Placebos in practice
BMJ 2004; 329 doi: https://doi.org/10.1136/bmj.329.7472.927 (Published 21 October 2004) Cite this as: BMJ 2004;329:927All rapid responses
Rapid responses are electronic comments to the editor. They enable our users to debate issues raised in articles published on bmj.com. A rapid response is first posted online. If you need the URL (web address) of an individual response, simply click on the response headline and copy the URL from the browser window. A proportion of responses will, after editing, be published online and in the print journal as letters, which are indexed in PubMed. Rapid responses are not indexed in PubMed and they are not journal articles. The BMJ reserves the right to remove responses which are being wilfully misrepresented as published articles or when it is brought to our attention that a response spreads misinformation.
From March 2022, the word limit for rapid responses will be 600 words not including references and author details. We will no longer post responses that exceed this limit.
The word limit for letters selected from posted responses remains 300 words.
Farre et al raised the question of off-label drug usage in their
discussion about the use of placebos. I beg to differ that the use of a
drug in a non-approved dose, frequency, duration, or even route
necessarily constitutes unconscious use of placebos, in that it
presupposes that the use of a drug in a non-approved manner must therefore
be ineffective, or harmful.
Whilst I am not sufficiently familiar with the laws regulating
medicines in various countries, I am given to believe that doctors do not
necessarily get punished if they use drugs in a manner not fully compliant
with licencing conditions. If there is sufficient clinical trial
literature that a certain unapproved method of drug use is therapeutically
advantageous, then the problem of negligence may not arise, even if such a
way of using a drug has never and is not approved anywhere in the world.
One instance of this type of unapproved (by the regulatory bodies)
but well established (by body of clinical trials) use of an anaesthetic
drug is the epidural administration of fentanyl. Although the tide of
opinion may well be turning against it now, it had enjoyed enormous
popularity amopngst the anaesthetic communities all over the world. Surely
this is not unconscious placebo usage.
It is perhaps salutary to note that in a discussion on off-label drug
usage, it has been quoted that up to 30% of all prescriptions in General
Medicine is off-label, ie fails to comply fully with licencing conditions.
Competing interests:
None declared
Competing interests: No competing interests
The placebo effect is undoubtedly real and it is right to use it but
both patients and their healthcare providers must acknowledge this and
their employment of it.
Doubts may have been raised about the usefulness of the placebo response
in conditions other than chronic pain but this, in fact, has very much in
common with other chronic disease states where many interventions without
good supportive evidence (and significant side effects) are daily advanced
and it should, therefore, be of worth in them too.
Guidelines for placebo use are appropriate as we, indeed, cannot afford to
dispense with any treatment that works even if we may not be certain how
it does; what we must do, however, is to confirm efficacy.
Considerable controversy may exist about the use of a biologically inert
or irrelevant substance with therapeutic intent but isn’t such use, in
fact, a daily occurrence? Doctors may not actively prescribe them but
many patients make use of them particularly in the form of homeopathic
remedies.
Homeopathic remedies are the best examples which, might I suggest, have
not, in fact, been properly tested. I am not aware of a single study in
which the homeopathic remedy itself has been tested as the only variable
(though one could be set up without difficulty) and, in this age of
evidence-based medicine and practice, I am concerned about the possible
untested extension into the NHS mainstream as has been recently proposed
by the Welsh Secretary for possible piloting in my region.
Competing interests:
None declared
Competing interests: No competing interests
In his editorial on placebo (1), David Spiegel writes that our meta-
analysis that compared placebo interventions with no treatment "had
numerous problems", including lumping of heterogeneous trials. This all-
embracing criticism is unwarranted.
We have previously responded to Spiegel’s comments (2) and explained that
a broad approach to meta-analysis is often appropriate (3), in particular
when there is no good a priori reason to exclude some conditions from
consideration. And we also analysed disease conditions separately, if they
had been investigated in at least three independent trials.
The recently updated review now includes 182 trials, with binary outcome
data for 4284 patients, and continuous outcome data for 7453 patients (4).
Thus, the relatively small loss of power by our preference for the
clinically more relevant binary outcomes cannot explain why our results
were largely negative.
