GENERAL MEDICAL COUNCIL
FITNESS TO PRACTISE PANEL (MISCONDUCT)
Tuesday 7 August 2007
Regents Place, 350 Euston Road, London NW1 3JN
Chairman: Dr Surendra Kumar, MB BS FRCGP
Panel Members: Mrs Sylvia Dean
Ms Wendy Golding
Dr Parimala Moodley
Dr Stephen Webster
Legal Assessor: Mr Nigel Seed QC
CASE OF:
WAKEFIELD, Dr Andrew Jeremy
WALKER-SMITH, Professor John Angus
MURCH, Professor Simon Harry
(DAY SEVENTEEN)
(Transcript of the shorthand notes of T. A. Reed & Co.
Tel No: 01992 465900)
A P P E A R A N C E S
MS SALLY SMITH QC and MR CHRIS MELLOR and MR OWAIN THOMAS of counsel, instructed by Messrs Field Fisher Waterhouse, solicitors, appeared on behalf of the General Medical Council.
MR KIERAN COONAN QC and MR NEIL SHELDON of counsel, instructed by Messrs RadcliffesLeBrasseur, Solicitors, appeared on behalf of Dr Wakefield who was present.
MR STEPHEN MILLER QC and MS ANDREA LINDSAY-STRUGO of counsel, instructed by Messrs Eastwoods, Solicitors, appeared on behalf of Professor Walker-Smith who was present.
MR ADRIAN HOPKINS QC and MR RICHARD PARTRIDGE of counsel, instructed by Messrs Berrymans, Solicitors, appeared on behalf of Professor Murch who was present.
I N D E X
Page No
RICHARD CHARLES HORTON, Recalled
Examined by MS SMITH, cont. 2
Cross-examined by MR COONAN 32
THE CHAIRMAN: Good morning, everyone. I have been given a message from Mr Hopkins that we will need to have a look at our yellow sheets of heads of charge again.
MR HOPKINS: You will recall that on the very first day of hearing the Notices of Inquiry were read out and admissions and various qualifications were then made and the next day we were given the yellow sheet to record those things. We have checked through and there are a few errors on the yellow which we need to correct. I can give you the cross-references for the transcript in doing that. You might wish to have the transcript for the first day to check what I am saying to you. It is page 82 of the yellow which relates to Professor Murch, paragraph 14, the allegations concerning child 9. That is where the allegation start mat paragraph 14. If you turn over to page 83 you will see sub-paragraph (e) which is 14 (e). That sub-paragraph was admitted.
THE LEGAL ASSESSOR: Our copies say admitted and found proved.
MR HOPKINS: Not on the one I have.
THE LEGAL ASSESSOR: There were two versions of this: one came out which did indeed have errors on it, which I had compared to the transcript and I gave it to the secretary. She then produced another one amending the errors that I had spotted. If you have spotted any more, please tell me.
MR HOPKINS: I need to see the version that you have to see whether these have been revised or not. Would you give me a moment whilst I just check. (Short pause)
In that case there is only one paragraph I need draw to your attention, page 89, paragraph 22, sub-paragraph (a), which is the stem. You will see in has in brackets “(with the observation that this is an incomplete account of the circumstances)”. That needs to be deleted so that the stem of 22 (a) is simply admitted and found proved. The other corrections I would have drawn to your attention I see the learned Legal Assessor has already done so.
THE LEGAL ASSESSOR: Paragraph 22, could we look at the transcript.
MR HOPKINS: Day 1, page 93G and 96E. You will see that sub-paragraph (i) below has the qualification “with the observation this is an incomplete account of the circumstances” and that is what was intended, but the stem 22 (a) there was no such qualification intended in the admission. That is it.
THE CHAIRMAN: I have been very helpfully reminded by the Panel Secretary can I ask all the other counsel as well to make sure that they can check through that if there are any inaccuracies or anything that does not meet their approval then before we print the next list if we could have those observations as well.
MR MILLER: We have checked through and we could not find any.
MR COONAN: I do not have Mr Sheldon here. I think he carried out the exercise.
THE CHAIRMAN: We will wait until Mr Sheldon can be here and then you will let us know in due course of time. I do not want another copy to be printed and then find there are other small inaccuracies.
MS SMITH: I will call Dr Horton to resume his evidence.
RICHARD CHARLES HORTON, Recalled
Examined by MS SMITH, continued
THE CHAIRMAN: Good morning, Dr Horton. Are you comfortable to start with your evidence?
A Of course, yes.
MS SMITH: You will remember we ground to a halt in the middle of your editor’s reply yesterday and we now have the last page of it. If you turn in your bundle 2, to page 821, you will remember we looked at the first paragraph and got down to the four parts in the story. If we could turn over to 821 (a) which I hope you now have in your bundle.
A Yes, I have it from yesterday.
Q I am going to run through that in summary and then the Panel can read the rest of that at some later point when they retire. These were your four parts:
“First, the decision to publish. There was no question in my mind that, subject to external peer review and editor debate, we should publish this work. The description of what seems to be a new syndrome and its relation to possible environmental triggers was original and would certainly interest our readers.”
Then going down to the second reason:
“Second, how to publish. As with any provocative report, we always consider the value of running a commissioned commentary in the same issue. In this instance, it was a necessity. Most of observers seem to agree that Robert Chen and Frank DeStefano wrote an important and helpful critique of Wakefield and colleagues work.
Third, how to report these data in the media. We choose not to include this study in our weekly press release. We let the paper and commentary speak for themselves. However, we did assist those who organised a press briefing at the Royal Free Hospital on Thursday February 26 by providing copies of the journal (with the commentary that Wakefield et al did not have) to journalists. Reported adverse comments about the safety of MMR vaccination were made at this press conference. By contrast, the views expressed in the paper are unambiguous and clear.”
You then quote the paragraph we have already looked at:
“We did not prove an association between measles, mumps and rubella vaccine and the syndrome described.
Finally, what has been the outcome? In particular, has harm been done.”
There are three end points: firstly, the press reaction, and you point out that in every report that you had journalists had urged readers to interpret the study cautiously; secondly, the number of children harmed by not receiving measles vaccine, and you indicated you do not know what that figure is but it should be easy to discover with time. You say the question needs to be asked “because its answer will help us all to do better in our reporting next time”. Thirdly, you say:
“Are Wakefield and colleagues’ observations reproducible? Again, we do not know but rather than dismiss what they have reported other investigators must urgently seek to confirm or refute their findings.”
Did that fairly encapsulate your views at that stage, i.e. without knowing what you may have subsequently discovered about the paper but certainly knowing that there had been a very considerable amount of correspondence and some criticism of your decision to publish?
A At the time, that is an accurate summary of my feelings for sure.
Q We looked at some of the correspondence yesterday and, as you know, the Panel had the opportunity to read that through. I want to turn to one particular letter which was in The Lancet in May of 1998 which made an allusion to Dawbarns Solicitors. You can put away weighed this bundle for the moment and get out your bundle 3, page 924. To put it into context, if we look at 923 this was the correspondence pages of The Lancet dealing with this very subject and so the controversy was still continuing at that stage. Again, the Panel will have the opportunity to look at that correspondence between 923 and 295 later.
I want to ask you about one particular letter on 924 from someone called A Rouse from the Department of Public Health Medicine, Wiltshire Health Authority. He said:
“Sir - After reading Andrew Wakefield and colleagues’ article I did a simple internet search and quickly found the Society for the Autistically Handicapped. I downloaded a 48 page fact sheet produced for the Society by Dawbarns, a firm of solicitors in King’s Lynn.
It seems likely then that some of the children investigated by Wakefield et al came to attention because of the activities of this Society and information from parents referred in this way would suffer from recall bias. It is a pity that Wakefield et al do not identify the manner in which the 12 children investigated were referred (e.g. from local GPs, self-referral via parents, or secondary/tertiary or international referral). Furthermore, if some children were referred directly or indirectly because of the activities of the Society for the Autistically Handicapped, Wakefield should have declared his co-operation with that organisation.”
There were letters in reply from the authors. At this stage I want to take you to that from Wakefield. I want to take you to the middle section which begins:
“A Rouse suggests that litigation bias might exist by virtue of information that he has downloaded from the internet from the Society for the Autistically Handicapped. Only one author (AJW) .... ”
That is Dr Wakefield.
