As reported, the Bavarian government has set aside almost half a million Euros for research to determine whether the over-use of antibiotics can be reduced by replacing them with homeopathic remedies. Homeopaths in and beyond Germany were delighted, of course, but many experts were bewildered (see also this or this, if you read German).
While the Bavarians are entering the planning stage of this research, I want to elaborate on the question what methodology might be employed for this task. As far as I can see, there are, depending on the precise research questions, various options.
IN VITRO TESTS OF HOMEOPATHICS
The most straight forward way to find out whether homeopathics are an alternative to antibiotics would be to screen them for antibiotic activity. For this, we would take all homeopathic remedies in several potencies that are commonly used, for instance D12 and C30, and add them to bacterial cultures. To cover even part of the range of homeopathic remedies, several thousand such tests would be required. The remedies that show activity in vitro would then be candidates for further clinical tests.
I doubt that this will generate meaningful findings. As homeopaths would probably point out quickly, they never claimed that their remedies have any antibiotic effects. Homeopathics work not via pharmacological mechanisms (there is none), they stimulate the vital force, the immune system, or whatever mystical force you fancy. Faced with the inevitably negative results of in vitro tests, homeopaths would merely shrug their shoulders and say: ‘we told you so’.
Thus it might be more constructive to go directly into animal models. Such tests could take several shapes and forms. For instance, scientists could infect animals with a bacterium and subsequently treat one group with a high potency homeopathic remedy and the control group with a placebo. If the homeopathic animals survive, while the controls die, the homeopathic treatment was effective.
Such concepts would run into problems on at least two levels. Firstly, any ethics committee worth its name would refuse to pass such a protocol and argue that it is not ethical to infect and then treat animals with two different types of placebo. Secondly, the homeopathic fraternity would explain that homeopathy must be individualised which cannot be done properly in animals. Faced with the inevitably negative results of such animal studies, homeopaths would merely shrug their shoulders and say: ‘we told you so’.
Homeopathy may, according to some homeopaths, defy in vitro and animal tests, but it is most certainly amenable to being tested in clinical trials. The simplest version of a clinical study would entail randomising a group of patients with bacterial infections – say pneumonia – into receiving either individualised homeopathy or placebo. Possibly, one could add a third group of patients being treated with appropriate antibiotics.
The problem here would again be the ethics; no proper ethic committee would pass such a concept (see above). Another problem might be that even the homeopathic fraternity would oppose such a study. Why? Because all but the most deluded homeopaths know only too well that the result of such a trial would be devastatingly negative for homeopathy.
Therefore, homeopaths are likely to go for a different study design, for instance, one where patients suspected to have a bacterial infection are randomised to two groups of GPs. One group of ‘normal’ GPs would proceed as usual, while the other group are also trained in homeopathy and would be free to give whatever they feel is right for each individual patient. With a bit of luck, the ‘normal’ GPs would over-prescribe antibiotics (because that’s what they are apparently doing routinely), while the homeopathic GPs would often use homeopathics instead.
Such a study would indeed generate a result alleging that the use of homeopathy reduces the use of antibiotics. Of course, to be truly ‘positive’ it would need to exclude any clinical outcome such as time to recovery, because that might not be in favour of homeopathy.
The problem might again be the ethics committee. Assuming they are scientifically switched on, they will see through the futility of a trial designed to produce the desired result. They might also argue that science is not for testing one faulty approach (over-prescribing) against another (homeopathy) and insist that science is about finding the best treatment (which is neither of the two).
There are, of course, many other study designs that could be considered. Generally, they fall into two different categories: if they are rigorous tests of a hypothesis, they are sure to produce a result unfavourable to homeopathy. Such studies will therefore be opposed to by the powerful homeopathic fraternity. If, however, studies are flimsily designed to generate a positive finding, they might be liked by homeopaths, yet rejected by scientists and ethicists.
A much easier solution to the question ‘does the use of homeopathy reduce the use of antibiotics’ might be to not do a trial at all, but to run a simple survey. For instance, one could retrospectively assess how many antibiotics 100 homeopathic GPs have prescribed during the last year and compare this to the figure of 100 over-prescribing, ‘normal’ GPs. This type of ‘research’ is a sure winner for the homeopaths. Therefore, I predict that they will advocate this or a similarly flawed concept.
Most politicians are scientifically illiterate to such a degree that they might actually agree to finance such a survey and then confuse correlation with causation by triumphantly stating that the use of homeopathy reduces over-prescribing of antibiotics. Few, I fear, will realise that there is only one method for reducing the over-prescribing of antibiotics: remind doctors what they all learnt in medical school, namely to prescribe antibiotics only in cases where they are indicated. And for that we evidently need no homeopathy or other SCAM.