Homeopathy: Rustum Roy drops the bomb

[wilsontown]

Interestingly, I got an e-mail from Homeopathy on our response to the Rao et al. paper. They wanted me to send them figures 1 to 3 that were missing from the original manuscript. I was a little puzzled by this, as there were no figures in the manuscript. Then I realised that they had taken references to the original figures in Rao et al. as being references to figures in our response. I was able to clear up this confusion, but it does make me wonder. Has the response been accepted for publication without anyone seeing the figures they thought were supposed to be included? There are alternative explanations, of course, but given what we've seen it perhaps wouldn't be surprising.

[shpalman]

Possibly, the person doing the typesetting saw references to figures but didn't read the text properly.

[Rolfe]

Originally Posted by wilsontown

Interestingly, I got an e-mail from Homeopathy on our response to the Rao et al. paper. They wanted me to send them figures 1 to 3 that were missing from the original manuscript. I was a little puzzled by this, as there were no figures in the manuscript. Then I realised that they had taken references to the original figures in Rao et al. as being references to figures in our response. I was able to clear up this confusion, but it does make me wonder. Has the response been accepted for publication without anyone seeing the figures they thought were supposed to be included? There are alternative explanations, of course, but given what we've seen it perhaps wouldn't be surprising.

Shpalman is probably right. I once wrote a noddy-type book on basic biochemistry interpretation for clinicians. I had a basic format of giving an approximate reference range for each analyte, a short description of the analyte's role in metabolism, then reasons for a decreased concentration, and for an increased concentration.

I gave the approximate reference range for bilirubin as (I think) 0-3 mcmol/l.

I got a letter from the editor asking me where the paragraph on reasons for a decrease in concentration had got to.

Rolfe.

[Deetee]

Originally Posted by Rolfe

Shpalman is probably right. I once wrote a noddy-type book on basic biochemistry interpretation for clinicians. I had a basic format of giving an approximate reference range for each analyte, a short description of the analyte's role in metabolism, then reasons for a decreased concentration, and for an increased concentration.

I gave the approximate reference range for bilirubin as (I think) 0-3 mcmol/l.

I got a letter from the editor asking me where the paragraph on reasons for a decrease in concentration had got to.

Rolfe.

I don't follow (....long day...)

Your range is wrong, no? So the editor was actually right, yes?

ETA - I see now.
In humans the range is 3-17.
But there are no pathological conditions associated with levels below this range afaik.

[Hellbound]

Deetee:

Rolfe works in veterinary medicine, so the range may well be right for the species she was discussing.

With that in mind, and the normal range starting at zero, how in the world could you have an abnormal decrease in concentration?



At least, that's my understanding of the post

[Procida]

Originally Posted by Rolfe

Procida's 4-hour visit seems to be over, for now anyway.

She was very quick to state her belief that Rao and friends were simply incompetent.

I do have a real life outside my working life.

If I prefer to read, and think, there is nothing wrong with that.

I very much appreciate your posts when you stick to science.

But not everyone is going to want to have an average of 7.5 posts a day, or to reply to either inuendo or epithet, even if they have to do that at least once, just to make the point.

[Deetee]

Originally Posted by Huntsman

Deetee:

Rolfe works in veterinary medicine, so the range may well be right for the species she was discussing.

With that in mind, and the normal range starting at zero, how in the world could you have an abnormal decrease in concentration?



At least, that's my understanding of the post

Yes - I got that eventually, but my brain was on holiday yesterday. I added in the comment about the human range for info, not to suggest Rolfe had got the vet range wrong.

[wilsontown]

Quote:

Possibly, the person doing the typesetting saw references to figures but didn't read the text properly.

Yes, that would seem to be a reasonable explanation.

[Rolfe]

Originally Posted by Deetee

I don't follow (....long day...)

Your range is wrong, no? So the editor was actually right, yes?

ETA - I see now.
In humans the range is 3-17.
But there are no pathological conditions associated with levels below this range afaik.