“... has agreed to help evaluate a small number of these children on behalf of the Legal Aid Board. These children have all been seen expressly on the basis that they were referred through the normal channels (e.g from GP, child psychiatrist or community paediatrician) on the merits of their symptoms. AJW had never heard of the Society for the Autistically Handicapped and no fact sheet has been provided for them to distribute to interested parties. The only fact sheet that we have produced is for GPs which describes the background and protocol for investigation of children with autism and gastrointestinal symptoms. Finally, all those children referred to us (including the 53 who have been investigated already, and those on the waiting list that extended into 1999) have come through the formal channels described above. No conflict of interest exists.”
When you read that letter, what did you understand Dr Wakefield to mean when he said one author has agreed to help evaluate a small number of these children on behalf of the Legal Aid Board?
A When I read that letter two statements stood out: first, the assertion that you concluded that paragraph with, “no conflict of interest exists”. At the time, in May 1998, I had no reason, no evidence before me, to suggest that that was an untrue statement so I took that statement on trust.
With respect to the sentence that you ask about specifically, has it agreed to help evaluate, I must admit I read that as something that happened after publication. To my knowledge in February 1998 and during the peer review process going back into 1997, I was completely unaware of any potential litigation surrounding the MMR vaccine. I was not aware of the involvement of a firm of solicitors Dawbarns. I certainly was not aware of any activity going on with the Society for the Autistically Handicapped prior to the 1998 paper. I was not aware of any other relationship between Dr Wakefield and Dawbarns and Richard Barr. When I read those statements I saw this as something that was triggered by the paper rather than the paper being in some senses a culmination of events up to February 1998.
Q Looking at the wording of the sentence you referred to “only one author that agreed to evaluate a small number of these children on behalf of the Legal Aid Board”, you say you took that to mean since the publication of the paper and we are now some three or four months on from publication of the paper.
A Yes.
Q Was there anything in particular about that wording which led you to think that?
A It is the has agreed. I know these are fine distinctions. If it had said “had agreed” then I would have thought that was more in the past tense. Reading “has agreed” in combination with the firm assertion that no conflict of interest exists, my suspicions were not raised at that time.
Q Did you accept that letter on its face value?
A We certainly did, yes.
Q I want to ask you about the situation, first in general terms in respect of declarations for conflict of interest. You will be aware that we are now talking about events a considerable period of time ago. I ask you to do the best that you can, and we all know it is difficult, to cast your mind back to your attitude then rather than your attitude now.
First of all, can I remind you and ask the Panel to look at the test for conflicts of interest in 1998 as far as The Lancet was concerned. That, I am afraid, is back in your volume 2.
If you turn to page 615, in fact on my copy there is no date on this, but I hope you will take it from me that in fact the date is 4 October 1997. This is in the article which we looked at briefly before, doctor, “Writing for The Lancet”, and we looked at it in the context of looking at early reports. Can I ask you, is this a particular item which you put in The Lancet relatively regularly to tell everyone who might want to write for The Lancet how to go about it?
A In those days, it was posted on the website and published every month or few months. Nowadays we just have it on the website.
Q If we look at page 616, in the left-hand column, there is a section in the middle which is entitled “Conflict of interest and funding”:
“The conflict of interest test is a simple one. Is there anything – eg, a shareholding in or receipt of a grant or consultancy fees from a pharmaceutical company or a contract from a medical devices manufacturer – that would embarrass you if it were to emerge after publication and you had not declared it? The Editor needs to be informed and will discuss with you whether or not disclosure in the journal is necessary. All sources of funding must be disclosed, as an acknowledgment in the text.”
Was that the 1997 test?
A It is part of the 1997 test. If you go back a page, at the top, in the box “Writing for The Lancet”, you will also see in the second paragraph:
“The Lancet is a signatory journal to the Uniform Requirements for Manuscripts, a document issued by the International Committee of Medical Journal Editors.”
The Lancet is a participating journal in that committee and we meet every year and have done since the late 1970s to discuss guidelines for authors.”
You will see that the next sentence says:
“The 1997 version … of that will be posted on our website … ‘Writing for The Lancet.’”
