It is ridiculous to assert that the placebo effect does not exist.1

We found little evidence in general that placebos had powerful clinical effects.2

The placebo effect, thought of as the result of the inert pill, can be better understood as an effect of the relationship between doctor and patient.3

Thus has the placebo been transformed in a few short years from a sham in medical practice and a control agent in clinical trials to a therapeutic ally.4

Most practitioners questioned in this study continue to use placebos.5

It is suggested that in select cases, use of placebo may even be morally imperative.6

These quotations are from recent articles debating the existence and the nature of the placebo effect. The discussion was brought up a decade ago, when Gøtzsche7 suggested that we should discard the term placebo and Kienle and Kiene8 questioned Beecher’s conclusions, in 1955,9 that placebos, understood as inert substances used in clinical studies, did have a considerable effect on patients’ well-being.

The recent debates have shown that the problem of the placebo effect seems to be partly semantic. There is a rather wide range of meanings for placebo and placebo effect, and some of the controversy has arisen when the proponents of various positions have ignored each other.10

Our aim in this essay is to clarify some of the conceptual issues related to this problem. We first consider the various uses of placebos as inert substances or treatments. We then examine the problem of the placebo effect. Finally, we suggest a terminological solution for the debate in order to put the argumentation on the nature and existence of the placebo effect on firmer ground.

THE MANY USES OF PLACEBOS

Placebos, understood as inert substances or treatments, may be used in medical practice and research in the following ways.

Deliberate deception

The clinician deliberately deceives the patient by claiming that the treatment works, although all available evidence is against any effect and the provider of the treatment knows it. This is the old, narrow meaning and use of placebos, which reflects the paternalistic spirit of earlier medical encounters. The practice is contrary to today’s medical ethical codes, which emphasise truthfulness in all professional interactions.11

Open disclosure

The physician tells the patients openly that they will receive an inert substance (eg, sugar pill) but that it may help them. This practice has been used in research settings and is, in fact, one possibility suggested by Lichtenberg et al., who discuss the ethical use of placebos in clinical practice.6

Good faith

The physician or the patient believes that the treatment is effective, although all available evidence is against it. For example, there is strong scientific evidence that cough medicines are not clinically effective,12 but their market is still huge.

Deception in research

The patients are told, falsely, that they will receive a medication or other active treatment. Although ruled out in contemporary biomedical research, this practice was not uncommon in the first two decades of randomised controlled trials.13

Genuine informed consent in research

The patients are told that after randomisation they may end up in a placebo group; they understand this alternative and give their informed consent. Although the principle of informed consent is imperative in contemporary research, the level of understanding among research participants is far from satisfactory. For example, a study in Lithuania showed that almost half of the participants in clinical trials thought that the aim of the trial was to improve their health, and almost 40% in one trial thought that their physicians knew exactly what (active drug or placebo) had been administered to their patients.14

PLACEBO EFFECT

In the light of the preceding examples, it is obviously problematic to speak of one placebo effect or of a uniform placebo effect, since the settings for the use of placebos vary widely. This is especially so because the often-quoted figure from Beecher’s classic paper,9 that 35% of patients exhibit placebo effects, has been shown to be an estimate without any scientific justification.

Kienle and Kiene8 re-evaluated the studies Beecher referred to in his 1955 article9 and claim that there is no scientific support whatsoever for Beecher’s arguments. They conclude that the apparent placebo effect “can be fully, plausibly, and easily explained without presuming any therapeutic placebo effect”. They support their argument with a detailed analysis of the material Beecher had used, and to our knowledge their conclusions have not been refuted as yet.

However, the idea that the therapeutic process itself would not have any effect on the therapeutic outcome, in addition to the expected effects of a particular treatment, is counterintuitive. Kienle and Kiene also appreciate this aspect, but in a roundabout way. They quote a study supporting the possibility of a true placebo effect but try to rescue their case by writing that “this was an example of a psychosomatic effect, not the effect of placebo application”. In another example, they acknowledge the effects of suggestion on asthma in experimental conditions, but they do not regard the result as proof of a placebo effect. Acknowledging psychosomatic phenomena and suggestion as effective components in treatment leaves, in fact, the question of the existence and nature of the placebo effect open. The authors consider the possibility of adopting the term “non-specific effect” as a substitute, but discard it as self-contradictory.

There have been attempts to clarify the issue of the placebo effect by renaming it. Benson and Friedman15 suggested remembered wellness, a term “chosen because ultimately the evocation of the placebo effect depends on central nervous system events that result in feelings of well-being”. The authors do not, however, deepen their suggestion with further analysis, and their suggestion remains somewhat cryptic.