It is correct that we reported a possible modest effect of placebo on
pain, but Spiegel fails to mention that we also reported that the effect
decreased significantly with increasing sample size, indicating possible
bias (5). In addition, as studies comparing a placebo intervention with an
untreated control group cannot be blinded, positive results were likely
influenced by reporting bias. Thus, Spiegel’s optimistic interpretation is
not tenable.
In the subtitle Spiegel claims that placebos ‘work in some conditions’.
Instead of this opinion it would have been more instructive if he had
clarified which conditions he refers to (besides pain), and presented
systematically reviewed evidence for clinically important effects of
placebo.
Peter C Gøtzsche, director
pcg@cochrane.dk
Asbjørn Hróbjartsson, senior researcher
Nordic Cochrane Centre,
H:S Rigshospitalet,
Blegdamsvej 9,
2100 København Ø,
Denmark
1. Spiegel D. Placebos in practice. BMJ 2004;329:927-8.
2. Hróbjartsson A, Gøtzsche PC. Is the placebo powerless? N Engl J
Med 2001;345:1278-9.
3. Gøtzsche PC. Why we need a broad perspective on meta-analysis: It
may be crucially important for patients. BMJ 2000;321:585-6.
4. Hróbjartsson A, Gøtzsche PC. Placebo interventions for all
clinical conditions. The Cochrane Database of Systematic Reviews 2004,
Issue 2. Art. No.: CD003974.pub2.
5. Hróbjartsson A, Gøtzsche PC. Is the placebo powerless? An analysis
of clinical trials comparing placebo with no treatment. N Engl J Med
2001;344:1594-602.
Competing interests:
None declared
Competing interests: No competing interests
We would like to extend some comments to the paper of David Spiegel
about placebos in practice (issue 23 October). Use of placebo can also
induce adverse effects (the named nocebo effect)(1). The prescription of
placebo by doctors is commonly conscious, but also could be unconscious.
The conscious use of placebo in clinical practice has one important
limitation, true placebos (inert substances) are usually not marketed and,
as a consequence, can not be prescribed easily. To our knowledge, placebo
is only present with its real name in some tablet packs of oral
contraceptives. In Hospitals, inert placebo capsules can be prepared by
pharmacist with non-active substances (e.g. lactose). As a consequence,
doctors usually should prescribe active placebos (substances with
pharmacological or therapeutic activity in other symptoms or diseases, but
not in the present indication) (2,3,4). In our opinion the unconscious
use of placebo by doctors is also frequent. In this case, doctors are not
aware really that they are using placebo. This phenomenon can appear in
the case of prescribing drugs when diagnosis is in doubt or is wrong, with
the incorrect use of right medicines (lower dose, shorter duration), with
the use of non-evidence based drugs or after the prescription of medicines
in non-approved indications. In all these cases, prescribing drugs which
have not proven beneficial effects may act only as a placebo but with the
side-effects of the active drug (harmful placebo)(5,6). The apparent
success of unconscious placebo treatment can mislead doctors into thinking
that the diagnosis was correct and the treatment effective, but this may
not be true. This phenomenon has been named “the therapeutic illusion”,
that can reinforce the conduct of incorrect prescribing. Patients with
therapeutic response to placebos can also present the therapeutic illusion
(7).
1. Ferreres J, Baños JE, Farre M. El efecto nocebo: la otra cara del
placebo [Nocebo effect: the other side of placebo]. Med Clin (Barc).
2004;122:511-6.
2. Grahame-Smith DG, Aronson JK. Placebos. In: Oxford Textbook of Clinical
Pharmacology and Drug Therapy. Grahame-Smith DG, Aronson JK, editors.
Oxford: Oxford University Press; 2002. p. 145-6.
3. Baños JE. Farre M. Efecto placebo. In: Principios de Farmacología
Clínica. Baños JE, Farre, editors. Barcelona: Masson 2002. p. 169-78.
4. Macedo A, Farre M, Baños JE. Placebo effect and placebos: what are we
talking about? Some conceptual and historical considerations. Eur J Clin
Pharmacol 2003;59:337-42.
