[Editor’s note: Once again, limitations in WordPress force us to attribute this article solely to returning author Tomasz Witkowski, but once again, it was written and submitted by both Dr. Witkowski and his colleague Maciej Zatonski, co-authors whose books have been favorably reviewed by Harriet.
Note that this post touches and expands upon a publication readers might be familiar with, as it was discussed by Harriet in January.]
When Mr A. finally reached the Emergency Department of his local hospital. “Please, please help… I swallowed all of my pills” he mumbled to the attending nurse. An empty, unlabelled medicine box dropped from his hands only seconds before he collapsed on the hospital floor. The staff immediately started the required emergency procedures. The twenty-six-year-old male appeared very pale and apathetic but was still in limited verbal contact. He managed to explain to the emergency room staff that as a clinical trial participant, he took 29 remaining tablets of an unknown experimental therapeutic investigated as a potential treatment for depression. His impulsive decision to commit suicide was soon followed by a reflection that he is not ready to die just yet. As he realised that his life might be about to end, he asked his neighbour to take him to the nearest hospital. Initial examinations revealed severe hypotension. His blood pressure was 80/40 mmHg and dropping, and his heart rate was elevated to 110 beats per minute. He was shivering and breathing rapidly. To prevent the blood pressure from falling even more, he was immediately given an intravenous infusion of 0.9% solution of sodium chloride. Over the next 4 hours, he received over 6 litres of intravenous fluids, which only partially helped to stabilise his blood pressure. The patient was still lethargic and apathetic. During the attempts to stabilise his condition, the staff has tried to learn more about the experimental treatment he was taking. It quickly turned out that he was indeed a participant of a registered clinical trial investigating new treatments for depression, however…he was in the placebo control group! In less than 15 minutes from informing Mr A about being in the placebo arm, his heart rate and blood pressure returned to normal, and he has recovered from his lethargic state. Luckily there was no permanent kidney damage caused by the administration of life-saving fluids.
Statins, sex and aspirin
Mr A’s history is one of the most famous documented examples of the nocebo effect – the evil twin of the well-known placebo effect. Placebo is derived from Latin and can be translated as I shall be pleasing. The first use of this word in the medical context most likely dates back to a book titled Mother’s New Dictionary from 1785. Placebo was used to describe a sham therapy or a sham medicine that was supposed to mimic an actual treatment. Today a placebo is a substance or treatment which is designed to have no therapeutic value. On the other hand, the placebo effect (or indeed, placebo effects) can be used to describe therapeutic outcomes that are derived from administering an inert treatment reported by people taking such placebo. The same inactive substance (placebo) can also cause quite severe negative results, as we have seen in Mr A’s case. Such a negative outcome caused by administration of an inert treatment is called a nocebo effect. The term “nocebo” means “I shall harm“. During the process of discovering and testing new medicines, the placebo and the nocebo effects occur very frequently. The nocebo is estimated to be responsible for 19-71% of all untoward adverse reactions to a medicinal product. Quite often, both of those effects cancel each other out in the data collected from the clinical trial participants. Among the subjects who received the placebo, the numbers of reported benefits are comparable to the numbers of reported adverse events due to exposure to an inert substance.
There are some terminological challenges related to reviewing both of those effects. Many authors choose to describe the negative impact of a placebo, rather than the nocebo effect. On the one hand, this could be related to a rather nonchalant use of medical terminology. On the other hand, it could be caused by the difficulty in distinguishing if the impact of a neutral substance is beneficial or harmful. For example: is a reduction in blood pressure a positive or negative consequence of administering a placebo?
For some reason, only the placebo effect has made it into the minds of the public, while the nocebo effect remains mostly unknown. Furthermore, the nocebo effect shows up for the first time in a medical context as late as 1961, due to the works of Walter Kennedy. Meanwhile, the under-appreciation of the nocebo effect allows it to continue to cause quite a bit of harm.
