Ben Goldacre
Saturday October 22, 2005
The Guardian
Right. Where were we? Oh yes: there is a small unaccredited laboratory in Northants called Chemsol, run by a man with a non-accredited correspondence-course PhD and no formal microbiology training, and he seems to find MRSA in hospitals where other accredited labs, in universities and the like, cannot. And, weirdly, almost every undercover tabloid journalist who gets an “MRSA hospital scandal” scoop involving positive swabs seems to have used this Chemsol operation. I include the Evening Standard, the Mail, the Sun, and of course the Mirror, for their fantastic “Mop Of Death” story.
So what is Dr Malyszewicz PhD finding on his swabs? It is a matter of public record, in Hansard and discussed in parliament no less, that various people have been trying to get samples out of Chemsol for years so that they can run their own tests. None had been forthcoming. Finally he released eight plates: I spoke to the microbiologists who tested them.
On six of the eight where Dr Malyszewicz PhD believed he had found MRSA, the lab found none at all. These plates were subjected to full-on forensic microbiological analyses. The first step was to take a few bacteria off the plate – you’d expect them to still be alive – and grow them on fresh plates. No MRSA. Then they did PCR (polymerase chain reaction) on them, part of the process commonly known as DNA fingerprinting. You use various enzymes to copy and copy and copy the strands of DNA, or “amplify” them, so that there is enough to examine it properly. (PCR, incidentally, was invented by LSD-toting punk genius Prof Kary Mullis: you can read his hilarious Nobel prize acceptance lecture here, and don’t say I never give you anything). So, even if the MRSA was dead, even if there had only been a tiny trace in the first place, PCR would find it. And on six plates, there were (note the plural, pedantic reader) none.
On two of the eight plates sent by Dr Malyszewicz PhD, there was indeed MRSA: it grew, and it showed up with PCR. But it was a very unusual type. Where was it from? This was a strain of MRSA that has never been found on a patient in the UK. In fact it has never before been seen in the UK. Using several DNA fingerprinting and sequencing techniques, they checked the genetic make-up against everything they could think of. They tried the library strains that professional microbiologists use to compare against and practise on. It was none of those. Finally they found it, on the MLST database at Imperial College London, which contains the gene sequence of almost all the MRSA strains ever found. It was an odd strain indeed, only ever found in Australia, and it’s a rare type even there. It is highly unlikely to have been out in the wild in the UK. It may have come from cross-contamination in the Chemsol lab, which of course does analysis work for Australian media toads as well.
And what of the other six plates? What was growing there? Not MRSA. A mixture of things, but mostly just bacilli, another group of common bacteria. Dr Malyszewicz PhD was unable to comment but suggested that “there was MRSA on those plates”. If you ever need to tell the difference between a bacillus and a staphylococcus: one looks like a rod, and the other looks like a ball. You’ll need a microscope that does about 100x magnification. The “Discovery World Great Value Microscope and Human Torso Set” at Argos is only £19.99, and should do the trick. In fact, instead of being one of those whining humanities graduates who goes on about how dreadfully non-generalist the British education system is, I suggest you go out and buy one right now, and blow your mind. It is Saturday, after all. And it’s another world down there.
Send your Bad Science to bad.science@guardian.co.uk
doctorvee / Magical MRSA lab said,
October 22, 2005 at 12:44 pm
[…] The man behind the Mop of Death — more on that lab that finds MRSA where others don’t. […]
WOWblog - The World’s Only Weblog » Blog Archive » Mad scientist said,
October 22, 2005 at 11:33 pm
[…] I hadn’t heard of Kary B. Mullis before, but when someone promises me a “hilarious Nobel prize acceptance lecture” from an “LSD-toting punk genius”, I follow the link. […]
Stephen said,
October 23, 2005 at 1:58 pm
Lets hope they don’t start sending him Bird-Flu samples!
Sam said,
October 24, 2005 at 1:05 am
Oh, you fools, don’t you know you can cure MRSA with homeopathy? See www.abchomeopathy.com/c.php/96 for details. Of course, they insist that they only deal with symptoms rather than the fundamental cause…erm but isn’t that the exact reason most alternative ‘medicine’ practitioners give that proper medicine is so bad?
Just Google MRSA homeopathy if your blood pressure is low and needs elevating. Apparently ‘bacteria are clever’ and deliberately evolve to kill you. The cunning things.
