Placebos: Awareness Does Not Protect Against Effects

28. November 2017

Doctors are reluctant to use placebo instead of medication as a therapeutic measure because they do not want to betray their patients. Perhaps they do not even have to. One study shows that placebos work, even when patients know that the preparation contains no active agents.

A sugar pill for migraine, or a salt injection for back pain: many physicians reject these notions. Eight medical doctors from various disciplines randomly interviewed by DocCheck stated that they generally did not use or had not used any dummy medications in practice. For them it is a matter of ethical as well as legal relevance.

Not jeopardising the relationship of trust

Dr. Herbert Vogl, a now retired general practitioner located in the vicinity of Munich (Germany), explained: “Patients would often come to the surgery with the expectation that I would treat them or at least prescribe them a drug”. Yet for some clinical situations that would be neither sensible nor purposeful. “What would then have been easier than giving them medication-free supplements?” Nevertheless, he rejects placebos: “As luck would have it, patients find out about that, whereby mistrust then remains forever. Especially in rural areas, that kind of thing gets about”. He adds: “The legal situation was also previously unclear, the German Medical Association had not yet made recommendations, as is currently the case. No one wants to risk going to trial on account of supposedly creative errors”.

Taking another look at hospitals: “Patients’ expectations have increased, especially with regard to emergency treatment”, Professor Dr. med. Roland Bingisser from University Hospital Basel reports. “Even if patients come to us with trivialities, they expect treatment – a patch or a few encouraging words, which can also be good placebo, are often not enough”. As an example, he mentions nausea without evidence of serious underlying diseases: “Instead of explaining this and telling the patient that he or she should not have to come to an emergency because of a trivial matter, the patient immediately feels better taking a medication”. Yet in strictly regulated clinical practice conducted according to guidelines, dummy medications do not come into use.

Not part of the therapy


Jeremy Howick © Oxford University

An older study from the UK provides more representative data. Jeremy Howick from Oxford University wrote to 1,715 GPs and general practitioners. Of these, 783 (46 percent) completed the questionnaire. They were to indicate whether placebos are generally used. If so, Howick wanted more information on the nature of the dummy medication and the reasons for its use.

‘Virtual’ placebos, in other words dietary supplements with no proven effect on a particular disease, were used comparatively frequently. 77 percent reported using such products at least once a week. The predominant reasons given were the psychological effects (50 percent) involved. In many cases, patients sought therapy, although therapy was not necessary, the lead author summarises. The most common example is having a flu-like infection. In second place was the explicit desire of patients to receive medication from the doctor (45 percent).

Only 12 percent of the physicians surveyed had ever used classic placebos. These are defined as medication-free dummy medications, such as lactose or saline. Asked why they did not use classic placebos, respondents said they did not want to betray patients. Their focus was on the highly valued relationship of trust. The doctors in many cases feared that sham medications, despite being potentially purposeful, could lead to a rift with long-time patients. However for the doctors it is not just about ethical values, but also about the economic situation of medical practices. Patients share information via web portals, and a bad rating can have devastating consequences.

Now be honest, Doc


James E.G. Charlesworth © University of Oxford

This paradigm is now being questioned by several working groups. They were able to demonstrate effects when using so-called open-label placebos. The term stands for preparations in which physicians openly inform their patients about missing ingredients.

James EG Charlesworth of the University of Oxford looked in literature databases for clinical, randomised studies on the subject. Five studies involving a total of 260 participants met the inclusion criteria. They dealt with drug-free preparations used against irritable bowel syndrome, depressions, allergic rhinitis, back pain and attention deficit/hyperactivity disorder (ADHD).

“Open-label placebos seem to have positive clinical effects compared to no treatment”, Charlesworth summarises. He cannot exclude methodical distortions, something partly related to the findings used for the aforementioned clinical conditions. Guidelines, for example, accurately define ADHD or depression. However, not all researchers in the examined studies adhered to the accurate definitions. Nevertheless, Charlesworth’s evidence is sufficient to investigate the use of placebos in more detail. He suggests focusing on a defined clinical finding or symptom.

Using effects in an ethically acceptable way

One study published a few months ago goes even further. Researchers focused on pain perception,using open-label placebos given as a cream.


