Elliott, M., Br Homeopath J. 2001 Jan;90(1):33-6.
This paper is just over three pages long, but it manages to pack in a quota of howlers worthy of a much longer document.
It is very easy to spot the deficiencies in normal scientific terms.
- Cushing's disease is very different in the two species investigated, and this must be the only paper in the entire literature canon which lumps them together as if they were the same condition (18 dogs and 21 horses are considered as a single group and treated in an identical manner).
- There was no conclusive evidence that any of the animals in the trial actually had Cushing's in the first place. Nearly half were only diagnosed on clinical signs, while the majority of the remainder only had basic routine lab tests done, not the specialist hormone tests usually regarded as mandatory. Even where the hormone tests were done, no information is given regarding results or cut-off criteria for inclusion. (As an aside, Mark is a client of my laboratory, and I was entirely unaware that he was doing this study. If he had consulted me, I would have offered him cut-price confirmatory hormone testing in the interests of his research, but he preferred simply to accept other veterinary surgeons' diagnoses, no matter how poorly supported, rather than attempting to confirm what was going on.)
- There was no control (untreated or placebo-treated) group, all 41 animals were treated.
- There was no blinding, everybody concerned seems to have known exactly what was going on.
- Whether or not the animals improved was assessed purely on subjective clinical examination - no follow-up laboratory testing was carried out.
- The person doing that assessing appears to have been Mr. Elliott himself.
- Many of the patients were only followed up for two months, which isn't very long in this context.
However, even more startling is the list of irregularities from the homoeopathic point of view. One feature of homoeopathic discourse is the propensity of homoepaths for dismissing negative studies carried out by non-homoeopaths on the grounds that proper homoeopathic procedures have not been followed. It is therefore quite astounding to see that most of these objections could equally well be applied to this positive study carried out by a VetMFHom.
- Inclusion in the study was based on an "allopathic" diagnosis of Cushing's disease. However, homoeopathy rejects the concept of aetiological diagnosis in favour of treating each case acording to the particular clinical symptoms exhibited.
- There was no individualisation, all patients were treated with the identical remedy irrespective of clinical signs - which must have varied quite a bit. In particular, the clinical signs of Cushing's disease vary markedly between the dog and the horse - for example, dogs go bald while horses grow a lot of hair, and horses get sore feet while dogs don't. One thing we are always being told about homoeopathy is the absolute necessity for individualising treatment to the patient's particular presenting signs, and this is usually seen as non-negotiable. However, Mr. Elliott just announces (in utter defiance of 200 years of homoeopathic practice) that he's going to see if a standardised approach will work, and carries on.
- The first remedy chosen was potentised ACTH. Cushing's disease is characterised (in the majority of cases) by pathologically increased blood concentrations of ACTH. This means that using potentised ACTH to treat it is not homoeopathy, but isopathy - something we're always reminded of every time there is a report of puppies which have received nosode prophylaxis going down with clinical parvovirus. To compound the issue, the author accepts that this approach would not be expected to work in dogs with the other (adrenal) form of Cushing's, where blood ACTH concentrations are abnormally low. However, no steps were taken to identify any dogs with this form of the disease and exclude them from the study.
- The other remedy used was Quercus robur. Although some reference was made to the proving symptoms of this remedy, it appears to have been selected not by symptom matching but by dowsing. Dowsing was not something advocated by Hahnemann, and it appears to have a poor track record in homoeopathy (McCartney et al. 2002).
- Both remedies were administered together to all patients. This is "polypharmacy", usually regarded as a heinous crime by homoeopaths, who insist that remedies must be administered singly.
If this approach were to have any chance of becoming accepted, it would be essential to repeat the study in a properly scientific manner - that is, if ethical approval could be obtained for such an exercise, something which we wonder if Mr. Elliott even thought about. Horses and dogs would have to be considered separately; definite confirmation of diagnosis would be required for every animal entering the study; animals would have to be split into treated and placebo-control groups and all participants blinded as to which was which; all animals would have to be followed up for a sufficiently long period; and assessment of clinical response would have to be strictly objective including the sort of laboratory studies which are mandatory when following up patients treated with the licensed treatment for this condition.
Supposing that were done, what do you think the outcome would be? Yes, we know. But the homoeopaths would have the answer for that. "This study was not conducted according to proper homoeopathic principles!" Allopathic diagnosis, absence of individualisation, isopathic selection of remedy and polypharmacy have all been invoked in the past to dismiss studies the homoeopaths didn't like. (We're not sure about the dowsing, because it's not the sort of thing scientific researchers usually include in their protocols.) Why, then are these acceptable (and all in the same study, no less!) just because a positive result is claimed?
Actually, that's precisely the point. Homoeopaths seem to be able to do things exactly as they please, without criticism, so long as they are within the faith and they claim success. However, as soon as success is not achieved, it is always possible to find some reason to explain away the lack of effect - as there is almost no consensus of agreement about how cases should be managed, there is plenty of scope for rationalising away lack of efficacy. The one explanation which will never be countenanced is that whatever the outcome of the case, it would have happened anyway irrespective of the administration of the magic sugar pills.
Another consideration is that this exercise was quite clearly a clinical trial. However, clinical trials have to be approved and licensed and authorised by ethics committees and things like that. Vets can't just blithely decide to give experimental medicines to their patients in the interests of science, on their own initiative. Any reputable journal would have confirmed that the requisite licences had been granted for any clinical trial they published. However, it is completely impossible that any licence would have been granted for this study, because of its glaring scientific inadequacies. It seems likely this trial was actually illegal.
It seems a real pity that Horizon chose to highlight this deeply flawed piece of work as evidence that homoeopathic remedies work in animals, without checking its credibility, or giving the slightest acknowledgement to the scientifically-conducted studies which have shown no effect.
back to Home Page | back to Horizon review