Ray said,
October 24, 2005 at 1:37 am
As I said in the previous thread, there’s a deal of weirdness about on the anti-MRSA circuit.. I was just looking at the website of Pearson and Black, who sell washing/disinfectant kits. They have some very funny sidelines like Acu Flight (magnetic gizmos that allegedly alleviate jetlag) and courses on weird -shit topics).
And then there are all the products involving silver: maybe there’s science in there, but it smacks of magic. The detail in the BBC story, Anti-MRSA silver fabric trialled sounded a bit strange (“The silver ionises when it is in the presence of bacteria and this ionisation process kills the bacteria that are present, on or near the fabric” (maybe like Frodo’s sword, Sting, that glowed in the presence of orcs). And when you go to fuller details – Precious Metals Technology Aids Health Care – the details ring serious alarm bells in places. “Finally, Toray fabrics also assist in the treatment of bed sores, where their function is dual purpose – killing bacteria and promoting the healing process through magnetic force“.
Mongrel said,
October 24, 2005 at 9:06 am
Ray. Silver impregnated dressings have been available for a while now on the NHS for the treatment of bed sores\ulcers because they do have an antibacterial action. I’m not a doctor so don’t ask me “How?” but just because something sounds woo doesn’t means it’s not worth a look (Acticoat is one of the popular brands)
David said,
October 24, 2005 at 9:58 am
Mongrel, while you’re no doubt correct about the use of silver in hospital dressings, it’s the language of the quote that sets alarm-bells ringing with me. It might as well say “The silver has a magic spell placed upon it which is activated when it is in the presence of bacteria and this magic kills the bacteria that are close enough to the ethereal field thereby created.”
If you’re going to describe a chemical or biochemical mechanism then you can’t substitute in simple words or simplify technical phrases just so it scans better. The author should either leave it in as was originally described, as a scientifically accurate description, or leave it out all together. They shouldn’t put it in as a mixture of science- and lay-speak. This is what purveyors of psuedo-science do as a matter of course and why the quote is so suspicious-sounding.
Alex B said,
October 24, 2005 at 12:25 pm
I would love to know how magnetic force could possibly promote “the healing process” (i.e. inflammatory response).
I am no physicist, but yet again the trusted google search gives you a little info on this fantastic magnetic force. unfortunately, and I say this with a trult heavy heart, the Lorentz Force Law does not SEEM to explain how the magnetic force would lead to a higher production of cytokines or t-cell proliferation. the equation is as follows:
Fqv x B
accordingly, the main premise is that “The force is perpendicular to both the velocity v of the charge q and the magnetic field B”
PERHAPS they are trying to tell us that the magnetic force is causing an increase in the velocity attained by our trusted T-cells to get to areas of damage??? Well I will be damned. Next time I am in a rush to get somewhere I will make sure to charge up my transportable magnetic field inducing backpack and activate it as to increase my velocity so I can get everywhere faster!
(And incase you are wondering, yes I do have more important things I should be doing, but I cannot be bothered to do some work on ‘sustaining the variety of life’ at this time of day- far too early).
Ray said,
October 24, 2005 at 12:55 pm
Quite. I’ve no problem with the idea that silver ions in solution might have a bactericidal action (a situation you’d get with a silver-doped dressing in direct contact with body fluids). Maybe that’s what they’re trying to convey, rather than “ionizes”, which has a very specific meaning. However, the promoting the healing process through magnetic force” is prime-cut pseudoscience. Silver isn’t magnetic.
Tessa K said,
October 24, 2005 at 9:51 pm
Apparentl,y magnets work on blood because there is iron in it and improve circulation. Except that blood is not ferro-magnetic.
DaveGodfrey said,
October 24, 2005 at 11:02 pm
‘s a good thing blood isn’t ferro-magnetic, Every time you listened to a walkman you’d runa a risk of having a minor stroke.
Ben Goldacre said,
October 24, 2005 at 11:07 pm
And everyone who went into an MRI scanner would explode.
Ray said,
October 25, 2005 at 12:07 am
And another thing: why does this Magic Cloth always work in an advantageous direction?
“Silver is thermally the most conductive and reflective element, and Toray products therefore assist in temperature regulation. Patients are effectively kept cool in hot climates and body heat loss is restricted in cool climates.”
Why not this? “Patients swelter in hot climates because the silver efficiently conducts external heat to their bodies, and freeze in cool climates as it conducts their body heat away”.
