Should doctors be allowed to prescribe homeopathic treatments on the NHS? John Worrall looks at the scientific and ethical status of homeopathy.

GPs were long able to prescribe homeopathic preparations on the NHS. Within the past few months, the UK government has banned this practice (though the ban is currently subject to appeal). Is the ban on NHS prescriptions for homeopathic preparations justified? In a previous post I showed that there is, surprisingly, an evidence-based argument for allowing such prescriptions – though I never, of course, suggested that the argument was, on its own, compelling. Here I reconsider that argument, and also introduce ethical issues into the debate alongside the scientific ones to come to a view of whether or not the ban is justified.

As before, let me be clear: I do not “believe in homeopathy” – the theory behind homeopathy is as pseudoscientific as can be; and, so far as whether or not it works is concerned, there is no evidence that homeopathy outperforms placebo. However, homeopathic treatments, of course, do no worse than (acknowledged) placebo; and there is evidence for a real placebo effect for conditions like mild-to-moderate depression, irritable bowel syndrome (IBS), the common cold and (especially) pain (though also Parkinson’s Disease and some other neurological conditions). So, while you certainly do not want to be taking homeopathic preparations if you are unlucky enough to have lung cancer, or congestive heart failure or any other major, non-neurological disease, there is evidence that those preparations can be effective if you suffer from one of the limited range of conditions I just cited (depression, IBS, pain …); and so, contrary to a widely held view, prescribing placebos, in particular homeopathic placebos, is not (always) the equivalent of doing nothing.

But, even if homeopathic treatments can be effective, why should it be NHS practitioners who prescribe them? Why shouldn’t patients be required to access those treatments via a private homeopath instead? Well, for one thing, a properly qualified NHS practitioner is much more likely than a homeopath to spot conditions that will not respond to a placebo and so more likely to ensure proper treatment of such conditions. But the central part of my earlier argument depends on an interesting, and again evidence-based, fact about placebos: that the effect of taking a placebo is, in general, greater if you don’t know that what you are taking is a placebo. This is because a major part of the placebo effect is caused by expectations of recovery – expectations that activate the body’s internal pharmacopeia (principally via discharge of endogenous opiates into the bloodstream). Patients who already have some degree of belief that homeopathic treatments are not simply placebos may have those expectations raised, and hence may experience a better result, if the treatment is prescribed, under the NHS, by an authoritative fully-qualified doctor.

In sum: there is scientific evidence that a homeopathic placebo may help a limited range of patients (those who are not sceptical about the theory of homeopathy but nonetheless accord greater respect to orthodox, rather than “complementary” practitioners), suffering from a limited range of conditions (depression, IBS, pain …); and there is scientific evidence that the help those patients receive will be greater if they are not told that what they are getting is a placebo; that is, if they believe, against the scientific evidence, that taking the homeopathic preparation has some “specific” or “characteristic” effect beyond placebo (and therefore qualifies for prescription under the NHS).

But surely medical practice needs to be ethical as well as evidence-based. Is it ethical to prescribe a placebo, even if it is likely to be effective? More specifically, is it ethical for a GP to prescribe what s/he believes to be a placebo while encouraging the recipient to believe that what s/he is receiving is a “regular medicine” with a “real” (more accurately: specific or characteristic) effect?

There is a widespread view that prescribing placebos is, in fact, downright unethical. Many would agree with Singh and Ernst (Trick or Treatment: Alternative Medicine on Trial, p.55) that “almost any form of treatment that relies heavily on the placebo effect is fraudulent”. But can a treatment of a patient really count as fraudulent or unethical if the evidence indicates that the treatment is effective for the patient’s condition? Naturally, if there is a treatment with a specific effect on the patient’s condition then, since that treatment can safely be assumed also to carry the non-specific placebo effect (especially if delivered by an empathic doctor), that treatment should, generally, be preferred to a “mere” placebo. And there is in fact now evidence that, for example, certain antidepressants are effective for IBS – a placebo-responsive condition. But what if you are a GP and your IBS patient does not want to take an antidepressant; or tries one and suffers unduly from side effects?  Since there is strong evidence that placebos are somewhat effective for IBS, it would not only seem ethical to prescribe one for such a patient, it would seem unethical not to. (After all, one good thing about placebos is that they do not carry side-effects.)

For many, the crucial issue will be what the patient is told while being prescribed a placebo and specifically a homeopathic placebo.  A “principle of autonomy” is nowadays widely considered to be an important ethical constraint on medical treatment. This principle, roughly speaking, requires the doctor to act, not as an external, decisive decision-maker, but rather as an advisor to her “patient” whom she must consider as an autonomous agent. What the principle more exactly entails is not at all clear, but it surely does ban as unethical outright lying to a patient, even if lying would be in the patient’s interests.