5. Alvarez Y, Farre M. La ética del placebo. FMC. 2001;8:201.
6. Laporte JR, Capella D. Useless drugs are not placebos. Lancet.
1987;1:1324.
7. Markus AC. The ethics of placebo prescribing. Mt Sinai J Med.
2000;67:140-3.
Competing interests:
None declared
Competing interests: No competing interests
It is interesting to note that while it is unethical to prescribe
placebos in clinical practice, dummy therapies are often an important
ingredient of the ‘gold standard’ research method in evidence-based-
medicine. These are certainly ingenious ways to control for and separate
the effects of specific therapies from non-specific factors such as hope
and expectations. However, when a significant proportion of trial
participants respond to placebos, this finding is perceived as a nuisance.
Furthermore, when a significant proportion of participants respond to both
the ‘active’ therapy as well as the placebo, investigators often conclude
that the treatment was ineffective. This ‘double-positive-paradox’ (1)
actually informs us that something quite powerful may be taking place.
Nitzan and Lichtenberg’s survey (2) highlights some of the
limitations of conventional medicine in treating chronic and functional
conditions and points to some of the current paradoxes in medicine.
Placebos are described as ‘inert’ and ‘inactive’ interventions, yet
placebo effects are ‘real’ and ‘active’ (3, 4). As pointed out in
Spiegel’s editorial (5), hopes, feelings, and human relationships can
influence the course of a physical illness and medical care. In a
systematic review of placebo-controlled RCT’s we found some rigorous
evidence to support this (6).
Doctors, at least in Israel, are giving placebos to patients without
telling them, and investigators worldwide are often not telling their
patients what treatment they got once their study is over (7). Similarly,
an increasing proportion of patients are turning to alternative and
complementary medicine, often without informing their primary physicians
(8). What is this telling us about the level of trust and collaboration
between patients and their providers? And how is this impacting the
outcome of health care?
Little is known around the mechanisms and therapeutic effect of
patient-practitioner interactions in medicine. A larger body of
scientific knowledge can be found in the psychotherapy literature. In
1975, a review of 91 studies evaluating various psychotherapeutic
interventions found that these were not different in terms of
effectiveness, and concluded that ‘everybody has won and must have prizes’
(9). In a recent textbook by the American Psychological Association, the
therapeutic alliance was identified as the strongest predictor of outcome
in psychotherapy (10).
Rather than continuing to evaluate the effectiveness of CAM therapies
by adopting the biomedical approach and controlling for the potential
confounders deriving from health care interactions, we should be examining
the impact of these relationships and helping to restructure our health
services around our findings.
We seem to have forgotten the extent to which each individual
possesses natural self-healing capabilities and that these can be
harnessed in a safe place with a care giver that is present, empathic and
encouraging. Our health care system does not support this, impacting the
quality of care we are able to provide (11,12).
For some, placebos may be a more gentle solution for functional
symptoms or chronic conditions that conventional therapy is unable to
treat, having less toxic properties than pharmacological therapies.
However, secretly prescribing placebos can only increase the sense of
mistrust between doctors and patients as well as the idea that physicians
are not taking the patients’ problem seriously.
The issue is not about whether we should prescribe placebos, but
rather about the need to increase our general knowledge around healing and
mind-body mechanisms (13), ways to understand and harness these in an
ethical and practical manner, encouraging a sense of trust and partnership
between the public and health care specialists.
Over thirty years ago McGuire described ‘three stages in the life of
an artifact’: first it is ignored; then it is controlled for its presumed
contaminating effects; and finally it is studied as an important
phenomenon in its own right (14). It is time we stop considering
perceptions, feelings and health care interactions as variables that need
to be controlled in the pursuit of medical science, but include and study
these as potentially meaningful mediators and moderators of therapeutic
outcomes in clinical trials.
References
(1) Reilly D. Randomised controlled trials for homoeopathy. When is
useful improvement a waste of time? Double positive paradox of negative
trials. BMJ 2002;325:41;
(2) Nitzan U, Lichtenberg P. Questionnaire survey on use of placebo.
BMJ 2004:329: 944-6.
