Anyone know about this homeopathy trial?

Diabolos

Thinker
Joined
Jul 25, 2005
Messages
211
First post - so please go gentle!

A friend who's open to the idea of homeopathy gleefully pointed out an article that appeared in the Sunday Telegraph (in the UK) entitled: Homeopathy Wins in Test With Medicine, knowing that I've always been rather skeptical that it can be anything more than placebo.

Sadly the article is very brief and I've been unable to trace any further information about it, so thought some of you guys might know about it:
http://tinyurl.com/7peqk

Apparently it measured homeopathy versus conventional medicine and the former worked better! There must be a flaw in there somewhere.
 
"Researchers in Germany"

No names, no references to the study itself. I'm not the research expert many of the other posters here are, but I'm skeptical towards the study's existence, not just its results.
 
BronzeDog said:
"Researchers in Germany"

No names, no references to the study itself. I'm not the research expert many of the other posters here are, but I'm skeptical towards the study's existence, not just its results.
I take your point, but the Sunday Telegraph is a respected newspaper, so it would surprise me if they just cobbled together some heresay. A reference would have helped though - I tried googling but nothing obvious came back.
 
Diabolos said:
I take your point, but the Sunday Telegraph is a respected newspaper, so it would surprise me if they just cobbled together some heresay. A reference would have helped though - I tried googling but nothing obvious came back.
I have a hard time evaluating what newspapers really deserve the respect they get, especially when it comes to health-related news. The sorts of things that show up on Stats.org has pretty much convinced me not to bother reading health news, unless it comes from a skeptical source.

Don't know how Reuters compares to the Sunday Telegraph, but they've had one problem with reporters making stuff up.
 
I'm not familiar with this study, it doesn't ring any bells, so I wonder if it's something new? If so, we'll probably get more details and a reference from another source soon.

From what they say, I think this is looking at different ways of administering placebo.
long-term health problems ranging from sinusitis to insomnia and depression
Just the sort of chronic, perhaps psychosomatic complaints that seem to respond to tea and sympathy. Without fuller and better details it's impossible to say, but I suspect that simply handing over a prescription for a painkiller or a sleeping pill or an antidepressant might not be the best way to approach people like that. Homoeopathy involves a great deal of patient/practitioner contact, and it can take an hour or more to take the sort of case history they routinely require. This has been described as a "therapeutic consultation", and seems to evoke a powerful Hawthorne effect in many people.

So, maybe homoeopathy "works" in this study in the sense that having a long consultation with someone sympathetic "works". If the non-homoeopathic group didn't have that, then it's no comparison at all.

The only way you can test homoeopathy as regards the efficacy of the remedies (as opposed to the tea and sympathy part) is to give everybody the same consultation routine, but only half of the patients get the prescribed remedies - the other half get a placebo. When this is done, typically both groups improve, but giving the remedy doesn't seem to make any difference.

This study may have interesting implications as regards the approach adopted by the "conventional therapy" arm of the trial, but unless there's a lot more here that wasn't in the report, I doubt if it tells us anything about the efficacy of the magic sugar pills.

Rolfe.
 
The first paragraph also tells us that the report is horsefeathers;

"Homeopathy has been proved more successful and cost-effective than conventional medicine in the first comparison of the two approaches."

The "first" comparison? There have been many comparisons and the better the controls are done the less likely homeopathy is to show any effect over placebo. End of story.
 
Diabolos said:
Apparently it measured homeopathy versus conventional medicine and the former worked better! There must be a flaw in there somewhere.
Such trials tend to be done by homeopaths themselves. Amazingly, they always find the effect they’re looking for.

They don’t normally do randomised, placebo controlled, double blind studies.

Doing open trials leaves the “researchers” open to all the confirmatory biases that lead them to find that which they wish to find. The flaw, in all likelihood, is that the trial was not a scientific one.
 
I saw that report in the Telegraph - I was considering writing a letter to the paper about it but haven't got around to it yet. If I do get around to it I'll report any response here.
 