Yes. That human reference range owes rather a lot to the old gravimetric units, and to the conversion factor - the old unit was mg/100ml, and the conversion factor to get to mmol/l (OK, I tried this time but hell's bells, Darat, if "µ" or "μ" would just work, life would be so much easier, I really, really want the html back!) well, anyway, it's 17.1. You will find that upper reference limit quoted as 17.1mmol/l in quite a lot of publications, although the precision of the assay actually doesn't allow a significant figure after the decimal point.

I really don't know about man, but that's way too high for dogs and cats, although a lot of old publications still quote the 17.1 regardless. I know I'm getting distinctly suspicious summat ain't right from 4mmol/l upwards, and from about 5 I start to see a subtle jaundice evident in the plasma.

The idea of having a lower reference limit for this analyte is daft. As you say, there's no such thing as a pathologically low concentration. I get 0 or 1 as the most common actual results in normal dogs and cats. I can only assume that this is not the case in man, and presumably the lower limit of 3 was obtained from an actual population survey. I just can't figure out why that's so, given that the upper limit is quite clearly a translation of an old upper limit ot 1mg/100ml. I would have assumed the old-unit range would have been given as 0-1mg/100ml, but the "3" seems to assume an old-unit range of 0.2-1mg/100ml, which is a tad unlikely. To be honest, you can barely read at that level of precision, and down there you're into minor artefacts of haemolysis more than actual presence of bilirubin.

[Of course, once you change species again all bets are off. In the horse the reference range goes up to about 50mmol/l, though pony breeds usually seem to be lower which may have more to do with diet than anything else. Any increase in this is overwhelmingly likely just to be an effect of fasting, as all horses get fasting hyperbilirubinaemia (similar to Gilbert's syndrome in man). Sheep and cattle are somewhere in between, maybe closer to the human range, but since cattle have naturally yellow plasma in any case due not to bilirubin but to beta-carotene, it gets even more complicated.]

</biochemistry geek stuff>

Rolfe.

[Rolfe]

Originally Posted by wilsontown

Interestingly, I got an e-mail from Homeopathy on our response to the Rao et al. paper. They wanted me to send them figures 1 to 3 that were missing from the original manuscript. I was a little puzzled by this, as there were no figures in the manuscript. Then I realised that they had taken references to the original figures in Rao et al. as being references to figures in our response. I was able to clear up this confusion, but it does make me wonder. Has the response been accepted for publication without anyone seeing the figures they thought were supposed to be included? There are alternative explanations, of course, but given what we've seen it perhaps wouldn't be surprising.

Back to the actual letter.

I caught on to a slight error in terminology in the letter last night while wathcing University Challenge. The question was, what do you call the science of comparing the intensity of a light source with a light source of known intensity? I didn't catch on at all, and when the answer came back from Paxman as "photometry", I just said, shoot me now!

Of course, photometry is what this is all about, it's just that doing it at different wavelengths makes it spectrophotometry. That's the term I always use, and it's the correct one. I just got sucked into using the word "spectroscopy" because of the Sigma description of the ethanol as "spectroscopic grade" and I think Rao et al. may also use the word in the paper. But spectroscopy is a lot more about actual absorption and emission bands I think. Maybe at a pinch you could call this spectroscopy, but really the correct term is spectrophotometry.

So, in the last paragraph of the letter, the sentence reading

Quote:

It is clear that the data presented are wholly inadequate to support the authors’ assertion that UV spectroscopy can differentiate between the two remedies, and between different potencies of the remedies.

Should really read ".... assertion that UV spectrophotometry can differentiate...." And there's another instance where arguably the usage should be the sam.

This is industrial-strength nitpicking, and I really don't think we're going to ask the editor to correct it.

Rolfe.

[JJM]

Rolfe,

I think there is also a difference between British and American terminology. A while back, a colleague tried to teach me the difference between -ometry and -oscopy; but few people, here, care. I think this is an application of the Law of Conservation of Syllables (-photometry vs. -oscopy). The original article uses -oscopy.

I also had a nit to pick (I once raised body lice, that is a fact, so I am good at it). The letter refers to the absorbance peak at 330nm. However, someone went to the trouble of providing us with a nice graphic showing it at 324nm. (Go to the page that lists all the threads, and this thread is marked with a paperclip. Click on that to see the graphic.) I am not sure it is worth changing.

[Rolfe]

Yes, Geni did the demonstration that the peak was (as far as he could tell) at 324nm, but he was using the figure from the powerpoint slide show that is on the web, not the printed figure from the paper. My impression was that the peak seems slightly further to the right in the printed figure. Also, 324nm cannot be substantiated from the data presented in the paper, it's far too precise - and it sounds absolutely obsessive! Given the eyeball appearance of the printed graph, I thought that "about 330nm" actually described it best, and better than "about 325nm" - in that version, it does look a bit closer to 350 than to 300. Also, it does sweetly corroborate your remark about acetone.

Rolfe.

[wilsontown]

I'm told that proofs of the letter should be available on or about December 3rd. I'll send them around to all involved. I don't think that there would be any problem asking for some such minor corrections at the proof stage, if required. They only get upset if you want to change quite a lot of stuff.

Incidentally, Rolfe, you seem to have a fancy animated graphic on your posts suggesting it's your birthday. If so, many happy returns.

[Rolfe]

Why, thank you sir. Currently tryng not to be entirely buried under dead farm animals, but seem to have managed to swap a dead Friesian cow for a pile of small animal biochemistry results, so it's not all bad.

Rolfe.

[wilsontown]

Just in case folks hadn't seen this, there's an interesting account of a debate with Rustum Roy at this here blog...

It's going to be very, very interesting to see what his response to the letter can possibly be.

[Rolfe]

Yeah, I read that earlier.

Quote:

I pointed out, as many others have, that this study is worthless because he did not demonstrate that the different preparations (which were just obtained from a homeopathic company) had no chemical differences - to which he had no effective answer.

I was quite in awe of his masterly summary of the entire letter in one pithy statement.

Rolfe.

[Mojo]

Rusty has shown up in the Grauniad today, bleating about "homeophobia":

http://www.guardian.co.uk/comment/st...229446,00.html

[Badly Shaved Monkey]

Rustum Roy has been seen in captivity recently.

http://www.guardian.co.uk:80/comment...229446,00.html

Would one of the authors of the rebuttal to his paper like to comment?

[Badly Shaved Monkey]

Originally Posted by Mojo

Rusty has shown up in the Grauniad today, bleating about "homeophobia":

http://www.guardian.co.uk/comment/st...229446,00.html

Oops, sorry Mojo, didn't see you'd already created a link there.

[Blue Bubble]

*Bump*

What's the state of the soon-to-be-published letter ?

[wilsontown]

Got citation information for the letter: Homeopathy 97/1, p44-45. As yet I'm not sure when it will actually appear, but it should be very shortly. I'll keep everyone posted.

[wilsontown]

DOI info for the letter: 10.1016/j.homp.2007.10.004

Also, DOI info for my response to Martin F Chaplin: 10.1016/j.homp.2007.10.002

These haven't actually appeared yet, but should be fairly soon.

I don't have info on any response from the authors of the original articles, but I'll post it as soon as I do.

[shpalman]

Volume 97 issue 1 of Homeopathy is now online.

[Badly Shaved Monkey]

Wow! I suspect I am not the only one to have a problem with their reply. It is pretty brazen to move seemlessly from a claim to be able to differentiate him remedies to saying they are pleased they have demonstrated a quality assurance tool that, amazingly, can distinguish clean ethanol from contaminated ethanol.

And, apparently, Roy is not a proponent of homeopathy.

Quite shameless.

[fls]

Q: Why couldn't the changes have resulted from differences in composition (contamination)?

A: I don't know and don't care.

Linda

[Badly Shaved Monkey]

Of course I meant "hom" remedies not "him" remedies. Predictive text on the iPhone being too clever for its own good!

[fls]

"Kerr et al's remarks concerning the probable contamination in the sample of the original solvent sent to us, although apparently not in any of the dilution materials..."

The implication being that it is more likely that all we know of medicine, chemistry and physics can be overturned than that the stock solution used in the remedies was from a different batch. Actually, I'm unnecessarily harsh. Since he's neutral on the issue of homeopathy, he probably gives an equal probability to both.

