Anonymous Criticism - Defending Ernst

Image via WikipediaAs with most bloggers who allow it I have received a variety of comments on my last blog, The self-importance of being Ernst.  Generally it is my policy not to allow anonymous comments.  Since I have the courage to identify myself in the blog I think anyone wanting to comment should do the same.  Maybe that's my self-importance or it is my belief that one should have the courage of ones convictions. 

Someone submitted a few comments in defense of Ernst to my blog and given that the email address is their web site and that their web site, which is just a tirade about various CAM methods, does not allow any commentary moderated or not from their readers I choose not to provide him/her with the links, i.e. advertising.  However, I won't hide behind my ability to block a comment and will paste each of the comments and I'll reply.  If my anonymous critic comes out from behind the curtain I might then publish as-is the words of this Wizard of Oz :)

Criticism 1

I noted in my previous blog posting that it was...unusual for Ernst to have a copy of Long's pre-submitted testimony, which was rejected by the CT Board of Chiropractic Examiners.  My anonymous critic says:

If you read the ‘Methods’ section of Professor Ernst’s original paper he said that “several experts were also contacted for further data”. Perhaps that is how he managed to see Preston Long’s document.

Well that would be reasonable except that, Long doesn't fit any one's definition as an expert.Secondly one needs to keep in mind that Ernst was conducting a "systematic review".  In a systematic review it is not uncommon to seek from other researcher, AKA experts their bibliographies of appropriate literature.  Clearly Long isn't a researcher on the topic (or any other that I can tell).  And one must remember that the data that Ernst was contacting others for was more papers in the scientific literature.  See a larger quote from the Methods:

In addition, our own departmental files and the bibliographies of the articles thus located were searched. Several experts were also contacted for further data.

Criticism 2

Stephen Perle wrote: “What is the common feature of Ernst's citation of these deaths whose information or rather misinformation was obtained from a web site and the deaths reported by Long? Neither of these data sources were from the scientific literature. In his reply to our letter Ernst says "These cases were, however, merely added for completeness and not included in my total number of 26 cases reported in my review." Completeness? So adding people who didn't die after chiropractic spinal manipulation and weren't in the scientific literature adds completeness to a review of the scientific literature?”

Note that Professor Ernst said in his Introduction that:

Quote
“A responsible approach to serious therapeutic risks, however, requires an open discussion of the facts. In this review I aimed to provide the basis for such a discussion by summarising all fatalities which occurred after chiropractic spinal manipulation and were published in the medical literature.”

Note that he didn’t say he would *exclusively* look at the medical literature. Indeed, in the Methods section he was more specific, informing us that

Quote
“Electronic searches were conducted in the following electronic databases: Medline, Embase, AMED, Cochrane Library (September 2009)… *In addition*, our own departmental files and the bibliographies of the articles thus located were searched. Several experts were also contacted for further data. Case reports were included if they provided information on human patients who had died after receiving one or more treatments from a chiropractor.”

He goes on to say that many other fatalities seemed to have remained unpublished and gives examples of the testimony of Preston Long DC (whom it is likely that he contacted personally) which listed the family names of nine victims. Dr Long also stated that ‘many others are unknown hidden behind legal agreements of silence’. Professor Ernst then cites ‘names’ of further North American fatalities from the website www dot whatstheharm dot com

Well this isn't exactly true.  The purpose says: "In this review, I aimed to provide the basis for such a discussion by summarizing all fatalities which occurred after chiropractic spinal manipulation and were published in the medical literature." Well the problem is my anonymous critic is grasping at straws.  The purpose of looking in the departmental files or contacting experts are to find other papers in the medical literature that the search failed to find.

IF Ernst had a purpose that included all possible sources of information then that would have been in his methods.  As I noted earlier if Ernst had put the information from Long in his discussion to coincide with the assertion that many cases are remain unpublished I would have only commented on the apparent collusion with critics of chiropractic not the propriety of including Long's "testimony".

My anonymous critic goes on to note the Ernst cites the names of other American fatalities found on the web site (cited above).  Except that 5 of the 9 deaths that Ernst attributes to chiropractic spinal manipulation (in that paragraph of his paper but not in the total death count) did not die as a result of chiropractic spinal manipulation.  So again if Ernst had only talked about these deaths to buttress his argument that the literature lacks a presentation of all deaths after chiropractic spinal manipulation I would not have criticized his lack of following his own purpose.  But then again would have criticized his zeal to attribute all the deaths found on the web site to chiropractic spinal manipulation. 

Criticism 3

Stephen Perle wrote: “Wenban has detailed the commonality of wrongly ascribing adverse events of manipulation to the care of doctor’s of chiropractic.”