This approach has been tried before I think, though then it was Stemetil (I think) which was the "conventional" comparison. In that case, however, they simply managed to show that the homoeopathy and the conventional treatment had a similar level of success.

Basically the trick is to pick a condition which is likely to be quite susceptible to tea and sympathy and which has a recognised "conventional" treatment which is quite possibly useless. Thus both the homoeopathy and the control group are being given useless pills. The homoeopathy group gets the tea and sympathy and useless pills. The conventional group gets a quick five minutes and a prescription for useless pills. Hey presto, homoeopathy does better than conventional medicine!

If you can cherry-pick your patients for the study, concentrating on those most prone to hypochondria (and frankly, if you're a homoeopath you get plenty of these) then you're virtually assured of success.

Rolfe.
 
Rolfe said:
This approach has been tried before I think, though then it was Stemetil (I think) which was the "conventional" comparison. In that case, however, they simply managed to show that the homoeopathy and the conventional treatment had a similar level of success.

Basically the trick is to pick a condition which is likely to be quite suusceptible to tea and sympathy and which has a recognised "conventional" treatment which is quite possibly useless. Thus both the homoeopathy and the control group are being given useless pills. The homoeopathy group gets the tea and sympathy and useless pills. The conventional group gets a quick five minutes and a prescription for useless pills. Hey presto, homoeopathy does better than conventional medicine!

Good point! I'd not picked up on the benefits of using useless pills (doled out after a brief consultation) as the drug comparator for the homeopathic remedy. You're right, it gives them a trial that looks pretty good on a casual or uninformed inspection, but is pretty well guaranteed to give them the positive result they want.

As with all medical trials, the real brains go into working out how to create good controls.
 
The only recent (as in this year) German study comparing homeopathy and conventional therapy I could find is this one, but I can't access the full paper, or read German, to make sure.
 
Donks said:
The only recent (as in this year) German study comparing homeopathy and conventional therapy I could find is this one, but I can't access the full paper, or read German, to make sure.
Two of the authors of that paper (Klein and Weiser) have, working together, published six studies indexed on PubMed, all of them comparing a homeopathic treatment with a "conventional" or "usual" one, and, as far as I can tell from the abstracts, none of them placebo-controlled. The other paper they have published together is a meta-analysis of similar clinical trials.

The best fit to the study mentioned in the Torygraph seems to be this one, at least in terms of the number of patients and the outcome, but it doesn't cover the variety of conditions mentioned in the newspaper story.
 
Donks said:
The only recent (as in this year) German study comparing homeopathy and conventional therapy I could find is this one, but I can't access the full paper, or read German, to make sure.
It doesn't look like the same thing. But it's the same sort of thing. Trialling a homoeopathic remedy against a minor OTC decongestant for "a bit of a cold". And they performed equally well.

Well, was it blinded? Were the two remedies administered in the same way? They admit that the aim was to show the "non-inferiority" of the homoeopathic remedy. In the abstract we aren't told what the variables were. Pick the right variables and you'd easily get the result you wanted.

Xylometazoline is quite good for unblocking a blocked nostril. That's all it does. I can't say I've tried spraying plain water to see if that works too, but it might do something. My point is that if you choose outcomes other than unblocking a blosked nostril, you won't find anything. Xylometazoline doesn't reduce inflammation or shorten the duration of your cold, and indeed you're warned that if you use it too often it might actually cause a bit of a rhinitis.

So, another fish in the barrel duly despatched.

Now, let's see you compare homoeopathy with insulin for type I diabetes. Oh, the nasty ethics committees won't let you? Oh, what a shame!

Rolfe.
 
Mojo said:
Two of the authors of that paper (Klein and Weiser) have, working together, published six studies indexed on PubMed, all of them comparing a homeopathic treatment with a "conventional" or "usual" one, and, as far as I can tell from the abstracts, none of them placebo-controlled. The other paper they have published together is a meta-analysis of similar clinical trials.

The best fit to the study mentioned in the Torygraph seems to be this one, at least in terms of the number of patients and the outcome, but it doesn't cover the variety of conditions mentioned in the newspaper story.
That doesn't fit either. Could it be the meta-analysis the Torygraph is on about?