That 1997 version is published by many medical journals and in fact was published by the Annals of Internal Medicine in January 1997, so was widely available, and that has a section on conflict of interest which very firmly states that authors are responsible for recognising and disclosing any conflicts of interest, including financial conflicts. So the statement you have just read out is supplementary to the statement that we already sign up to and is meant to offer some clarification.
Q Can I ask you, the meetings you have referred to, the International Committee of Medical Journal Editors, is that the Vancouver Group?
A That is exactly right, yes.
Q I shall take you to the actual documentation in relation to that in a moment, but that is a group of which you were a member at the time, is it?
A That is right.
Q As far as the test as it was set out in The Lancet, did you expect authors who were submitting to you to be aware of the rules as to conflict?
A Yes, indeed. That is why we put the section on conflict of interest in what is effectively our instructions for authors, because if you look on the previous page, we say in the third column, “What to do before submission”. So these are kind of the ground rules if you are going to submit anything to the journal. It tells you literally what to put in the envelope, which is what we used at the time. So we would expect authors to have read this for sure.
Q You told us yesterday that in fact Dr Wakefield had submitted a series of papers on related subjects to The Lancet in the past.
A He had published several papers in the journal, indeed.
Q So would you expect him to have made himself aware of the rules relating to this?
A Most certainly.
Q Before we go to what some of the other journals and the Vancouver Group requirements were, I think it is right that you in fact had yourself in 1997 published in The Lancet a paper about conflicts of interest, referring to those documents. I would like to take you to that first of all, please. It is at page 426. Again, I am going to adopt the same tactic, Dr Horton, of reading it out to you. I hope you will forgive me for that; it is rather tedious, but I am not going to read it out in its entirety and I am going to invite the Panel to look at it for themselves later. I think this arose as a result of a particular instance which had come to light, but I am looking at it for its general views. It says:
“The quarrelsome matter of conflict of interest is colourfully brought to light once more in this week’s correspondence pages … The research community has become extremely exercised by this issue in recent years. Now The Lancet is charged with neglecting to inform readers of a financial conflict of interest supplied to us by Scott Tenenbaum and colleagues at Tulane University.”
So that was unrelated to the matters with which we are concerned, but triggered you to put your views down again on paper. Is that correct?
A The whole subject of conflict of interest in the 1990s and going into the 2000s, so to speak, has been the subject of great concern to the research community, so we were all writing about this, yes.
Q You go on:
“The weight of opinion about such conflicts is that they should be disclosed to editors and readers alike. For instance, the majority view of the International Committee of Medical Journal Editors is that ‘Published articles and letters should include a description of all financial support and any conflict of interest that, in the editors’ judgment, readers should know about.’ Why? Because ‘Bias can often be identified and eliminated by careful attention to the scientific methods and conclusions of the work. Financial relationships and their effects are less easily detected.”
Then you say:
“The New England Journal of Medicine, for example, has argued that ‘most academic institutions and journals have not gone far enough in dealing with this problem’; the BMJ agrees and now publishes a conflict of interest statement at the end of every paper (often declared as ‘none’); and other journal editors have supported a strict policy of full disclosure. However, the case in favour of full disclosure rests, it seems to me, on three large fallacies.
First, there is the fallacy of objectivity, the notion that scientific writing can be free form the common prejudices found in other literature or journalism – or that if prejudice does exist it can be easily neutralised by a statement of disclosure. Yet interpretations of scientific data will always be refracted through the experiences and biases of the authors. Scientific writing can never escape from being a rhetorical exercise. Advocates of disclosure may argue that some of these potentially malign influences could be limited by focusing on the most serious conflicts. This brings us to the second fallacy – that it is financial conflicts of interest which ‘cause the most concern’. Financial conflicts may be the easiest to identify but they may not be the most influential. Academic, personal, and political rivalries and beliefs are less easily recognised, but each may affect an interpretation. Such biases render the declaration ‘conflict of interest: none’ an impossibility. To put financial conflicts to the fore is to provide a smokescreen for more covert and possibly more influential commitments.