As Professor Ernst says in his response to Whedon et al, his original review contests that with good references. He also says he can not reasonably be expected to know of the ‘personal correspondence’ that Wenban and Bennett cited and asserts that the small discrepancies in numbers (about which they argue) are almost irrelevant.

You know if we were talking about hundreds or thousands or even hundreds of thousands of deaths one or two discrepancies wouldn't matter would they?  But and this is the amazing fact we are talking about a very small number - 26.  So adding one that doesn't belong is big.  I mean with a total of 26 in 115 years we are talking a 3.8% error.  OK the number is a small point - it just goes to the sloppy nature of the review.  Add a couple take away a couple and pretty soon you got nothing of substance in Ernst's "systematic review".
Criticism 4

Stephen Perle wrote: “Ernst says Cassidy's study has been repeatedly criticized for being flawed. The key word is repeatedly. This obviously repeatedly means more than once, which means that there must be more than one reference cited for that criticism. In fact, I think most would agree that repeatedly probably means many more than one. Well there is only one reference cited and what is that: Ernst E. Vascular accidents after chiropractic spinal manipulation: myth or reality? Perfusion 2010; 23: 73–4. No it can't be the only criticism in the scientific literature (not the blogosphere mind you - this is science we are talking about here) that Ernst can find is his own (now you get the idea about the title for the blog).”

Professor Ernst did not qualify where the study had been repeatedly criticised. Further, it you’re going to discount the blogosphere, then your blog post here should not, in any way, be taken seriously by the scientific community.

Well actually Ernst did qualify where the study had been repeatedly criticized.  He provided ONE reference to the journal he is the editor-in-chief of.  Since I know that my anonymous critic isn't a scientist (I know this from their former blog name) I'll forgive his/her ignorance about citation (and purpose above).  However, given the fact that Ernst had no problem citing a web site as a source for the purported deaths, I am sure if his source for "repeated criticism" of Cassidy was the blogoshere or other web sites (even my anonymous critic's) there's no reason to suppose he wouldn't cite them.  The fact of the matter remains that there have been no substantive criticism of Cassidy's methods in the scientific literature.  There has been only one letter to the editor by Maigne. It is my opinion that Cassidy et al adequately responded to Maigne but clearly there isn't "repeated criticism" nor criticism substantive enough to deal a death blow to the validity or importance of Cassidy's findings. 

I think one of the most important points one needs to keep in mind regarding cervical manipulation and stroke is that the only people going ape about this issue are critics of chiropractic.  It just isn't an issue to epidemiologists nor any of the stroke associations.  For example I was at a conference this year in Connecticut put on by the Connecticut Department of Public Health and the American Stroke Association on stroke in the young adult.  Not one speaker mentioned cervical manipulation the whole day.  The critical problem when it comes to stroke in the young adult is that despite its prevalence many health care providers don't recognize the condition because they think stroke is a condition of older people.  Here are the stats I heard for strokes in CT for 2009, keep in mind the population of CT is about 3.5 million

                Age          Number of deaths

45-65             2047

25-44               354

15-24                 34

4-14                   10

<4                        8

Clearly stroke is a horrible event to occur to anyone young or old.  I know first hand from family members how devastating it is but right now the evidence just does not support a causal relation between cervical manipulation and stroke.

Finally I agree completely with my anonymous critic that no one in the scientific community should take my blog or any other blog seriously from a scientific stand point.  That is why I didn't post anything on this blog until my letter to the editor was published.  As I noted in the previous posting the way science progresses is not only the publishing of ones results but that there is an opportunity for critique in the same venue via letters to the editor.  The more scientifically minded should read Ernst's paper, the letters to the editor and his response.  In fact it is this reason that I have not posted my anonymous critics comments as is.  Because I don't want to encourage people to see this person's opinion without the opportunity for rebuttal.  If they posted their name and not their URL I would have not written a whole blog post and just let the comments go and commented on them as I have done herein.

Criticism 5

Stephen Perle wrote: “Now to his critique. I won't reproduce it here…”

Why not? For anyone wishing to read it, there’s a link to the full text of it in the chiropractic section of my website. It really is quite revealing.

Well the simple answer is I don't violate copyrights.  The paper is copyrighted and neither my anonymous critic nor the other web site where I found the paper have anything about permission granted  from the copyright holder to publish the paper on the web.  The journal nor Ernst on his web site have a PDF available for free.  I guess if one is really agitated about the chiropractic profession it is acceptable practice to violate copyrights.  Zealotry trumps legality.

