Bad Science in The Guardian

Guardian “Bad Science” columnist Ben Goldacre today criticizes this “tedious” article on “Complementary Medicine” that appeared in his own newspaper. It will be interesting to see if The Guardian publishes Ben’s letter. The article, by Osteopath Nicola Sturzaker, was apparently inspired by the negative response to the recent proposed certification of homeopathic remedies. Her article is the usual collection of fallacies and woolly thinking.  She starts by playing victim:

The launch this month of the national rules system, designed to bring homeopathic remedies into line with licensed medicines and allow packaging on homeopathic products to describe the illnesses they claim to treat, brought a predictable collective sneer from the established medical profession.

Yes, but she fails to mention the sneering is predictable only because homeopathic medicine predictably doesn’t work. Sturzaker is claiming this response is unconstructive. If the quoted comments were all that were offered, I would agree – it would just be ad hominem. But the negative comments are backed by numerous studies. For example, this recent review of 110 homeopathy trials published in the Lancet that shows homeopathy is nothing more than placebo. The sneering comments are made following an examination of the evidence that shows homeopathy doesn’t work, and the sneering is because homeopaths and alties in general continue to ignore these studies. What else is left for real doctors to do? Quoting scientific studies is apparently useless as homeopaths just ignore them: sneering at the way homeopaths ignore science is a valid response.

The issue at stake (the lower levels of evidence-based proof that homeopathic treatments have to provide compared with conventional drug treatments) is well worth debating and, in truth, the latest decision in favour of homeopathy may actually have tipped the balance too far.

This is a strange comment. Too far? How does she know? On what basis is it “too far”? And how far “too far” has it gone? Homeopathy either works or it doesn’t; there is either evidence it works or there isn’t. If the evidence is that it doesn’t work, it shouldn’t be certified at all – there is no degree of partial certification that is far enough but not “too far”. And if evidence is not going to be used as the criterion, how does one know how far to go before it is “too far”? This is just woolly thinking.

Michael Baum, emeritus professor of surgery at University College London got his in early: "This is like licensing a witches' brew as a medicine so long as the bat wings are sterile," he said, before going on to further dark murmurings about "witchcraft".

Yes, I think that is a pretty good analogy..

The first is the implied infallibility of the mainstream health professions, namely that their methods are always right, and ours are always wrong. Clearly, neither is true. Anecdotally, we all know of cases where GPs failed to notice something serious and doled out take-two-aspirin-and-go-to-bed-early advice when the realities were far more serious.

Here we start with the fallacies – a straw man. No one is claiming evidence based medicine (EBM) is infallible – “always right” as she puts it. What is claimed is that (a) there is evidence EBM works – better than chance and better than placebo, and (b) homeopathy doesn’t work. Of course, it’s easier to attack this straw man version of EBM than it is to show evidence that homeopathy works.

The second half of the paragraph, and much of what follows, is just a fallacious appeal to science was wrong before. Of course doctors are sometimes wrong and they sometimes miss a serious illness, but that doesn’t mean that homeopathy works.

Again and again, doctors refer to evidence bases as their catch-all for ruling out complementary medicine. In practice this is fine - though it comes from the profession that brought you thalidomide and is beginning to wonder whether rushing herceptin through the net was really so wise.

Another appeal to science was wrong before. As Ben writes: “thalidomide pops up from 1957”. It’s just a red herring: she  doesn’t describe any alternative to “evidence bases as [a] catch-all for ruling out complementary medicine”. This criticism of “evidence bases” without any alternative demonstrates the vacuousness of Sturzaker’s article and of alties in general.

However, the medical profession is a multi-million pound industry backed by pharmacological giants. The complementary sector cannot compete - we cannot pay for trials, we do not have multi-nationals encouraging treatment dependency.

Nonsense. Homeopathy is a multi-million dollar business with plenty of money for poor studies biased to show it works. They have money for trials, but they shy away from  well designed trials because well designed trials all show homeopathy doesn’t work.

It is the exact opposite - if there is a theory underlying all complementary medicine, it is that the human body works quite well on its own and needs tweaking as little and as naturally as possible. If you want clinical evidence - how about millions of years of human history?

LOL – “if there is a theory underlying all complementary medicine…” If? At least she acknowledges this is in doubt. The claim that things were just fine before EBM is just plain absurd: before 1900 life expectancies in the US were in the 30s and 40s, many women died in childbirth and many children died before reaching adulthood.  Millions of years of human history, indeed.

Ironically, one of the few areas that a large-scale trial has been done is the area that started this current row. Homeopathic medicine is indeed controversial, as in order for a homeopath to treat a patient, the person's individual symptoms have to be taken into account in order to make an individualised prescription. This means that homeopathy does not perform exceptionally well in random controlled trials - where one group of people are all given the same medicine and another group are given a placebo.