GWO said,
October 25, 2005 at 10:58 am
It’s not? Damn, I’ve been totally lied to by my X-Men comics.
Adrian Gaylard said,
October 25, 2005 at 12:48 pm
Reflectivity is a surafce property, depending on surface finish (as well as composition). To benefit from the reflectivity of silver preventing overhheating in hot climates (or reverse the shiney side to keep the heat in, in cold climates), as claimed, the magic cloth would require a polished (shiney) silver finish (silver foil?). As has been pointed out, these thermally based claims are, to be polite, sheer bunk.
Edward Bozzard said,
October 25, 2005 at 2:10 pm
A group in Canada is claiming that Ginseng does actually have an impact on the common cold – While i’m loathed to believe this, their methods do appear ot be quite sensible. Can anyone reading this suggest any reason why their results wouldn’t be Kosher? www.cmaj.ca/cgi/content/full/173/9/1043
Tessa K said,
October 25, 2005 at 6:53 pm
Edward: it might help to know how many colds the people in the treated group had in the same four month period in the previous year. They may well have had less than the control group during the trial, but it was a small test and some people are less prone to colds than others. That’s all I can think of.
Adrian Gaylard said,
October 26, 2005 at 8:35 pm
This is a small study conducted over a short period of time. The results demonstrate its limitations. If you look at Table 2 in the paper, the entries for “1 cold” have 95% CI’s (for the differences between placebo and treatment) which span zero (i.e. “no difference”.) This is saying that their is a small but significant chance of the treatment group faring worse than the placebo group, not a result to be proud of!
The CI’s for other entries are broad, and their lower limits are close to zero. For example, the outcome for mean number of colds per person (Jackson+cold) is 0.25 with a 95% CI from 0.04 to 0.45. In other words these data indicate that their is a 95% probability the advantage of the treatment, for this outcome, is between 0.04 (very small!) and 0.45 (subatantial). The data do not really allow one to conclude whether the treatment is worth the candle or not.
Add to this that the 95% confidence level allows for a 1 in 20 chance that the value estimated lies outside the very broad CI’s and the statistics do not look very compelling.
Finally the sub-division of the data could be hiding more problems. Why split the data into 1 cold and >2 colds? If it was split into 1,2 and >3 colds what would the result look like? With this sort of choice there is always the chance that the data have been analysed in such a way as to give the most publishable outcome (inadvertantly, I am sure).
A statistician said,
October 27, 2005 at 5:45 pm
Adrian,
I’m not sure the study is ‘small’. Generally statistiicans are employed to make sure the study is powerful enough to detect differences by making the study large enough.
Given the calibre of statistician employed in this trial (Allan Donner is world class) I doubt the statistics are dodgy.
Although the 95% confidence interval (1 in 20 chance) is reported, they also report the actual p-value of 0.017, which is much stronger evidence there is a true difference
– more like a 1 in 60 chance there is no effect of ginseng.
other results for the secondary analyses are even more compelling, suggesting 1 in 1000 or greater chance there is no effect.
“there is a small but significant chance of the treatment group faring worse than the placebo group” I would caution your use of the word significant here. It doesn’t aid explaining statistics to people to use that word in any other context than statistical significance, or clinical significance. Any non-significant confidence interval is by definition going to include the potential for ‘negative’ results
the answer to be question may well be that this is one of the studies that is incorrect by chance. See
medicine.plosjournals.org/perlserv/?request=get-document&doi=10.1371/journal.pmed.0020124
and
bmj.bmjjournals.com/cgi/content/full/322/7280/226
for more on this ‘fascinating’ topic. Of course it may well be that ginseng does in fact work, but as neatly summarised in the companion paper
www.cmaj.ca/cgi/content/full/173/9/1051
” The proposed mechanism of action of ginseng is unclear. The authors provide a summary of the immunologic effects of North American ginseng, but it is not clear how these relate to viral respiratory infection.” And until they tell me how i shall remain sceptical
Adrian Gaylard said,
October 27, 2005 at 6:35 pm
Statistician: 300 odd people is hardly a large study either. I do, however, take your comment on the use of the word significant. I still cannot get excited by 95% CI’s that include zero. I wonder, to what you would ascribe the reason for this?