(3) Leuchter, A. F., Cook, I. A., Witte, E. A., Morgan, M., &
Abrams, M. Changes in brain function of depressed subjects during
treatment with placebo. American Journal of Psychiatry 2002; 159 (1), 122-
129.
(4) Moseley, J. B., O'Malley, K., Petersen, N. J., Menke, T. J.,
Brody, B., Kuykendall, D. H., Hollingsworth, J. C., Ashton, C. M., &
Wray, N. P. A controlled trial of arthroscopic surgery for osteoarthritis
of the knee. NEJM 2002; 2: 81-88.
(5) Spiegel, D. Placebos in practice. BMJ 2004;329:927-928
(6) Di Blasi, Z., Harkness, E., Ernst, E., Georgiou, A., &
Kleijnen, J. Influence of context effects on health outcomes: a systematic
review. The Lancet 2001; 357:757-62.
(7) Di Blasi Z, Kaptchuk TJ, Weinman J, & Kleijnen J. Informing
participants of allocation to placebo at trial closure: a postal survey.
BMJ 2002; 25:1329.
(8) Eisenberg DM, Davis RB, Ettner SL, Appel S, Wilkey S, Van Rompay
M, Kessler RC. Trends in alternative medicine use in the United States,
1990-1997: results of a follow-up national survey. JAMA 1998; 280:1569-75.
(9) Luborsky, L., Singer, B., & Luborsky, L. Comparative studies
of psychotherapies: Is it true that "everyone has won and must have
prices"? Archives of General Psychiatry, 1975. 32, 995-1005.
(10) Hubble, M. A., Duncan, B. L., & Miller, S. D. The Heart and
Soul of Change: what works in therapy. 1999 (pp. 1-462). Washington D.C.:
American Psychological Association.
(11) Mercer SW, Hasegawa H, Reilly D, Bikker AP. Length of
consultations. Time and stress are limiting holistic care in Scotland.
BMJ. 2002;325:1241.
(12) Freeman GK, Horder JP, Howie JG, Hungin AP, Hill AP, Shah NC,
Wilson A. Evolving general practice consultation in Britain: issues of
length and context. BMJ 2002; 324:880-2.
(13) Reilly D. Enhancing human healing. BMJ 2001;322:120-121.
(14) McGuire, W. J. (1969). Suspiciousness of experimenter's intent.
In R. Rosenthal & R. L. Rosnow (Eds.), Artifact in behavioral research
(pp. 13–57). New York: Academic Press.
Competing interests:
None declared
Competing interests: No competing interests
Excellent paper.
For a placebo to be useful it is necessary for the practitioner to foster,
or create belief in the efficacy of a treatment (s)he knows to be
ineffective. This necessarily involves deception of the the patient and
breaches the principle of patient autonomy.
The utilitarian argument against this is that the patient is seeking
relief from a psychological illness and cannot, or will not compehend the
psychogenic nature of the illness, and the most benign method of
intervention is to provide a placebo.
Is our role as medical practitioners one to use science where science is
capable of providing evidence of sufficent strength to support a decision,
and satisfy clients wants and wishes when it is not? What is clearly
inappropriate is for practitioners to foster interventions known to be
ineffective simply to get rid of an irritating client, or because they are
too lazy to seek evidence on what is effective.
Competing interests:
None declared
Competing interests: No competing interests
David Spiegel's comments on cin practice points to the need
for more research into the underlying biological mechanisms that mediate
the placebo effect. Such research cannot even begin, however, without
novel hypotheses to test.
I have recently proposed that one pathway by which placebos might
produce their effects is via the suppression of the acute phase response
(inflammation and sickness behaviour). See:
Dylan Evans, Suppression of the acute-phase response as a biological
mechanism for the placebo effect, Medical Hypotheses, In Press, Corrected
Proof, Available online 6 October 2004, .
(http://www.sciencedirect.com/science/article/B6WN2-4DGMR85-
6/2/64052c7a51cbf15cc8a6259b77c6ba35)
See also:
Dylan Evans. 2003 Placebo: The Belief Effect . London: Harper
Collins. Republished by Harper Collins as Placebo: Mind over Matter in
Modern Medicine in 2004.