The third fallacy is that disclosure can heal the wound inflicted by a financial conflict. By way of an example, this week’s letters on the paper by Tenenbaum et al all carry conflict of interest statements. But how is the reader to interpret these disclosures? Most of the authors have some conflict to unveil. Some of those reporting a correlation between the results of an assay for antipolymer antibody and symptoms among the recipients of breast implants have served as witnesses for plaintiff attorneys in breast implant litigation …
According to prevailing rules concerning disclosure … executive editor at the New England Journal of Medicine, also has an interest to reveal.”
Then you point out that she has written a book.
“Do these disclosures help us to understand the data reported by Tenenbaum et al? I doubt it. Their paper is criticised on largely methodological grounds and the authors respond in detail. Statements of supposed conflict of interest may divert the reader’s attention away from the scientific argument by seeming to focus on pro or con positions concerning [the] manufacturers. Disclosure may not only oversimplify a complex issue but also in itself mislead, sometimes damagingly so.
The difficulties of applying a policy of full disclosure are illustrated by the 1996 New England Journal … ”
You then quote an example from there. Going down to the bottom of the page, you say:
“The only way to minimise among interpretations is to allow maximum dialogue from all parties, irrespective of their interests. The Lancet also prefers a pluralistic solution to one based on censorship.”
You say at the bottom:
“Here, it seems to me, our obsession with conflicts of interest may harm free discussion in science.
Perhaps part of the difficulty with conflict of interest lies in the phrase itself, which has disparaging connotations. The Annals of Internal Medicine uses the term ‘dual commitment’ and asks authors to disclosure these to editors. The Lancet’s policy is much the same.”
Then, if I may, Dr Horton, that paper which you wrote was informed by a number of different papers to which you refer and I just want to take you to four of those, which is what other journals were saying on the same subject really from a significantly earlier period, starting in 1993, which was the New England Journal of Medicine. You will have to go to another volume, I am afraid. It is bundle FTP7, page 205. This is an article in the New England Journal, an editorial, in August 1993, so some five years before the time we are considering. Again, I am going to invite the Panel, sir, to look at this themselves, but I will read you just the basic position:
“Financial conflicts of interest in medicine – what they are and how to deal with them – constitute one of the most contentious issues in our profession. Organised medicine and its critics have debated whether clinicians should gain financially from their practice in any way other than through their direct fees or salaries for service, and the academic community and its critics have debated whether medical researchers should gain financially from companies whose products they are evaluating. Opinions vary greatly – from those advocating that all conflicts of interest be proscribed to those arguing for disclosure and caveat emptor to those maintaining that there is no problem. There are even disagreements about the definition of a conflict of interest. Many attach qualifiers, such as ‘potential’ or ‘apparent’, in an attempt, it often seems, to be delicate or nonjudgmental.
In this issue of the Journal, [a particular author] provides a thoughtful and compelling analysis of the controversy. He defines a financial conflict of interest as a condition, not behaviour; clinicians or researchers who might benefit financially by distorting their work have a conflict of interest regardless of whether they actually distort their work. We agree with this definition. The circumstances determine whether there is a conflict of interest, not the outcome.”
Then going on to the next page, page 206, the Journal indicates that they have a more stringent standard for review in relation to conflict of interest and they say in the middle of the page:
“Many arguments have been raised against our policies. The most frequent are that disclosing conflicts of interest impugns the integrity of honest researchers, that prohibiting conflict of interest in authors of review article and editorials is a form of censorship, and that our policies unfairly ignore intellectual biases in favour of financial ones. Let’s look at these arguments.
Disclosing a conflict of interest does not impugn the honesty of authors … On the contrary, if problems were apparent during the peer-review process and not corrected in revision, the paper would not be published. Furthermore, the issue is not primarily one of honesty but of unconscious bias, which may be quite subtle and difficult to detect. This is one reason for disclosure; bias not appreciated in the peer review process may still be detected by readers.”
They then go on to deal with the question of censorship and the last sentence is:
“Our decision to exclude authors with financial conflict of interest is simply another criterion, designed to ensure that the articles are of maximal credibility and interest to our readers.