However, there is no need for a physician to lie to a patient when prescribing a placebo.  Kaptchuk pointed out some years ago that IBS patients, for example, could quite honestly be told the following:

I am prescribing you placebo pills made of an inert substance, like sugar pills, that have been shown in clinical studies to produce significant improvement in IBS symptoms through mind-body self-healing processes.

Kaptchuk’s recommendation can be greatly improved (i) by dropping the implied endorsement of Cartesian mind-body dualism (a doctrine that has long been accepted as unscientific); (ii) by recognising that it is not the pills that produce the effect (how could they if they are ‘inert’?) but rather the treatment (or “treatment ritual”); and finally (iii) by bringing it better in line with current conceptions of patient autonomy. Taking these into consideration, a patient in the situation we are positing can truthfully be told the following:

I am offering you a treatment involving taking placebo pills made of an otherwise inert substance, like sugar pills – a treatment that has been shown in clinical studies to produce significant improvement in IBS symptoms through activating and enhancing the body’s self-healing processes.

This is undeniably honest and the evidence – perhaps surprisingly – indicates that giving that information to such patients will not entirely eliminate the placebo effect that the patient would have experienced had s/he been unaware that what they were receiving was a placebo. However, giving the information that the pill is inert, because it generally lowers expectations of a positive outcome, will very likely reduce that effect, and perhaps reduce it greatly.

Is this reduction in the placebo effect the unavoidable price of practising ethically?  This is a tempting view, but I am not so sure. First, let’s recognise that, since Kaptchuk made his recommendation for how to prescribe placebos ethically using IBS as his example, evidence has arisen that some antidepressants (marginally) outperform placebos for IBS.  Suppose, however, that an IBS patient has not flourished on the antidepressant; and suppose that the GP knows that the patient is positively inclined toward homeopathy, and says:

We tried you with the anti-depressant and you suffered side effects that you were unhappy with, so I would like to offer you instead a homeopathic pill that has been shown in clinical trials to have a positive effect on the symptoms of IBS. It seems to work via the activation of the body’s self-healing processes, but, in any event, the evidence is that the treatment does work; and, certainly, it is unlikely to carry any side-effects.

Is it an affront to the patient’s autonomy that this fails to add that the evidence-based view is that the homeopathic pill is a placebo, and that it might just as well be a sugar pill?  It is not clear to me that it is.

On the other hand (philosophers, of course, possess an endless supply of other hands), what seems ethical often depends on how widely we construe the decision problem at issue. Perhaps if we concentrate simply on whether or not the treatment is likely maximally to ameliorate the patient’s condition, we might be inclined to answer the above question “No: withholding information is not on an ethical par with outright deception and if, as in this case, withholding the information (that the homeopathic pill is a placebo) will likely increase benefit to the patient, then that’s compatible with the autonomy principle and therefore ethically acceptable”. Notice, however, that by acting in this way, the GP is knowingly exploiting a false belief that the patient has: the belief that homeopathic potions have effectiveness beyond the placebo effect. If the outcome of the treatment is indeed some relief of the patient’s symptoms, then acting in this way is very likely to reinforce that false, indeed pseudoscientific belief. That, in itself, seems ethically suspect.

So, when we take a wider-screen view – one that takes into account the impact of the decision to prescribe a homeopathic preparation not just on the patient’s symptoms but, more widely, on the health of his/her intellect – the ethics again become murkier. Indeed, my own view is that the ethical misdemeanour of giving encouragement to pseudoscience and of missing an opportunity to challenge a pseudoscientific belief outweighs any advantage to the patient of, say, an extra degree of relief from their symptoms of IBS.

In the end, then, I support the government’s decision to ban NHS prescriptions of homeopathic preparations. But, as we have seen, the considerations in favour of the ban are much more complex and nuanced than the ban’s supporters have supposed. Those supporters invariably claim that the ban is justified because homeopathic remedies have been “proven not to work”. But this is plain false. What is true is that there is evidence that homeopathic remedies do not work any better than placebos. But placebos, and so in particular homeopathic placebos, do work – for a limited range of conditions. What the pro-ban lobby needs, and what the above supplies, is an argument for banning NHS prescriptions of homeopathic treatments, even though they work for the condition presented by the patient – that is, an argument for banning prescribing such treatments despite the fact that they work.

By John Worrall


John Worrall is a Professor in the Department of Philosophy, Logic and Scientific Method at LSE.  As well as Evidence-based Medicine, he specialises in the philosophical issues raised by major theory-change in science; and is especially known for his work on structural realism. He is also the founder and lead guitarist of the Department’s house band, The Critique of Pure Rhythm.



Further reading

  • Benedetti, Fabrizio, 2014. Placebo Effects. Oxford University Press.
  • Singh, Simon and Edzard Ernst, 2009. Trick or Treatment?: Alternative Medicine on Trial. Corgi.
  • Worrall, John, 2016. “Homeopathy and Evidence-Based Policy“. The LSE Philosophy Blog.



Featured image: Giancarlo Foto4U / CC BY 2.0 (cropped from original)