Also. I suggested that it is far better for people to listen to the cross-examination of Dr. Cassidy at the hearing that Long never came to.  Again here is a link to the CT-N video of Cassidy's half day on the witness stand.  Every point Ernst made and then some were tried by the two attorneys who cross-examined Cassidy.  Hear the answers from the horses mouth. 

Criticism 6

I think readers should be told that Dohos and Tragiannidis, of the Aristotle University of Thessaloniki’s Medical School in Greece, wrote to the International Journal of Clinical Practice in support of Ernst’s paper:

Quote:
“…we agree with Professor Ernst on the following points

• Vascular accidents after upper spinal manipulation can cause severe vertebral artery dissections;
• Numerous deaths have been associated with chiropractic neck manipulations;
• Many other cases are unknown behind legal agreements of silence
• Therefore the risks of chiropractic neck manipulations by far outweigh their benefits.”

Ref: Critique of review of deaths after chiropractic, 3. Dokos C, Tragiannidis A.
Int J Clin Pract. 2011 Jan;65(1):103-4.

The problem with  Dohos and Tragiannidis is that they provide no substantive evidence for any of their bulleted opinions above.   

Criticism 7 

I also think readers should be made aware of the following in the US:

Quote
"The National Quality Forum lists 28 ‘never events’ healthcare mistakes that should never happen and need to be reported. Death or serious disability from spinal manipulation is listed as no.16. But chiropractors do not have to report this because they have a loophole. The National Quality Forum demands it of clinics and hospitals but no reporting is mandated for individual doctor's offices where 99% of spinal manipulation is done. Chiropractors generally do not practise in hospitals or clinics."
Ref: Britt Harwe, open letter to the Editor, Focus on Alternative and Complementary Therapies, Volume 15, Issue 1, pages 87-88, March 2010.

OK and the point?  There are a lot of events that occur in private practices for which no reporting is mandated. 

In conclusion I am waiting for the wizard of oz to step out from behind the curtain or not. 

SMP

The self-importance of being Ernst

Image via Wikipedia

Some may know Oscar Wilde's play The Importance of Being Earnest from which I've modeled the title of this blog entry.  The play is a farce where obligations are shirked by assuming false identities.  There is no discussion of false identities in this blog but there is a false presentation of objective science on the part of Dr. Edzard Ernst as Wilde writes in Act 1

 I don’t play accurately – anyone can play accurately- but I play with wonderful expression.

In July the International Journal of Clinical Practice published a paper by Dr. Ernst titled: "Deaths after chiropractic: a review of published cases."  When I read the paper I noticed some obvious errors.  As opposed to some so-called science bloggers, rather than blog about it, I scienced.  (I read once where a scientist said that one should use science as a verb, one sciences.)  
I understand that Web 2.0 allows anyone to be a critic and possibly have great influence on public opinion (gee I am writing a blog about this).  However, one of the fundamental differences between the scientific endeavor and other domains of thought is that knowledge progresses though the publishing of ones findings in a public forum after peer review and then subject to the possibility of letters to the editor which help clarify or critique or add information to that publication.  When one submits a letter to the editor they are peer reviewed, as were the original manuscripts, and if one is using a bibliographic database such as PubMed the letters to the editor are hyperlinked to the record for the original manuscript so that people can find these important commentaries.

Here is a link to our letter to the editor on PubMed's web site.  Copyright issues prevents me from providing the letter to my readers.  We initially wrote a much longer letter but were informed after submission that the journal's maximum size is 500 words for a letter.  I will provide the added detail in this blog.
The paper the letter is in regard to is: Ernst E. Deaths after chiropractic: a review of published cases. Int J Clin Pract. 2010 Jul;64(8):1162-5.  In this paper Ernst's purpose was 

In this review, I aimed to provide the basis for such a discussion by summarising all fatalities which occurred after chiropractic spinal manipulation and were published in the medical literature.

In the results section of the paper he has a paragraph that is, to say the least, very interesting when one thinks of the purpose.   Herein he describes "the testimony of the chiropractor Preston Long for a court in Connecticut recently listed the family names of nine victims."  Ernst cites this with: "Presentation by Dr Preston Long DC to State of Connecticut. Connecticut State Board of Chiropractic Examiners. Hartford Connecticut.
25 October 2009.  There was never a hearing before the State Board of Chiropractic Examiners on October 25, 2009.  The hearing was in January of 2010 and because Long did not appear to under oath attest to his pre-submitted testimony it was, in accord with the procedural rules in CT not accepted and thus not part of the public record.  One wonders how does Ernst get this document when it was not public.  I have a copy as I was on the witness list and these pre-submitted testimony documents I was told were not to be made public until the board hearing.  