Papakostas and Daras16 treat the placebo effect as “the response to the healing situation”. In our opinion, this phrase addresses the problem correctly but is too long to be useful. Moerman and Jonas17 suggested that much of what is called the placebo effect is a special case of the meaning response, which is defined as the physiological or psychological effects of meaning in the treatment of illness. This choice, however, describes not so much the effect as the nature of the outcome, as the term meaning response itself suggests.

FROM PLACEBO EFFECT TO CARE EFFECT

Since consistent definitions of the concepts placebo and placebo effect seem unattainable, Gøtzsche suggested that the concept of placebo should be discarded altogether.7 He argues that the focus should be on the choice of outcome measure and the magnitude of the effect rather than on interventions that are difficult to define. We appreciate his proposal to discard the concept, but we wish to develop the issue in a more constructive way.

We suggest that the concept of placebo should be limited to a research context only. The term placebo would refer, thus, only to the procedures that are used as inert controls for so-called active treatments in medical research.

When inert or only vaguely effective substances or treatments are used in a clinical context, they should not be called placebos. If a method of treatment is ineffective in its own right, let it be called an ineffective treatment for that particular patient or problem. There is no reason to adopt a particular concept for that in a therapeutic process.

But when we refer to the phenomena that take place during the consultation and lead to beneficial therapeutic results while the treatment given has no apparent causal connection to the outcome, we suggest that that phenomenon be called, not a placebo effect, but a care effect, for the following reasons.

First, the term placebo effect is contradictory by definition. If something is inert, it cannot have an effect. This confusion can be found, for instance, in the opening chapter of Arthur and Elaine Shapiro’s widely cited book The powerful placebo.13 They define placebo as “any treatment … that is used for its ameliorative effect on a symptom or disease but that actually is ineffective or is not specifically effective for the condition being treated”. Then placebo effect is defined as “primarily the nonspecific psychological or psychophysiological therapeutic effect produced by a placebo, but may be the effect of spontaneous improvement attributed to the placebo”. If the word placebo in the latter definition is replaced with its definition, we end up with the following: “... therapeutic effect produced by any treatment ... that actually is ineffective or is not specifically effective for the condition being treated … may be the effect of spontaneous improvement attributed to the [treatment]”. The first part makes no sense and the second part limits the therapeutic result to spontaneous improvement only and rules out all other components of the therapeutic encounter.

Second, a care effect is necessarily present in any therapeutic encounter, providing the patient is not unconscious. Whatever the treatment, the patient is being treated. Thus, when a patient reports subjective responses that cannot be fully explained in terms of the supposed mechanisms of the treatment given, that may be considered to be the effect of being treated or cared for. As the late RD Laing once remarked, “there is nothing that affects our chemistry more immediately than other people”.18 That is, for all of us, just being acknowledged, heard, understood, assured and comforted can be very alleviating in itself. This can be considered, ultimately, a substantial part of any treatment, irrespective of the care giver’s frame of reference, be it that of a scientifically educated physician, an alternative therapist or an indigenous healer.19 20 The care effect may be small or even negative with respect to the therapeutic intention, but it can never be excluded from the therapeutic encounter. A care effect may be present in a clinical trial, too, but, as our classification of the different uses of placebo demonstrates, the clinical context and the research context are fundamentally different settings. If present in a clinical trial, a care effect can be considered a confounding factor, not an ally as in clinical medicine.

Third, the term placebo has many unscientific and pejorative connotations in the clinical context, which have infiltrated also into the concept of placebo effect, regardless of what has been written about the importance, existence and supposed mechanisms of it. For example, it is common to use expressions such as “it is merely a placebo effect” when reference is made to positive outcome of a treatment that the speaker does not consider theoretically plausible. In contrast, the terms care and caring have neutral or even positive connotations (eg, of concern and responsibility). They are free of the burden that placebo and placebo effect have acquired as referring to something that is somehow not real. The term care effect respects the reality of the effect of care and suggests a need for more research into the multitude of phenomena behind it.

Fourth, in the present literature discussing the placebo effect in the clinical context, the term could in most cases be readily replaced by care effect without causing any damage to the basic argumentation. That replacement would, in fact, clarify the discussion, since the concept care effect is free of the connotations and contradictory meanings of the term placebo.

In the research context we may replace the term placebo effect with the expression “the effect in the placebo group” when discussing the outcome of the research. That would free us from using such an illogical misnomer as placebo effect in medical science.