© Uni Basel

Cosima Locher from the Department of Clinical Psychology and Psychotherapy at the University of Basel recruited 160 healthy volunteers and divided them into four groups. They received …

  1. no therapy (control group)
  2. a white cream with no active ingredient; explanation absent
  3. a white cream with no active ingredient, but with the indication that it was a placebo,
  4. a white cream with no active ingredient, but with the note added that this was the local anesthetic lidocaine.

Thereafter, Locher determined for all groups the perception of heat pain on a scale of 0 to 100 points. Subjects reported on average a pain index of 60.4 (group 1), 62.1 (group 2), 59.5 (group 3) and 57.3 (group 4). Contrary to previous assumptions, not only was there a clear effect evident in group 2 (with no indication of a placebo) and in group 4 (with a false promise of action). Patients in group 3 who were openly and honestly informed about the absence of active agents experienced less pain as well. There was a 15-minute, generally understandable explanation given before the start of the test. Group 1 served as a comparison and involved no intervention.

“The previous assumption that placebos only work if they are administered by deception should be reassessed”, Locher concludes. And the co-author Jens Gaab, also a researcher at the University of Basel, adds: “The open administration of a dummy medication offers new ways for using the placebo effect in an ethical manner”.

Psychotropic medications – not always necessary

Locher sees antidepressant use with adolescents as one potential example. Together with colleagues, she analysed 36 drug trials. The analysis included 6,778 children and adolescents up to the age of 18. They had received serotonin reuptake inhibitor (SSRI), serotonin norepinephrine reuptake inhibitors (SNRI) or placebo.

Unsurprisingly, verum worked better than placebo. That’s what the licensing authorities had already assumed anyway. The difference was, however, small. At the same time, the scientist found differences depending on the clinical condition. Placebos used for depression showed more pronounced effects than for anxiety disorders. Locher and colleagues hope that factors that contribute to the placebo effect could be more precisely exploited in treatment of depression.

Ethical and moral difficulties

So what then is stopping doctors from “treating” by using dummy medications? First and foremost ambiguities relating to practical application. The German Medical Association (GMA) released a comprehensive – perhaps even too extensive – study years ago. In it, experts point to the fundamental obligations of offering effective therapies, alleviating ailments and informing patients.

The GMA considers the use of placebos beyond clinical trials to be justifiable where …

  • no effective pharmacotherapy exists,
  • they are not used to treat a serious disease,
  • the patient expresses this through his or her own wishes,
  • there is a chance of success.

An effect is hardly able to be detected, according to Professor Dr. med. Robert Jütte of the Robert Bosch Foundation, when placebos are used for chronic obstructive pulmonary diseases (COPD), osteoporosis and pulmonary hypertension. In contrast, placebos used for arterial hypertension, Parkinson’s disease, partial epilepsy and rheumatoid arthritis perform considerably well in comparison to active agent-containing medications and “could therefore be fully included in decisions relating to medication therapy”.

2 rating(s) (5 ø)


Mr William Bradley
Mr William Bradley

Many conditions will always improve without treatment. Many years ago we had a very cynical saying – the patient got better despite our treatment. One has only to look back at the days when there was no really effective treatment. Recovery was dependent on trust in the consulting person and as Florence Nightingale over 150 years ago good nursing can have a very benficial effect. Having been involved in drug trials we were very much aware of the placebo effect, which this and consultant trust problably helps to explain the effectiveness of homeopathic medicines.

#3 |

Am I missing something in the Basel placebo study data? It appears there were 40 subjects per group. No indication is given of standard deviation within the groups, but given that pain tends to be very personal, I would imagine that individual scores must have been at least plus or minus 10 from the average rating. The average difference between the control and test groups was no more than 3. The average difference between control and open placebo was only 1 – on a scale of 0 to 100. I don’t need to run the statistics to know that there is no statistically significant difference between groups. Even if this study had been done with enough subjects to reach statistical significance (I’m guessing thousands would be needed), can one really argue that lowering an average pain score from 60.4 to 59.5 is clinically meaningful?
The use of an open placebo is an interesting idea, but I see no evidence in this article to support any utility.

#2 |

In Germany placebo administration is dangerous not for the patient but for us MP because there is a high risk to be condemned by justice as the ” golden calf ” of pharmaceutical industry is called ” evidence based medicine ” what means when you do not work like it is prescribed (avoiding natural treatment) you are in danger .

#1 |

Copyright © 2021 DocCheck Medical Services GmbH
Follow DocCheck: