Georg Ivanovas From Autism to Humanism - systems theory in medicine
2.4 The placebo effect
The basis of current medicine and therapy is that of a specific disease, a disease that can be described in terms of a pathophysiology, either on a visible or on a cellular/molecular level (physiological or biochemical pathways, gene expression and so on). In order to treat the specific disease a specific therapy is designed.
This creates a frame of observation which is characterized by this specificity. Everything that is not describable with this approach is unspecific. There are unspecific diseases and unspecific therapies. Such unspecific effects are an integral part of every therapy (Walach/Sadaghiani 2002).
That is, a lot happens outside the defined frame of specifity, like conditioning, socialisation, expectation and other, just as seen before. It is impossible to exclude these unspecific factors. The exclusion, as has been shown, is an unspecific factor itself.
The relation between specific and unspecific therapy is the same as the relation between verum and placebo. In fact, they are identical only applied to drug therapy
That is the content of the most used definition of placebo by A. Shapiro: “We define the placebo effect as the non-specific, psychological, or psycho-physiological therapeutic effect produced by a placebo, or the effect of spontaneous improvement attributed to the placebo” (Shapiro/Shapiro 1907: 12).
The placebo effect happens always in an undefined area. This has to be well understood. The placebo effect is by definition the unexplained. There is no possibility to make statements on the placebo effect as such. Doing this leads to the known problems in the investigation of the placebo effect.
Of course it is possible to investigate certain parameters as pill size and colour, the way of administration and so on. By that these factors become specific leaving other factors in the unobserved state of unspecificity. In investigating ‘expectation’ certain expectation related results will be found. All other aspects will remain undefined. That is, statements on the placebo effect as a whole are meaningless according to the basic definition of the specific disease.
The unspecific nature of the placebo effect explains the contradicting results seen in placebo research:
The formalization of the placebo effect reveals another weak point in the usual design:
The specific model of a disease induces a specific therapy.
- The specific pharmaceutical medium is the verum (V).
- All effects not specific to its physiological mechanism are called placebo (P).
- As people are able to recover without being concerned about V and P there is also a spontaneous recovery (S).
This spontaneous recovery is logically a little bit strange. Is it a constant (like the radioactive process of disintegration) or does it depend on unspecific factors, than becoming an element of P? In fact S cannot be distinguished analytically from the placebo effect (Fields/Price 1997: 96-100), though it is something different. The placebo definition of Shapiro has taken this into account, although it is not precise in this regard. Thus, what is normally talked about is the so-called placebo response which represents P+S. That is, in the frame of the usual randomised trials no distinction between a placebo effect and a placebo response is possible. The only way to do this would be to compare it with an untreated cohort.. This has be done, for example, in a meta-analysis of antidepressant drugs. The result was (Kirsch/Sapirstein 1998):
- 25,16 % of the effects can be attributed to the drug
- 50,97 % of the effects can be attributed to the placebo effect
- 23,87 % of the effects can be attributed to the natural history
In order to be regarded as effective, a therapy has to fulfil the following requirement
V (+S) > P (+S).
In words: The verum (including spontaneous remission) has to be superior to placebo (including spontaneous remission). Or: “Subtracting mean placebo response rates from mean drug response rates reveals a mean medication effect” (Kirsch/Sapirstein 1998).
What does this imply? If we assume a linear process, we can find a lot of combinations. V, P and S might use the same therapeutic principles, work additively, complement or impede each other. This is not only a theoretical construct. For example, it could be proved that the opium and endorphin antagonist Naloxone is able to block placebo induced analgesia (Levine et al 1978). Naloxone antagonizes even placebo induced heart rate reduction, whereas propranolol antagonizes only the heart rate reduction but not the analgesia (Pollo et al 2003).
Theoretically, V + P + S could be added to create an extraordinary healing process. Naturally the version P + S - V = 0 (the verum impedes self-healing abilities of the body) is also possible and probably quite common. But every other combination is also thinkable.
Assuming linear relations everything can be neatly depicted. There are, for example, diagrams where an additivity of specific and unspecific effects, placebo effect, artefacts and regression is assumed (Walach 1999a).
But: Does all this arithmetic make any sense? Do such diagrams have any semantic meaning? Do such distinctions represent a kind of reality? Does all the agony around the placebo effect represent any truth? Or is it just a result of a certain study design? Playing around with certain vague notions?