Nonsense. There is a perfectly straightforward way to test homeopathic individual remedies and procedures. The subjects of the trial all visit a homeopath, and all get the full consultation and individualized homeopathic prescription exactly as specified by the homeopath. Then, using a randomized double-blind protocol, half the patients get the specific individualized homeopathic remedy as prescribed by the homeopath, and the other half get sugar pills. (Yes I know they’re both getting sugar pills – you know what I mean.) Not that difficult, really.

When homeopathic trials are based upon individualised prescriptions we see a very different picture. At the end of 2005, the results of a large six-year study of 6,500 patients at Bristol Homeopathic Hospital reported 75% improvement in their health.

Nonsense, again. As Ben points out, the study was not a trial at all. As I wrote here, all that happened is that patients who visited a homeopath were simply asked by the homeopath treating them if they “felt better”, and 75% of them apparently said yes. Big deal. As Ben ends his letter:

It is this level of ignorance about the most basic concepts in evidence based medicine which makes the debate with alternative therapists so fabulously circular.

And, I would add, frustrating. But it’s a debate we can’t shirk.

Comments

[bsa]

One word: AARRRGG! One of the Science blogs mentioned "a pinata of stupid" earlier this month; seems apt in this case.

[Jimmy_Blue]

I seem to recall another facet of the 2005 study that puts a big hole in the homeopath argument, some of the patients included in the homeopathy trial as success stories were also still recieving conventional medical treatment at the same time as the homeopathy? So, no surprise they decided they felt better surely?

Anyone else heard about this or is my memory playing games with me?

[Big Al]

I was interested in the mention of thalidomide: perhaps Ms Sturzaker ought to read up on its history.

In the UK, thalidomide was leapt upon by the Distillers' Company, who saw it as a safe way to enhance one's lifestyle. They weren't Big Pharma by any stretch of the imagination.

Early studies did show an increase in foetus resorbtion in rabbits (often a bad sign) and Sir George Somers, the only proper pharmacologist to work on the drug said that on the basis of the animal trials, he would have "thrown it out the window".

However, he was released from his contract and effectively gagged. Chemie Gruenenthal, the originators of the drug, backpedalled and denied any problems with thalidomide, but the evidence began to mount.

One Australian doctor, McBride, examined a phocomelic ("flipper-limbed") child, and the affliction was so rare he had to look it up in a Victorian textbook. In the next few weeks, he saw several more.

The fact is that,even as a flawed and dangerous drug, thalidomide's effects were so strong and evident that even Gruenenthal's big-shot lawyers couldn't explain them away, try as they might.

The science of drug administration and testing was tightened as a result, and Distillers dropped out of the "recreational drug" business. Heads rolled - contrast that with the "Big Pharma will cover it up!" mantra.

Contrast that with homeopathy. Hasn't changed for 200 years, and its effects, if any, are buried in the statistical noise floor. It never changes, because it doesn't have to: it doesn't do anything in the first place.

[outeast]

She does (implicitly) raise one important point: homeopaths often get good results through having lengthy, detailed consultations, while conventionally qualified doctors often '[dole] out take-two-aspirin-and-go-to-bed-early advice when the realities [are] far more serious.'

The pitch is queered. Conventional doctors need to be given more time for proper diagnoses - this would reduce the misdiagnoses and inappropriate treatments that result and which power the CM market as well as causing suffering.

(Bear in mind this is from the perspective of someone familiar with NHS care, where doctors are well-paid but given ridiculously small amounts of time per patient.)

"homeopaths often get good results" - they do?

Did I miss something?

[outeast]

Did I miss something?

I guess so. If homeopaths didn't get good results then homeopathy wouldn't be so popular... or so successful in non-placebo-controlled 'studies'. Note that I wrote through having lengthy, detailed consultations - the consultation process itself gives the treatments effect: treatment provided by a trained homeopath can constitute a very effective placebo (not all placebos are equal).

One reason this queers the pitch is that conventional patients do not benefit from the added fillip of the placebo effect (so it's 'effective placebo + ineffective treatment' against 'minimal placebo + effective treatment': 'effective placebo + effective treatment' would undoubtedly be more effective).

The second reason is that, as I said, doctors do not (get the chance to) discuss symptoms in as much depth as might be necessary. Patients are pelied upon to communicate their problems, but a patient visiting the doctor because of recent severe headaches (say) may not think to report the pains in his legs that he's been having for months - even though both might be due to the same parasite. The knowledge of a patient that a 'holistic' practitioner such as a homeopath or acupuncturist elicits would be invaluable to a conventional doctor.

The trouble is that people assume that any apparent improvement (usually just that they think they feel a bit better) is down to the tap-water "remedies" themselves, and some of the patients then eschew conventional treatments.

