According to a wide-spread belief, the demand to ‘first do no harm’ originates from the Hippocratic oath, which all doctors take when finishing medical school. Few of us appreciate that both of these assumptions are incorrect. Firstly, doctors do not normally take the Hippocratic oath – you only need to read it to understand why; it contains many points that would make little sense today. And secondly, the Hippocratic oath does not actually include the phrase ‘first do no harm’.
Nevertheless, you might assume, doctors are obliged to do no harm. After all, isn’t this is an important principle of medical ethics? In fact, this assumption is also not entirely correct.
Strictly speaking, doctors need to do harm all the time. Their injections hurt, their diagnostic procedures can be painful, their medications can cause adverse effects, their surgical interventions are full of risks, and so on. None of this would be remotely acceptable if doctors had to adhere to the principle of first doing no harm.
The ethical imperative of doing no harm has therefore long been changed to the demand of doing more good than harm. Of course, doctors must be allowed to do harm, even quite serious harm, as long as their actions can be expected to generate more good overall.
In medical terms, we speak of the risk/benefit balance of an intervention. If the known risks of a treatment are greater than the expected benefits, we cannot ethically administer or prescribe it; however, if the benefits demonstrably outweigh the risks, we can consider it a reasonable option.
But what about the many treatments where there is uncertainty regarding either the risks or benefits, or both? In such cases of incomplete evidence, we need to look at the best data currently available and, together with the patient, try to make an informed judgement.
Perhaps this is best explained by running through a few exemplary scenarios in which homeopathy (the classic example of a therapy that is promoted as being entirely harmless) is being employed.
Scenario 1: Patient with a self-limiting condition
Let’s assume our patient has a common cold and consults her physician, who prescribes a homeopathic remedy. One could argue that no harm is done in such a situation. The treatment cannot be expected to help beyond a placebo effect, but the cold will disappear in just a few days, and the patient will not suffer any side effects of the prescription. This attitude might be common, but it disregards the following potential for harm:
- The cost of the treatment
- The possibility that our patient suffers needlessly for several days with cold symptoms that might easily be treatable with a non-homeopathic therapy
- The possibility of our patient getting the erroneous impression that homeopathy is an effective therapy (because, after all, the cold did eventually go away), and therefore opts to use it for future, more serious illnesses.
What if the physician only prescribed homeopathy because the patient asked him to do so? Strictly speaking, the above issues of harm also apply in this situation. The ethical response of the doctor would have been to inform the patient what the best evidence tells us (namely that homeopathy is a placebo therapy), provide assurance about the nature of the condition, and prescribe effective symptomatic treatments as needed.
And what if the physician does all of these things and, in addition, prescribes homeopathy because the patient wants it? In this case, the possibility of harms one and three still apply.
Scenario 2: Patient with a chronic condition
Consider a patient suffering from chronic painful arthritis who consults her physician, who prescribes homeopathic remedies as the sole therapy. In such a situation, the following harms need to be considered:
- The cost for the treatment
- The possibility that our patient suffers needlessly from symptoms that are treatable. As these symptoms can be serious, this would often amount to medical negligence.
What if the physician only prescribed homeopathy because the patient asked him to do so, and refused any conventional therapies? In such cases, it is the physician’s ethical duty to inform the patient about the best evidence as it pertains to homeopathy as well as effective conventional treatments for their condition. Failure to do so would amount to negligence. The patient is then free to decide, of course. But so is the physician; nobody can force them to prescribe ineffective treatments. If no agreement can be reached, the patient might have to change physician.
And what if the physician does inform the patient adequately, but also prescribes homeopathy because the patient insists on it? In this case, the possibility of the above harms still applies.
Scenario 3: Patient with a life-threatening condition
Consider a young man with testicular cancer (a malignancy with a good prognosis if adequately treated). He consults his doctor, who prescribes homeopathic remedies as the sole therapy. In such a situation, the physician is grossly negligent and could be struck off because of it.
What if the physician prescribed homeopathy because the patient asked him to do so, and refused conventional therapies? Again, in such a case, the physician has an ethical duty to inform the patient about the best evidence as it pertains to homeopathy and to the conventional treatment for his cancer; failure to do so would be negligent. Again, the patient is free to decide what they want to do, as is the physician. If no agreement can be reached, the patient might wish to change his doctor.
And what if the physician does inform the patient adequately, makes sure that he receives effective oncological treatments, but also prescribes homeopathy because the patient insists on it? In this case, there is still the possibility of harms regarding cost potential to leave the patient with the impression that homeopathy is effective, which may lead to further harm in the future.
In conclusion
These scenarios are of course theoretical and, in everyday practice, many other factors might need considering. They nevertheless demonstrate why the demand ‘first do no harm’ is today obsolete, and had to be replaced by ‘do more good than harm’.
The latter principle does not support homeopathy (or any other ineffective so-called alternative medicine [SCAM]). In other words, the use of allegedly harmless but ineffective treatments is not ethical. But what if a clinician strongly believes in the effectiveness of homeopathy (or other SCAM)? In this case, they are clearly not acting according to the best available evidence – and that, of course, is also unethical.
The conclusions of all this are, I think, twofold. First, the ethical imperative of ‘first do no harm’ is often misunderstood, particularly in the realm of SCAM. Second, it cannot provide a sound justification for employing therapies that are (allegedly) free of adverse effects.