Alex B said,
October 28, 2005 at 2:09 pm
from someone whose doing an experiment using 2 populations of 5 organisms, 300 people is a huge study 😀
A statistician said,
October 28, 2005 at 3:06 pm
Adrian, The reason you shouldn’t get exicted is that your interpretation is correct!
At the end of the day, all ‘good’ studies should state a priori what their primary outcome is, and not fudge it when they report. Here it was number of colds. This was found to be statistically significant. I’m no clinician, I don’t know whether it is important from that perspective. The main inference should be on that. Secondary outcomes are so called as they are less important. The fact that one of the 3 reported is non-significant doesn’t affect the primary outcome.
As to what is a large trial, well we could argue that all day. Given that they report how the sample size was calculated, and that their actual sample size was greater than this, even after dropout, i am convinced it was large enough. You may not be. Thats life. The really relevant question is ‘how long before the daily mail has a 2 page article touting the ‘miracle’ of ginseng?” (and also flogging it through the advert pages). I give it a week.
Tessa K said,
October 28, 2005 at 5:42 pm
Statistician: >
If they can find a way to get sex into it, I say even sooner than that.
Oli said,
November 5, 2005 at 10:21 pm
Would this be the same Kary Mullis who believes that HIV doesn’t cause AIDS, that the ozone layer isn’t damaged by CFCs, and that global warming isn’t caused by CO2 emissions?
Sounds like good science indeed!
Wonder if he has a shed.
wayscj said,
November 21, 2009 at 7:32 am
ed hardy ed hardy
ed hardy clothing ed hardy clothing
ed hardy shop ed hardy shop
christian audigier christian audigier
ed hardy cheap ed hardy cheap
ed hardy outlet ed hardy outlet
ed hardy sale ed hardy sale
ed hardy store ed hardy store
ed hardy mens ed hardy mens
ed hardy womens ed hardy womens
ed hardy kids ed hardy kids ed hardy kids
MMR and MRSA – health scares and bowel movements | Lack of Environment said,
September 28, 2011 at 7:13 am
[…] Therefore, having lived through both of these health scares, imagine my surprise to read in the final chapters of Ben Goldacre’s Bad Science that both the MMR and the MRSA health scares were entirely manufactured by journalists whose sole aim was to increase their readership! This is the scale of Goldacre’s claim and, as with everything else in his book, he backs it up with great wit and style (oh, and a lot of logic and evidence too). […]
Rianabiology said,
September 8, 2014 at 7:10 am
This article was really interesting because the experiment was conducted by a man with non-accredited correspondence-course PhD and no formal microbiology training and how he found MRSA in hospital. He thought he found them, but the lab didn’t find any. How is this a good science? He is not even qualified to do something like this and I also think it was talking about the experiment really casually and it was not that formal at all. Like, at the end of the article it said “I suggest you go out and buy one right now, and blow your mind. It is Saturday, after all. And it’s another world down there.” For a scientific research you don’t really ask people. You just say what you did and why, make it really detailed. This article was far from that.
mdafne said,
September 8, 2014 at 7:13 am
This goes further to prove that people exaggerate over things which are not even proven to be true. In order to create a scandal the media, in this case, has augmented a situation which if handled correctly like it was later done, would have been basically non-existent. One question that arises in my mind is: why would these samples be sent to this unknown lab run by a man with no microbiology training instead of a well-known and respectable lab? Obviously other factors would have to be looked into for example how much did this man charged and who sent in the samples but nevertheless this goes further to prove that if you need to do something (i.e. sending in a sample to be analysed for a disease) then it is better to be safe than sorry.
lovish21 said,
September 14, 2014 at 11:31 pm
Methicillin-resistant Staphylococcus aureus (MRSA) is a bacterium responsible for several difficult-to-treats infections in humans. There are many people who are tries to imply an assumption to either gain momentum in their life or simply because they believe so without any concrete data. The above example is a clear prove of the need of quantitative data rather than someone believing it so. An aspect of quantitative date is being replicable and the fact that the accredited labs couldn’t discover the results of what the man found shows how researcher bias is common in several experiments. The man may have something against the hospital or may be in involved in the research for too long that it starts influencing the results as he began observing things he wanted to find. This also brings the need or single or double blind in a experiment. Single blind is where the participant is unaware of what the experiment is about and double blind is where both the participant and research (or the person collecting data) is unaware of the purpose of the experiment. Single and double blinds allows validity of the results to remain reliable and replicable.