Linda

[JJM]

Although Rao was the corresponding author, I was pissadointed that Roy did not join in the response. He likes to make insulting remarks about other scientists. Either Rao did not show him our letter, or he realized that his incompetence was unarguable.

[Rolfe]

Er, calling Pipirr - do you have any news for us? Is the PhD official yet?

Rolfe.

[Pipirr]

News? It's in the hands of the thesis committee! I'll let you all know when I know. But I hear it's been well received.

[Pipirr]

I thought that Rao et al. provided a non-answer. It's one thing to say that they provided homeopaths with 'potential quality control tools'. It's quite another to have actually demonstrated that. They really didn't show much of anything.

[Mojo]

Originally Posted by Pipirr

I thought that Rao et al. provided a non-answer. It's one thing to say that they provided homeopaths with 'potential quality control tools'. It's quite another to have actually demonstrated that.

I was kind of assuming he meant a quality control tool for assessing the quality of different batches of stock solvent.

[Mojo]

Anyway, why would they need a quality control tool for the potentised remedies? They are supposed to produce such distinctive results that I'm sure the homoeopaths would notice if anything was amiss.

[Rolfe]

I'm sorry, but I'm still gasping at the lameness of that reply. I suppose Darat would rap my knuckles if I posted it all here? Oh well.

First two paragraphs (at least half the reply) don't address the letter at all, just bang on again about pressures and structures. Total waste of paper.

Third paragraph appears first to concede that the "plain ethanol" was indeed contaminated ("probably"!), then to try to rescue something from that wreck. I think she actually means that they have made the great discovery that you can detect contamination in ethanol by UV spectroscopy!!! Wow! Whoda thunkit?!

Final paragraph, and this one I am going to quote as "fair use".

Quote:

All the analytical data shown in the paper are the result of reproducible analyses, although we appreciate the suggestion of representing such data as an average with a standard deviation, we emphasize that our key identification by display of an envelope demonstrates that, there are indeed differences beyond the standard deviation range among individual homeopathic remedies, as used in practice.

Translation: OK, we could have presented SDs and done some statistics, but we didn't so shut up about it, we assert that the data are reproducible and differences are there. Honest. How can anyone say "there are indeed differences beyond the standard deviation range" with a straight face, while refusing to present the actual standard deviations? (And anyway, tell me about +/- 2 SDs, and I might start to listen!)

She seems to be implying that the "plain ethanol" was sent by Hahnemann Labs with the rest of the stuff. I'm not at all sure I believe her. It's a bit telling that fact isn't in the M&M, when all the extraneous rubbish about temperature sensors and unique identifying numbers is. And I don't really think it's all that likely Hahnemann Labs would have ethanol as filthy as that kicking about the shop, as it were. I still think Roy picked up a dusty old bottle from a store room, probably opened multiple times over the past ten years!

But, let's assume she's telling the truth. So, the "plain ethanol" sent by the manufacturer demonstrates marked absorbances which simply aren't in the prepared samples, and aren't characteristic of ethanol at all. What do you do? Jump to the conclusion that succussion alone removes massive amounts of absorbance from ethanol? Likely? Even there, that ignores the demonstrable fact that ethanol just doesn't have these absorbances - if we can find the Sigma catalogue, and so could Gavinimurthy when he started all this (OK, I see he got it from JJM), then so can she.

What does a true "inorganic materials scientist" do? Well, anything but order up some fresh ethanol in a variety of grades and run some spectra on these for comparison, obviously. I mean, no, you just accept any old rubbish and publish it, no matter how implausible, surely!

I think we done right in not attacking homoeopathy or her stupid theories at all, just demolishing the actual results. It removed a huge potential for come-back. I mean, for all she knew, we could all be raving woos! Frankly, she's toast. And I suspect she knows it.

Another little bonus - several veterinary homoeopaths of my acquaintance subscribe to that journal. They'll recognise my name. They are likely to be chewing the crockery about now.



Rolfe.

[Mojo]

Originally Posted by Rolfe

Another little bonus - several veterinary homoeopaths of my acquaintance subscribe to that journal. They'll recognise my name. They are likely to be chewing the crockery about now.