---
Dylan Evans
Senior Lecturer in Intelligent Autonomous Systems
Faculty of Computing, Engineering and Mathematical Sciences,
University of the West of England,
Frenchay Campus,
Coldharbour Lane,
Bristol BS16 1QY
Web: www.dylan.org.uk
Competing interests:
None declared
Competing interests: No competing interests
A study published in the 1979 edition of Advances in Pain Research
and Therapy offered a tantalizing glimpse of a possible mechanism for
placebo analgesia.
A hundred or so patients who had wisdom teeth extraction were
assigned (random doule blind) to a fixed dose of an opiate (morphine, I
think) or the same volume of saline for post-operative analgesia. There
was no statistically significant difference in the proportion of patients
in the opiate versus the saline group who expressed satisfactory pain
relief. Placebo analgesia worked in a case of organic pain, post-operative
pain.
The researchers then broke the code after collecting analgesia data
and then randomised (again double blind) the saline responders to saline
or a dose of nalaxone. All the saline responders who got naloxone
complained of their pain again. This suggests that endogenous analgesic
systems involving enkephalins and/or endorphins might be involved.
So is placebo analgesia all in the mind? Or does the mind work via
known neuropharmacological pathways?
Those interested in looking the paper up will have to make do with
the following reference in non-standard manner:
Fields HL and Basbaum AI in Advances in Pain Research and Therapy, Vol.3
Ed Bonica JJ et al. Raven Press, New York 1979 pp427-440
Competing interests:
None declared
Competing interests: No competing interests
you rightly state that it is not because pain is relieved by placebo
that it is not
real. I would go one step further: as one of my teachers told me almost 30
years ago (Pr Raymond Villey, in Caen, France):
"beware of the pain that cedes to placebo: it's most certainly
organic".
I have seen that proven again and again. I have no explanation other
than the
one given for the soldiers at anzio: The patient with "real" pain wants it
to go
away so much that any straw will be clutched at to relieve the pain,
including
placebo.
On the other hand, the patient with "psychological" pain gains from
the pain
in some manner. There will be much less incentive to see the pain
relieved,
and placebo may be as ineffective as the other pain medication.
As for the dose-response to placebo, the adverse reactions to
placebos of
high-dose NSAIDs are much more frequent that those to placebos of low-
dose NSAIDs, in clinical trials. Go explain.
have a nice weekend
Nicholas Moore
Competing interests:
None declared
Competing interests: No competing interests
Placebo is effective and it differs from sham drugs or "inadrugs"
Dear Sirs,
As it is written, placebo comes from Latin "I will please". Nocebo is the
opposite to placebo. It is a pity that many physicians and especially the
pharmacological field have adopted the term placebo to a sham drug used in
comparative randomized clinical trials (RCT).
A RCT is never a perfect patient-physician relationship because the
patient cannot be sure that her/his doctor is giving the best treatment.
Also the Doctor's healing (placebo) effect differs from a situation, where
the Doctor (or a therapist) is giving treatment that he/she trusts in.
All medical personnel should always try to add their own therapeutic
healing potential (the positive placebo effect) to the given
pharmacological or surgical teatment. The effect of a postive placebo is
often more than 25 or even more than 30 % of the total therapeutic effect
of a drug. Why should we weaken the therapeutic by so much in neglecting
the use of placebo in our routine daily work. In the same ways we should
always avoid use of nocebo.
A good solution would be to use PLACEBO in its original meaning -
pleasing the patient - trying to do the best to treat the patient. Some
members in the alternative medicine field talk about energy healing.
Placebo can be understod also as "energy healing" - whatever it means.
If the word placebo is used in its proper meaning a better word
should be used in randomized clinical trials. We have such words, e.g a
SHAM DRUG, or a SHAM TREATMENT. An alternaive is PHARMACOLOGICALLY
INACTIVE DRUG. A new word could be invented. Could it be "INADRUG". I am
sure that much better words could be found. Whatever a word is used, it
should tell to us that we do not believe that this drug has any
pharmacological activity. Placebo drug does really not equal calcium
tablets that are used in clinical trials.
Competing interests:
None declared
Competing interests: No competing interests