A particular illustration of this problem was recently highlighted in the recently published research (discussed in January by Dr. Hall) conducted at the Imperial College London. The researchers showed that the nocebo effect is largely responsible for patients’ discontinuations of their treatment with statins – a group of medicines usually prescribed to prevent heart attacks and strokes. The research was conducted at the Hammersmith Hospital in London among hospitalised patients who previously discontinued taking statins due to the perceived side effects of the treatment. Each patient received a 12-month supply of medicines in separate containers: four boxes contained statins, four contained placebos, and four were empty so the patient took no pills for those months. Patients did not know which of the containers with tablets contained statins and which contained a placebo. All participants were asked to carefully describe how they felt on each day of the year. Regardless of whether patients were taking statins or placebo, they have reported a similar number of significant adverse events. The symptoms disappeared only in the months when the trial subjects did not take any pills. Just as Mr A., they became the victims of the nocebo effect.
Statins are not the only victims of the nocebo effect. In 2002 Arthur Barsky with his colleagues demonstrated that 97% of reported allergic reactions to penicillin (often a life-saving antibiotic) can actually be attributed to the nocebo effect. Their analysis has shown that adverse effects experienced by the vast majority of patients are caused by the widespread public awareness of the possible risk of allergic reactions to penicillin. It should be noted that penicillin has one of the largest therapeutic windows among all known medicines.
The adverse reactions caused by acetylsalicylic acid (aspirin) show a similar pattern. Aspirin was compared to another anticoagulant in three research centres. In two of the centres, but not in the third one, the patients were informed about the possibility of experiencing gastrointestinal disorders. The analysis of reported adverse events showed that patients warned of potential risks reported them three times more often than patients who were unaware of the danger. Objective assessment of actual damage to the digestive tract, including stomach ulcers, showed absolutely no differences between the groups.
The nocebo effect can potentially ruin the sexual life of men and their partners. Urologists from the University of Florence explored the adverse events of finasteride, often used in the treatment of enlarged prostate. Their research showed that the nocebo was responsible for a significant percentage of reported sexual dysfunctions among men treated with finasteride. The effect occurred primarily among those informed about this potential side effect, regardless if taking finasteride or placebo.
The power of information
In 2017 Science published an article reporting on the application of different topical medicines in the treatment of inflammatory skin conditions. One cream was presented in expensive-looking packaging, while the other was placed in a box deliberately designed to look cheap. None of the creams contained any therapeutic substances. Subjects were informed that their medicines can increase their pain sensitivity. It turned out that skin areas covered with placebo creams were, in fact, more sensitive to pain when compared to the untreated areas. The effect was also more pronounced with the more expensive-looking placebo, likely to being perceived as “stronger”. What is even more interesting, the reported sensations were not only subjective: MRI imaging of the subjects’ brains showed changes characteristic of other painful conditions.
The nocebo effect, just as the placebo effect, is caused by the expectations generated by acquired information. During clinical trials, patients receive information about expected positive outcomes of medicine, as well as possible side effects. In some subjects, the combination of the provided information and the administration of inert placebo can lead to positive or negative outcomes. Both the placebo and the nocebo effects are virtually impossible for physicians to predict. Doctors would require an in-depth knowledge of patients’ personalities and a record of their experiences with various treatments. The nocebo effect can be triggered by merely reading about potential side effects from the patient information leaflets included with the medicine. Some doctors even forbid their patients from reading the package inserts. Relatively frequent nocebo effects reported with statins is likely caused by their unfounded negative reputation amplified by the media. The challenge is that statins were unquestionably proven to prolong the lives of the patients who take them, yet the implied side effects often lead to poor compliance and low adherence to prescribed treatments. Patients pay the ultimate price: a significantly increased risk of heart attacks and strokes.
Suppose the nocebo effect is as frequent as the placebo effect. Why does the former remain largely unknown, while virtually everyone appears to know about the latter? The answer could be simple: the placebo effect helps to sell worthless therapies and products. Our familiarity with the placebo effect is caused mainly by marketing activities. Defenders of alternative medicine often repeat some variation of “It might be worth to try, even if this does not work! Perhaps the placebo effect will work?” They neglect to mention the opposite effect is just as likely. “Why oppose homoeopathy? Even if there is nothing in it, it will not harm anyone, and the patient might benefit from the placebo effect?” Well, shouldn’t an honest proponent of sham treatments also discuss the likelihood of the nocebo effect?