Let's face it, "what have you had published in homeopathy" is going to be a great comeback to the perennial claims from homoeopaths that their critics know nothing about homoeopathy.

[Rolfe]

You know, I think I spent half the night marvelling at the lameness of the reply. I mean, she has barely addressed anything we said!

She apparently concedes that the "plain ethanol" was contaminated. However.

• No mention of our (rather daring) assertion that one of the traces presented was an accidental duplication of a completely different trace.
• No mention of our assertion that the same traces were presented on one graph as "typical" and on another as representing extreme high or low examples.
• No confirmation (or denial) of what was meant by "envelope of differences", or how the traces presented in that figure were chosen.
• And most of all, not a single syllable about what was arguably the most important point, that the subjective differences apparent between the three succussed preparations could be explained by each having been prepared using a different bottle of stock solvent.

All she seems to have said is, "well, here's my theory again in case you missed it" (no, we didn't comment on your theory, only on your results). Then, "we're not homoeopaths, we're only poor scientists, so all we did was order up what the manufacturer produces". As if that absolves her of all responsibility for addressing possible confounding factors in what the manufacturer might send! Then, "well, we didn't do any statistics at all, or present our data in such a way that it could be statistically analysed, but trust us, there really are differences!"



Looking at how little grasp she seems to have of the criticisms, my suspicion is that she had little to do with either the experimental work or the writing up, and is now simply delegated to say something to provide a fig leaf over the emperor's nudity. I thought these results were really Roy's baby, but maybe he couldn't be trusted to say anything either coherent or polite? I also have a suspicion that Rao now realises they've been comprehensively exploded. I suspect the next tactic on the part of the homoeopathy establishment will be to keep referencing the paper, with never a mention of any debunking two issues later.

By the way, the letter following ours seems also to be a debunking job by a proper scientist on one of the other papers in that ghastly "memory of water" issue. I wonder if it was one of the Bad Science people?

Rolfe.

[Rolfe]

Originally Posted by Mojo

Let's face it, "what have you had published in homeopathy" is going to be a great comeback to the perennial claims from homoeopaths that their critics know nothing about homoeopathy.

Yeah, but do I list this one in my CV? Oh, the embarrassment!

I wish now I'd thought to put JJM's name first - this was really all down to him noticing the wrongness of the "plain ethanol" trace, and we ran with that. But, having written the letter, I automatically signed it.

Rolfe.

[shpalman]

Originally Posted by Rolfe

By the way, the letter following ours seems also to be a debunking job by a proper scientist on one of the other papers in that ghastly "memory of water" issue. I wonder if it was one of the Bad Science people?

You mean Adrian Gaylard? Yes, he is. I also notice that the letter by Philippe Leick (who has previously debunked Walach's ideas regarding WQT in the German magazine Skeptiker) gets no response.

[wilsontown]

Yes, the reply doesn't actually address any of the points raised. At all. I've just published to my blog on it, but I don't really make any different points to those already made here.

Following our letter, Adrian Gaylard has a go at the Vybiral and Voracek (apologies, no attempt to reproduce the diacritics...) and Milgrom papers. Neither really have much of a comeback. Gaylard posts at Bad Science, and has an excellent blog here.

I had a letter in responding to Martin Chaplin's review. At least Chaplin actually addressed some of the points I made, although I still don't agree with him.

I think the crowning glory of the issue, though, is Philippe Leick's demolition of the papers on quantum nonsense by Milgrom and Weingartner, with bonus shout-out to shpalman's blog.

[Rolfe]

Well let's hear it for the Web. Ten or 15 years ago it's very doubtful they'd have got responses at all, never mind like that. The Web drew our attention to the articles, early enough after publication to allow for a response. The Web allowed us to read the stuff, when we would probably not have bothered ordering it up from the library. But not only that, it allowed extensive discussion, with contributions from a variety of people who otherwise would never have met.

JJM, Wilsontown, Pipirr and myself have never met in person, and indeed inhabit three different countries (four if I have my SNP hat on!). Yet the Web has allowed up to collaborate and produce a publication on the subject.

Who says the woos have it all their own way?

Rolfe.

Edited to add: Your own letter comes over very well too, Wilsontown.