II. Only from the early 1960s, with the phocomelia produced by thalidomide, did the comparison of an intervention with placebo become the defining and legally necessary question for scientific medicine and governmental regulatory agencies. The biomedical concept of efficacy underwent a gradual and subtle shift. The question of efficacy evolved from demonstrable clinical effects and gradually become an issue of a demarcation line defined by a placebo intervention in a clinical trial.

In fact, we need to remember that there is more than one approach to this question and that many issues are still being debated. The most common approach to this question is the perspective of fastidious efficacy, a name coined by Alvan Feinstein. Assuming that clinical outcome "is equal to its active effect plus its placebo effect", the fastidious approach takes an additive/subtractive approach for defining the boundary line of efficacy. The assumption of fastidious efficacy -- embedded in the adoption of the RCT -- both creates the question of what is beyond the placebo and supplies the answer with its "blinded-placebo-controlled" approach. Its answer is quantitative and precise and has the objectification that a chemo-mechanistic biology requires. Whether this information accurately portrays clinical medicine is problematic as the method assumes that verum and non-specific effects are stable, separable, linear and relatively constant during a clinical trial.

Most reviewers of acupuncture's efficacy using this perspective tend to be either slightly positive or inconclusive. Unfortunately, most of the clinical trials upon which these opinions are based have major flaws. For example, it is unclear whether the quality of acupuncture, the quality of therapist and the length of time for a reasonable expectation of outcome were adequately taken into account when these randomized clinical trials were performed. A second approach towards approaching this question is the perspective of pragmatic efficacy, a term created by Schwartz and Lellouch. This model of randomized clinical trials assumes that various non-specific effects are interactive among themselves and among any verum effects and that is, therefore, difficult to make a clear distinction between verum and placebo. The pragmatic approach blurs the verum-placebo dichotomy embedded in the question, generates a less scientifically-satisfactory answer but may be of greater clinical relevance. The growth of the outcomes research movement is consistent with this pragmatic efficacy perspective. I am unaware of any genuine pragmatic trials where the practitioner is able to practice without any restraints on "optimal" treatment for acupuncture. (There are many hybrid-pragmatic trials where acupuncture is compared to another real treatment, but all other factors are kept equal, so that it is not "optimal" acupuncture.) The pragmatic trials in other complementary fields such as manipulative therapy are very positive and there is a need for genuine pragmatic trials in acupuncture.

A third approach to the question is what Tambiah calls "performance efficacy." This approach is cultural and anthropological and accepts symbols, belief, suggestion, expectation and persuasion as central to illness and health. This approach avoids any reliance on objectivity, controls and measurement. The contrived environment of an experiment might be the only clearly labeled sham from this perspective. This approach raises the question of who decides what is a sham. Performance efficacy is grounded in the language of solidarity, holism and unity; it is immune to the distancing and neutrality requirement of randomization, blinding and p values. This method generates clear answers about the experience of a self-selected, biased, individual person and has nothing to do with generalizability and replicability.

This approach turns the question of what is beyond the placebo on its head -- instead, it asks: who decides whether a therapy is merely sham or more than sham? Patient satisfaction surveys of complementary medicine point to the possibility that one answer to what is beyond the placebo is that sham intervention is defined in the eyes of the beholder -- who may or may not recognize the scientific method as the final arbitrator. Generally speaking, patient survey of satisfaction with alternative medicine is very high, always in the range of 80%.

There are no published surveys of acupuncture or Oriental medicine patients that I am aware of, but I suspect the satisfaction rate is equally high. The performance perspective also raises fundamental questions as to whether or not randomization introduces significant bias; whether informed consent itself is a potent non-specific effect and whether patient preference itself can have decisive clinical effects.

References