The nocebo effect is not often mentioned at all. Why would it be? After all, it does not help to sell any sham therapeutics. Information-induced adverse reactions can pose a real threat to our health or even life, especially when we consider the impact on discontinuing treatments aimed at management of genuine medical conditions. Unnecessary suffering caused by the nocebo effect, such as the decreased tolerance to pain, is an equally significant price that patients pay. The marketing use of the potential benefits of the placebo effect without informing of the impact of nocebo is at least dishonest. It is no different to selling tobacco, alcohol or other medicines without the information about the harms they may cause. There are no control measures related to the provision of information about any potential harmful effects of therapies. Media outlets are free to discuss subjective opinion statements about medicinal substances. Some treatments get an undeserved bad press that is not reflected in the available evidence. Other medicines enjoy wide public acceptance, even when their therapeutic value is hardly justifiable.
Another issue is related to the product information leaflets provided with placebos, such as homoeopathic “remedies”. Homeopathic products usually contain no active ingredients, or their concentrations are so low that it is often impossible to find a single molecule in the entire pack. Despite this, homoeopathic products are often accompanied by packaging inserts designed to create an impression that we are dealing with a real medicinal product. The manufacturers of homoeopathic products often warn about adverse events such as hypotension, breathing difficulties, dyspnoea, itchy rash, nausea, vomiting, various kinds of pain, gastrointestinal disturbances, agitation, hallucinations, even…paralysis! In 2011 sceptics from around the world “overdosed” massive amounts of homoeopathic products in a large international campaign called “Homeopathy – there is nothing in it”. From tens of thousands of volunteers, not a single one reported any of the adverse reactions that the makers of those products warned against. However, all of the volunteers were critical thinkers. They also had an in-depth knowledge of the manufacturing processes and of the actual contents of homoeopathic products. How can we guarantee that there are no patients like Mr A. among people who take such products?
Placebo instead of morphine
The conscious and deliberate prescription of placebos is often discussed by therapists and physicians. The vast majority agrees that the use of placebo is unethical. Such views are not by caused fears of the nocebo effect, nor by the conviction that placebo treatments exactly resemble selling of red-coloured water as wine. If placebos do not yield measurable results, patients can lose trust in the treatments and in the physicians who prescribe them. In some cases it can lead to learned helplessness – if nothing has helped so far, they may stop any further attempts to address their condition.
Liberal physicians do allow the use of placebo in certain situations only. For example, a hypochondriac may be harming him or herself with real medicines and refuses to stop to take them, arguing that they are helping him. Swapping potentially harmful medications with placebo seems to be justified in such cases. Similarly, during the treatment of drug addictions, the active substance may be slowly replaced by inert substance, which could have a measurably positive impact on the patients. Such an approach, however, is often discussed with the patient during the consent process. Placebo treatments are sometimes used by physicians in cases where there are no effective treatments available, and doctor’s knowledge of the patient allows them to assume that placebo might help. Such an approach is mostly used (often consensually) in the management of pain in terminally ill patients. Placebo has been shown to reduce pain just as effectively as morphine in some instances. An effective application of placebo requires significant clinical knowledge and experience. This is why placebo treatments should only be used in exceptional circumstances by trained medical staff – just like any other pharmacotherapy. Only specialist physicians should be assumed to have sufficient knowledge and judgment to assess the risk and benefit ratio of the intervention.
Another important consideration related to placebo and nocebo should be taken into account. Although both of the effects can seem impressive, they are related exclusively to subjective symptoms and conditions. They include pain, insomnia, impotence, nausea and vomiting, anxiety, mood disorders, itching, gastrointestinal problems, etc. Symptoms are not diseases, and placebos have never been shown to act against the natural course of an organic condition. Administration of placebo has never resulted in the regrowth of an amputated limb, disappearance of a tumour, remission of metastases, functional improvement in a permanently damaged kidney, or any other permanent damage in any organ, process or bodily function. Similarly, despite widespread legends of voodoo deaths, it was never proven that inert substances or practices have ever caused it.