[Editor’s note: With Scott away on vacation, this Thursday we bring you the inaugural post of Tomasz Witkowski, a psychologist, academic, founder of the Polish Skeptics Club, and author of two books favorably reviewed by Harriet. Welcome Tomasz!]

First, do no harm – a rule that enjoys practically unconditional acceptance by all of human civilization, at least when it concerns healthcare interventions. Students of medicine have this hammered into their heads beginning with their first lesson. It is so universal that we treat it as a part of the Hippocratic oath, while it in fact never was one; Hippocrates’ authorship itself is also frequently called into question.

If we consider the matter more carefully, it turns out to be not only inhumanly difficult to adhere to, but even cruel in a certain sense. A doctor, a physiotherapist, or a nurse is faced with a person who is suffering, in need of help, sometimes immediately; but before that help is given, it is necessary to take into consideration all of the surrounding circumstances and professional knowledge in order to be entirely sure that no harm will be done. And if they hesitate? If they aren’t completely sure? Should they refrain from intervening? The rule is clear – yes. Not doing harm takes priority over doing something which may be helpful, but which also bears too great a risk of harm. Thus, sometimes the principle of primum non nocere means we should refrain from providing assistance to a person who is suffering. It means that not-acting is better than rash, uncertain action.

And yet, when taking a closer look at medical practices, it turns out that this overarching principle is giving way to one that stands in opposition – melius anceps remedium quam nullum – an uncertain treatment is better than no treatment at all. Those who give care prefer to do anything at all rather than nothing. And what if it brings relief in suffering? The motives of the people inclined to follow the latter rule are diverse. They may attempt to do something out of pure sympathy, or due to pressure put on them by the patient. They may be pushed into doing something by their environment, the demands of the sick person’s family, or even their superiors. Sometimes they are driven by aversion to the helplessness which they feel when refraining from action, or by the need to verify their own competences and a feeling of their own professional value. Not infrequently, they are also motivated by economic considerations – after all, patients prefer to return to doctors or therapists who have done something, prescribed some medicine or performed some intervention, rather than to those who simply shrug their shoulders and say there’s no tried-and-true method, so the rule of primum non nocere means that nothing is to be done.

An unequivocally negative and categoric judgement of submission to the principle of melius anceps remedium quam nullum would be nothing but a manifestation of moral absolutism, which is not something to which I am prone. A fair assessment of such conduct cannot be offered without taking into account motive on the one hand, and consequences on the other. This is expressed in the risk-benefit ratio, a measure broadly accepted in medical studies. Rather than refraining from acting in the face of even the smallest risk, we calculate the relation of the risk of harm to the potential benefits that the patient may receive, silently accepting the tenets of utilitarianism.

Psychotherapy and primum non nocere

And yet, when I observe the practice of medicine, particularly psychotherapy, I am unable to avoid the impression that many of those offering their services in medicine and psychotherapy have simply forgotten about the existence of the principle of primum non nocere, and their decisions have nothing in common with a rational calculation of the threats and benefits. But is the application of ineffective therapeutic methods consistent with that principle? The cautious answer is yes – as long as they bring positive psychological effects comparable to those that come from the use of a placebo. Yet this cautious answer most certainly requires elaboration. Yes, but only when we are entirely certain that effectiveness was measured in comparison to a placebo group, and not to a control group that was not subjected to any treatments. Furthermore, if this condition is fulfilled as well, we should also obtain the unshakeable conviction that there are no methods in existence that would be more beneficial in the very same situation the patient is in. Indeed, the suffering of a patient treated improperly may deepen, and that patient, firm in the belief that the treatment has been successful, will not seek more beneficial methods; this is fuel for needless suffering.

The issue of ineffective treatment should also be viewed from the perspective of the difficulties the patient goes through. This frequently involves abandoning other activities (family, social, professional, etc.) to devote time to treatment. Patients devote a portion of their financial resources to treatment. That alone is already a difficulty. It becomes an even greater burden when such therapy is offered to terminally ill patients – rather than live out their final days in the company of loved ones, doing the things that give their lives sense, they participate in empty rituals of ineffective therapies. I would thus submit that ineffective treatment is equivalent to the infliction of harm on a patient, and I will attempt to demonstrate how common this phenomenon is in psychotherapy.

In 2017, a team directed by Evangelos Evangelou from the University of Ioannina in Greece did something out of the ordinary – a meta-analysis of meta-analyses concerning the effectiveness of psychotherapy. Let us recall that meta-analysis is a method that gathers in one place the results of studies from many experimental works. A meta-analysis can only include those experiments that include a control group and in which random assignment of participants to groups was employed, and thus studies meeting the highest methodological standards. What Evangelos Evangelou and his team did was to gather 247 finished meta-analyses and combined them into one single meta-analysis. The result was striking. While the majority of meta-analyses favoured psychotherapy as an intervention method, a mere 7% of them contained convincing evidence that psychotherapy was effective! Does the remaining 93% guarantee that we will remain faithful to the rule of primum non nocere?

At present it is estimate that there are over 600 different modalities (approaches, schools) of psychotherapy practiced around the world. Can the average psychotherapist, deciding on the choice of a specific therapeutic intervention, be sure that it is the very best choice for the patient? Well, I am afraid that none of them (or at least nobody I have met) boast the brainpower to study, remember, and ultimately engage in measured reflection on the benefits and threats of them all. And what about those who are creating new modalities? Do they begin with an analysis of all existing ones, arriving at the conclusion that one more is necessary for improving the effectiveness of psychotherapy? If this were the case, it would mean that every new modality is better than the previous one, and the vintage ones would be slated for extinction. But this is not happening – the majority of them are doing quite well, and rather than decline, their numbers are growing. In such conditions is it really possible to implement the principle of primum non nocere?