That's OK if we're just talking about the odd headache (I'm a firm believer in avoiding medication when not needed), but the worry is that cancer patients or people with other life-threatening diseases will waive much-needed mainstream treatment.

The homeopathy brigade like to slag off "allopathic" treatments, since, supposedly, they don't treat "the whole person". I therefore do not find it inconceivable that some homeopaths would never consider handing over their "valuable insights" to a doctor for nothing.

Another thing: do all homeopaths and acupuncturists carry out these detailed examinations? Do they understand the implications of certain symptoms? On learning of a pain in a certain part of the body, do they go to the trouble of pinning it down precisely, as a good doctor would?

Or is it a case of "Stomach pains? Just take some homeopathic arsenic"?

I'd bet that there are many homeopaths who only perfrom the most perfunctory survey, just as there are doctors who take the time to get to the root of their patients' problems.

It's as easy to generalise about one discipline as the other. However, consider when homeopathy came into being (18th century). Medicine as we understand it was non-existent then, but there was no sudden fall in death rate due to the introduction of Herr Hahnemann's marvellous discovery. If homeopathy was any good whatsoever, surely it would have had a noticeable effect on the huge mortality rates?

Yet the introduction of Salvarsan, penicillin, antisepsis, anaesthesia and other medical advances all had dramatic influences on public health. No arguments. You didn't have to believe in them for them to work.

At the beginning of the 20th century, average life expectancy in the developed world was, I believe, 38 years. Homeopathy had been around for over a century, but it didn't seem to do much, despite the supposed value of lengthy consultation.

Modern "allopathic" medicine has also been around for about a century, but the average life expectancy now is around 80 years.

[outeast]

You're preaching to the choir, unnamed one!

What got me thinking about the consulitation issue, though, is that I recently met up with an old friend who is now an acupuncturist. While I have - let's say - doubts about the efficacy of acupuncture per se, it became clear in conversation with him that an acupuncturist really does seek to learn about all that ails a patient rather than relying on the patient's first reported problem.

While he then goes on to give (probably) worthless treatment, the investment of time and effort he puts into learnng about his patients is laudable and is something that should be more common among doctors. My own experience of doctors has been that they tend not to be very rigorous about eliciting reports of symptoms - despite the fact that many diseases produce a wide array of symptoms, often not obviously related, without which a correct diagnosis may not be forthcoming.

This is about doctors' attitudes (yes, I know it's far from universal) combined with the problems of heavy patient loads. As an example, take this comment from the Endometriosis Association of Victoria:

...[F]or many women the road to diagnosis is long and drawn out. ...it took the women an average of four years to obtain a diagnosis after they had first reported their symptoms to a doctor... partly due to doctors' attitudes to the symptoms of endometriosis. Many doctors do not listen seriously to women's complaints about their symptoms ... and therefore do not consider that there may be a physical cause for such complaints.

It's easy to sneer at 'holistic' approaches (I do!), but doctors were to elicit that level of information and dedicate a similar amount of attention to diagnosis then conventional medicine might not have such a tarnished image.

Again, I am not endorsing altie treatments. I am exclusively pro-EBM - jst think that diagnosis it not what it could be, and that a greater investment of time in diagnosis would also reap health benefits through placebo and thus further boost the success rate of contemporary medicine.

PS Mortality rates are only a good measure of success in treating fatal illness. Quality of life measures are important in the treatment of chronic disease, and it's there that alti treatments often score well (see the study I referenced above) because it is there that the placebo effect can be most effective. Dont scorn the placebo!

[Big Al]

Sorry, outeast, I forgot to sign the last post. I do agree that medicine would benefit if doctors had more time to spend with patients.

However, I also think that it's in the interests of alties' bank balances to find as much wrong with you as possible - the fad spa "doctors" of the 18th centuries certainly had the knack of finding all sorts of ailments you never knew you had.

I think a non-medically-qualified homeopath is likely to interpret every little ache and pain as a major problem, muddying the water and raising a plethora of possibly insignificant health issues.

It's not just length of consultation, but quality. I'd like to see some real evidence that the detailed quizzing of altie practitioners gets right to the nub of the patients' health problems.

Does anyone, ever, go to an altie practitioner who says, "You're perfectly healthy"?

I doubt it. I imagine if I went to a homeopath, he'd read equal importance into every least twinge, ache or discomfort,and I'd end up convinced that I was one step away from being a basket case.

So that's treating the "whole patient", is it?

Doctors can ask the right questions to zero in on a problem, not just scattergun every least pain. No,they can't spend all the time they'd probably want to on each patient, but at least they're likely to get to the crux of the matter.

The fact is, everyone may have a whole plethora of ailments at any one time, but if it was really affecting your quality of life, you wouldn't fail to bring it to the doctor's attention.