Anyone wishing to see the hearing where Long's testimony is discussed should go to the CT-N web site here.
In that same paragraph Ernst cites nine deaths from chiropractic  spinal manipulation that he found on a web site.  When I read Ernst's paper one name stood out in my mind, Kimberly Lee Strohecker.  I know a fair bit about Ms. Strohecker's very unfortunate death.  I wrote about it in a 2004 ethics article titled J'Accuse...! (I Accuse).  This article is about the death of Ms. Strohecker after a chiropractor got her to stop taking her anti-seizure medication.  At least the federal court barred the chiropractor from ever practicing chiropractic again along with prison and a fine.

Yes her death was the result of a chiropractor's actions but they weren't due to chiropractic spinal manipulation.  So I reviewed all the other people Ernst cited as dying from chiropractic spinal manipulation on the web site and discovered the Ms. Strohecker was one of 5 whose death might have been because of a chiropractors actions but weren't the result of chiropractic spinal manipulation.

What is the common feature of Ernst's citation of these deaths whose information or rather misinformation was obtained from a web site and the deaths reported by Long?  Neither of these data sources were from the scientific literature.  In his reply to our letter Ernst says "These cases were, however, merely added for completeness and not included in my total number of 26 cases reported in my review."  Completeness?  So adding people who didn't die after chiropractic spinal manipulation and weren't in the scientific literature adds completeness to a review of the scientific literature?

Now some reading this will say - gee that's picky.  Well the point is ones method must fit ones stated purpose.  Now if he had said in the discussion that there may be many more deaths after chiropractic spinal manipulation that aren't reported in the scientific literature and then cited these he would have at least stood on safer ground.  Still 5/9 cases he said were after chiropractic spinal manipulation weren't.  As we noted in our letter to the editor: "When obvious facts are wrong, the veracity of other facts, not as easily verified, must come into question."  Science works in part because we have to have faith that what people write in their papers is in fact the truth.  It is not truthful to say 9 people died when only 4 did.  Thus can we trust the rest of what Ernst tells us?

Also Ernst cites an event that never occurred, Long's non-testimony before the CT Chiropractic Board.  So again he cites something that is obviously not true.  Thus can we trust the rest of what Ernst tells us?  Some might say it is an honest error which if he was a newbie to scientific publishing I'd say give him a pass but I just did a PubMed search for Ernst E and found 1426 papers.  I don't think he can claim to be a newbie.

Ernst cites a paper by Dziewas et al as presenting another case of a person who died after chiropractic spinal manipulation.  However, one of the great features of modern electronic bibliographic searching technologies, such as  PubMed, which Ernst used, is that letters to the editor also are found with the citation for every paper.  When searching PubMed for Dziewas one finds a letter to the editor by Wenban.  Wenban reported that the treatment was not provided by a chiropractor and thus cannot be chiropractic spinal manipulation.  Wenban has detailed the commonality of wrongly ascribing adverse events of manipulation to the care of doctor’s of chiropractic. 

A couple of the letters to the editor note that Ernst missed in his discussion the study by Cassidy et al that investigated the association between stroke and seeking care of a chiropractic physician or a medical physician and found no excess risk for seeking care with a chiropractic physician.  I've blogged about this previously.   In Ernst's reply he says about Cassidy's study "it has been repeatedly criticised for being seriously, perhaps even fatally, flawed e.g. (8). The ‘inconvenient truth’ might thus turn out to be a 'convenient untruth’."  A long time ago I read an article that talked about somethings one can do to simply determine the validity of a paper one reads.  One of those is references.  Above Ernst says Cassidy's study has been repeatedly criticized for being flawed.  The key word is repeatedly.  This obviously repeatedly means more than once, which means that there must be more than one reference cited for that criticism.  In fact, I think most would agree that repeatedly probably means many more than one.  Well there is only one reference cited and what is that: Ernst E. Vascular accidents after chiropractic spinal manipulation: myth or reality? Perfusion 2010; 23: 73–4.

No it can't be the only criticism in the scientific literature (not the blogosphere mind you - this is science we are talking about here) that Ernst can find is his own (now you get the idea about the title for the blog).  Anyway, you mean to tell me that none of the critics of this study had the testicular fortitude to write a letter to the editor?  Cassidy's study was published in Spine and was part of the WHO's Bone and Joint Decade's Cervical Spine Task Force reports.  Well OK Ernst didn't see fit to write his critique in the journal it was published in, as my colleagues and I did about Ernst's paper.  You know those newbies to scientific publishing don't know that you should submit critiques to the journal it was written in.  Well you remember those 1426 citations to Ernst in PubMed just 204 are letters to the editor.  OK but let's look at his cited critique.  