The consequences of the existence of so many ineffective interventions are borne by patients. Therapists continue to benefit from them. Patients subjected to ineffective interventions become the object of a sort of “recycling”. Disappointed by the absence of effects, they reach successive therapists who take a portion of their resources, and if those turn out to be ineffective, there is a chance that still other representatives will get the chance to demonstrate their effectiveness. “Record holders” that I have encountered in my life have been in psychotherapy for over 25 years, and I’ve heard of a single patient who was treated in succession by as many as 18 psychotherapists. Primum no nocere or perpetuum mobile?

The issue of faithfulness to the principle that has become the title of this essay stands out even more when we shift our perspective from the quite common absence of effectiveness towards the negative effects of treatment undertaken. In medicine this is the well-understood – but still shameful – phenomenon of iatrogenesis. In the United States alone, mistakes in treatment cause between 230,000 and 248,000 deaths each year. Of those, around 12,000 are caused by unnecessary surgical interventions, 27,000 by mistakes made in hospitals, 80,000 by hospital infections, and around 107,000 people die as a result of side effects from medicines applied properly. And how does the situation look in psychotherapy?

To tell the truth, it looks a lot better, because psychotherapists and researchers have done their best to ensure that as few such numbers as possible have seen the light of day. Literally. In 2014, an article was published whose authors had reviewed all studies on the effectiveness of psychotherapy published in 2010. From among the 132 works they gathered, a mere 28 of them (21%) contained any information that would suggest their authors had controlled the negative effects of the examined therapeutic interventions. Only four of them (3%) contained descriptions of those effects and methodologies for collecting data. Five of them did contain a description of negative effects, but there was no information about the manner in which they were measured. In four additional reports (3%) it was only briefly stated that no undesirable effects had occurred, while 15 studies (11%) supplied merely information about declines in patients’ condition or made mention of the fact that such a decline was being monitored. The remaining 104 papers (79%) contained no information about the monitoring of negative effects of psychotherapy.

Unfortunately, these results are not unique. Conor Duggan and collaborators analysed 82 psychotherapeutic procedures confirmed by the National Institute for Health Research in the United Kingdom in the period 1995-2013. None of them mentioned potential negative effects of psychotherapy, although the majority of pharmacological therapy protocols did.

Knowledge on the negative effects of psychotherapy is also absent among American psychologists. In 2006, Charles Boisvert from the Rhode Island Centre for Cognitive Behavioral Therapy and David Faust of Brown University Medical School conducted a survey among 181 psychologists and APA members operating their private practices across the United States. As many as 28% of that group was unaware of the existence of negative effects of psychotherapy. In turn, other studies have shown that therapists frequently assess the effects of conducted psychotherapy inaccurately – they regard it as effective even when it does not generate any positive results. As the article’s authors warn us, this involves the risk of continuation of ineffective therapy, and in consequence a worsening of the patient’s condition. This is confirmed in studies by Michael Lambert and his collaborators, who asked 40 clinicians to make predictions on the potential for negative effects to occur at the conclusion of psychotherapy among 550 of their patients. It turned out that even despite information that the general proportion of patients experiencing negative effects was around 8%, they predicted that virtually no negative effects could occur among their patients. They only indicated three cases in which the patient’s condition could decline. The prediction was accurate in one case. The clinicians failed to identify 39 other patients whose condition deteriorated in the course of therapy.

While the issue of iatrogenesis in medicine is studied systematically – a mere browsing of the literature shows a range of books and monographs on the subject – I have found a mere handful of articles on the subject of iatrogenesis in psychotherapy, the majority of which were theoretical. I have not found a single book dedicated in its entirety to the subject, at least not in English.

When I look at this data, I am convinced that the principle of primum non nocere, allowing us to identify threats to patients in psychotherapy, has been replaced by the suggestion made frequently by contemporary youngsters – “stay out of it!”

The most well-known public declaration of adherence to the principle of melius anceps remedium quam nullum was a declaration in 1891 by the Swiss psychiatrist Gottlieb Burckhardt, who said:

Doctors are different by nature. One kind adheres to the old principle: first, do no harm (primum non nocere); the other one says: it is better to do something than do nothing (melius anceps remedium quam nullum). I certainly belong to the second category.

It evoked outrage among his contemporaries, and Burckhard’s career ended in ridicule and indignity. But his attitude found many admirers, particularly among neurosurgeons at the end of the 19th and in the first half of the 20th century. While they did not flaunt it, they did follow the advice of Burckhard, infamous proof of which can be found in the thousands of victims of lobotomies and other, often haphazard, surgical interventions performed on psychiatric patients.

Equally famous to Burckhardt’s declaration is the riposte of the French psychiatrist Armand Semelaigne, who declared that “an absence of treatment was better than a bad treatment,” thus demonstrating the fundamental meaning of the principle primum non nocere. Where is psychotherapy to be found among these ethical precepts? Is it following the simple path marked out by Armand Semelaigne? Or rather seeking possibilities for action in the footsteps of Burckhard?


Posted by Tomasz Witkowski

Tomasz Witkowski, PhD, is a Polish psychologist, skeptic and science writer. He is known for his unconventional campaigns against pseudoscience. He specializes in debunking pseudoscience, particularly in the fields of psychology, psychotherapy, and diagnostics. As a founder of the Polish Skeptics Club, he also engages in debates on pseudoscience-related topics, emphasizing scientific skepticism. He has authored over a dozen books, over 40 scientific articles, and about 200 popular science articles. Two of his books have been published in English: Psychology Led Astray: Cargo Cult in Science and Therapy (2016) and Psychology Gone Wrong: The Dark Sides of Science and Therapy (co-written with Maciej Zatonski, 2015). He writes regularly for BPS Research Digest.