These guys have got a nice little scam going here. Compare magic sugar pills and tea and lots of sympathy with some minor "conventional" treatment which probably doesn't do a lot and no tea and no sympathy. Who do you think is going to say they feel better? This is quite obviously not blinded, and basically it's shoo-in to give homoeopathy a victory.

Nice one, Jerry.

Rolfe.
 
Eos of the Eons said:
Yeahhhh, that's right, quote the flawed study :rolleyes: :rolleyes: :rolleyes:

I quoted the poster's (Peter's) comments.

Which brings us back to the perennial question. Alt Med researchers have no legitimate excuse for not performing properly controlled studies. But they keep performing exactly the kind of trials that permit their biased observations to give them the result they need. Well, I don't think we can really say they are pursuing the truth of what is going on, so are they really this stupid or is it a cynical attempt abuse the methods of science to give them results that make them look good?
 
Eos of the Eons said:
Yeahhhh, that's right, quote the flawed study :rolleyes: :rolleyes: :rolleyes:

I quoted the poster's (Peter's) comments.
More of what Peter said.
The study is yet another reflecting the trend for many acupuncturists, homeopaths, chiropractors (with at least certain conditions) and others to have given up on the task of showing that their methods act as anything other than an elaborate placebo c.f. the comment: "The study ignored the question of how homeopathy might work and focused on how well it performs."

But there is another way of looking good. What you do is compare the apparent outcomes of two very different medical scenarios.

One scenario involves a boring treatment of limited efficacy and patient appeal i.e. bog-standard, ho-hum normal medical care.

The other scenario includes the "special treatment". The "special treatment" is little more elaborate, theatrical, mysterious, intensive, or "hands on".

It also involves practitioners who are a little more desperate to get the right result and who may thus try so hard as to trigger answers of politeness or experimental subordination in their subjects, or who may unconsciously fudge the results slightly when involved in unblinded assessment of them, as in the above study.

A third component is to carefully select conditions where placebos can appear to have pronounced effect (note the "selected chronic diagnoses" chosen for this trial) if only because most patients get better over time.

All that has been proved is that a lot almost any kind of nurture makes peoples feel better or makes them less likely to say they have NOT been helped by the care.
He's saying pretty much the same as we've been saying in this thread. The study is really designed to demonstrate how well homoeopathy applies the placebo effect.

Barb, whenever a well-designed, blinded and controlled study shows null effect for homoeopathy, you seem to find some way to declare that homoeopathy isn't susceptible to scientific study. Why, then, when an allegedly "scientific" study seems to have given the results you wanted, do you simply chip in with the "p<0.001" comment (which, to be frank, makes you sound exactly like Xanta)?

If you were trying to point out that this was an alleged effect on children, do we really have to go over yet again how the problem of observer bias rears its ugly head when dealing with patients by proxy (parents or animal owners) and how an unblinded study is even more likely to give the desired result under those circumstances?

It's obvious to anyone with even half a brain cell that this study design is deliberately contrived to show homoeopathy in a positive light, by concentrating on the Hawthorne effect (or therapeutic consultation, if you like) rather than actually trying to demonstrate whether the sugar pills are doing anything.

Why do you think this is? Why do you think that studies designed to show that "any kind of nurture makes people feel better or makes them less likely to say they have NOT been helped" invariably do show exactly that, but as soon as you design the study to find out whether the sugar pills themselves contribute anything, you find that there's nothing there?

You repeatedly say that the people you "treat" say they feel better, and their response validates your beliefs. Can you not see that this study is yet one more piece of evidence that it is your sweet smile and sympathetic manner that is doing the trick, and that the suger pills are a sham and a deception?

Rolfe.
 
I've just ordered a copy of the paper from the RCVS library so I can read the protocol in detail.

I think I'll write to the authors and ask them to explain why they designed it the way they did, given that it is obvious that the design could not lead to defensible conclusions. I wonder whether they'll reply.
 

Back
Top Bottom