Ernst E. Vascular accidents after chiropractic spinal manipulation: myth or reality? Perfusion 2010; 23: 73–4.

Don't search too hard for this in PubMed.  It's not there. The journal name is wrong.  I know newbies sometimes get their citations wrong especially when they don't know the journal too well.  Here is the proper citation.

Ernst E. Vascular accidents after chiropractic spinal manipulation: Myth or reality? Verlag Perfusion GmbH. 2010;23:73-4.

What do I know about Verlag Perfusion GmbH.  If you look at their web site you'll see it is a German publication and the first editor listed is...Ernst.  So I guess he's unfamiliar with the proper name for the journal.
Now to his critique.  I won't reproduce it here but instead suggest listening to Cassidy answer these questions as they were posed to him before the CT Chiropractic Board you can see his testimony here.

Ernst in his paper on deaths after chiropractic spinal manipulation also says that there is no benefit to spinal manipulation.  What's his reference for that.  You guessed it, his own narrative review.  Thus he ignores a growing body of systematic reviews showing benefit to spinal manipulation (1- 9)

There is one more item.  In that list of 1425 papers in PubMed written by Ernst E there are 299 systematic reviews of these I count 37 which there is only one author.   Of the 299 systematic reviews 17 are about the chiropractic profession.  Has a real fixation on us doesn't he.  Of the 37 systematic reviews with only one author 10 are about chiropractic.  Or to put is another way of the 17 systematic reviews about the chiropractic profession 10 are single author systematic reviews.  This is astounding.  Why?  Because standard practice in systematic reviews is to have at least two people.  What is considered to be the preeminent international group conducting systematic reviews, The Cochrane Collaboration never does any reviews with one reviewer.  Its just not done by almost anyone but...Ernst.  In fact from the Cochrane Handbook:

It is essential that Cochrane reviews be undertaken by more than one person. This ensures that tasks such as selection of studies for eligibility and data extraction can be performed by at least two people independently, increasing the likelihood that errors are detected. If more than one team expresses an interest in undertaking a review on the same topic, it is likely that a CRG will encourage them to work together.

 Well the self-importance of being Ernst is that he's the only expert whose opinion counts to Ernst.

SMP

1. Oliphant D. Safety of spinal manipulation in the treatment of lumbar disk herniations: a systematic review and risk assessment. J Manipulative Physiol Ther. 2004 Mar-Apr;27(3):197-210. Pubmed record
2. Fernandez-de-Las-Penas C, Alonso-Blanco C, Cuadrado ML, Pareja JA. Spinal manipulative therapy in the management of cervicogenic headache. Headache. 2005 Oct;45(9):1260-3. Pubmed record
3. Chou R, Huffman LH. Nonpharmacologic therapies for acute and chronic low back pain: a review of the evidence for an American Pain Society/American College of Physicians clinical practice guideline. Ann Intern Med. 2007 Oct 2;147(7):492-504. Pubmed record
4. Hawk C, Khorsan R, Lisi AJ, Ferrance RJ, Evans MW. Chiropractic care for nonmusculoskeletal conditions: a systematic review with implications for whole systems research. J Altern Complement Med. 2007 Jun;13(5):491-512. Pubmed record
5. Bronfort G, Haas M, Evans R, Kawchuk G, Dagenais S. Evidence-informed management of chronic low back pain with spinal manipulation and mobilization. Spine J. 2008 Jan-Feb;8(1):213-25. Pubmed record
6. Hurwitz EL, Carragee EJ, van der Velde G, Carroll LJ, Nordin M, Guzman J, et al. Treatment of neck pain: noninvasive interventions: results of the Bone and Joint Decade 2000-2010 Task Force on Neck Pain and Its Associated Disorders. Spine. 2008 Feb 15;33(4 Suppl):S123-52. Pubmed record
7. Bronfort G, Haas M, Evans R, Leininger B, Triano J. Effectiveness of manual therapies: the UK evidence report. Chiropr Osteopat. 2010;18:3. Pubmed record
8. Gross A, Miller J, D'Sylva J, Burnie SJ, Goldsmith CH, Graham N, et al. Manipulation or mobilisation for neck pain: a Cochrane Review. Man Ther. 2010 Aug;15(4):315-33. Pubmed record
9. Miller J, Gross A, D'Sylva J, Burnie SJ, Goldsmith CH, Graham N, et al. Manual therapy and exercise for neck pain: a systematic review. Man Ther. 2010 Aug;15(4):334-54. Pubmed record