Our conflict of interest policy is directed toward financial arrangements because, as Thompson points out, they are both widespread and optional, as well as seductive. Many intellectual conflicts of interest – for example, the desire for positive or important results in a research study – are not optional. They are inherent to the endeavour. Fortunately, these conflicts are not hidden and they are well appreciated, since nearly all researchers share them; there is therefore no need for disclosure. Other conflicts are unusual or idiosyncratic (for example, a Seventh-Day Adventist doing research on the health effects of the sect’s lifestyle), and there is no practical way of anticipating all of them or dealing with them with a blanket policy. Like Thompson, we believe that the fact that a policy cannot cover all contingencies is hardly a reason not to have one.
Financial conflict of interest are a matter of choice. Researchers do not have to have them, nor do restrictions based not hem constitute a violation of rights. In many areas of life, such restrictions are the norm. For example, judges are expected to excuse themselves from cases involving companies in which they have an interest, and no one with such an interest would be selected for a jury, no matter how much he protested his objectivity. Why should those in the biomedical research community consider it an affront to have to submit to similar restrictions? Furthermore, when researchers choose to invest in health-related companies rather than in other types of business, it raises the question of whether they are attempting to profit from the specialised knowledge they gain in the course of performing research. In other settings, this would constitute insider trading. Can this be the face we want to present to the public?
Given the current climate of clinical investigations, financial conflict of interest will probably continue to be a fact of academic life, but they must be responsibly regulated, with disclosure being a minimal requirement. Most academic institutions and journals have not gone far enough in dealing with this problem.”
Dr Horton, I am going to run through, if I may, the chronological papers you have referred to and then ask you to comment. Just in passing, the New England Journal of Medicine, can you tell us what the status of that journal is as a general medical journal?
A If you were sitting in North America, it would be seen as the world’s leading general medical journal.
Q You would acknowledge that it came fairly high up the list. Would that be fair?
A Absolutely. These things are measured in quantitative terms and it is the number one impact factor general medical journal, yes.
Q If we go on to the BMJ, British Medical Journal at page 209, this is an editorial by Richard Smith, the editor of that journal and is in 1994, a year later. Going to the second paragraph:
“Recognising the growing concern, the International Committee of Medical Journal Editors (the Vancouver group) last year produced a statement on conflict of interest. Dennis Thompson...”
He was the gentleman from Harvard:
“...recently defined just what is meant by conflict of interest ‘a set of conditions in which professional judgment concerning primary interest (such as patients’ welfare or the validity of research) tends to be unduly influenced by a secondary interest (such as financial gain). He emphasises that conflict of interest is a condition not a behaviour. We should pay attention to conflict of interest not only when it is clear that a judgment has been influenced by conflict of interest but simply when it might have been. Bias works subtly – which is why the double blind randomised controlled trial is such a crucial development – and most of us have lifted insight into our own motives, let alone the motives of others. Suggesting that somebody has a conflict of interest is thus far removed from accusing them of dishonest behaviour. But conflict can have important effects: several studies have shown that doctors are more likely to refer patients for tests, operations, or hospital admission when they will benefit financially than when they will not.”
Going down to the bottom of the page:
“Editors need to deal with conflict of interest in order to make sure that the quality of research, judgments, and information in their journals is not reduced by secondary interests. They must also pay attention to the issue in order to play their part in maintaining public confidence in the judgments of professionals.”
Going over on to the next page:
“The common remedy for conflict of interest is disclosure. We plan as soon as possible to include the source of funding for a research study in all scientific papers, and we want authors and referees to let us know of any conflicts of interest they may have. We will send them a document explaining the nature of any conflict. Sometimes we may decide our readers should know about a conflict of interest and we will then publish a note on the conflict – after consultation with the authors or reviewers. To disclose a conflict of interest about a piece of work does not mean that the work is worthless (otherwise there would be no point in publishing it); but readers will want to consider that information along with many other factors in making their own judgment on the work.
The BMJ has for several years subscribed to the uniform requirements of the Vancouver group that ask authors to let us know about conflicts of interest, but people rarely do so. Now we are moving the policy along by always recording the source of funding for research by asking people to sign a document, and sometimes disclosing conflicts. Perhaps we will eventually have to do more.
The editors of the New England Journal of Medicine have said that ‘most academic institutions and journals have not gone far enough in dealing with this problem’ and that is still truer on this side of the Atlantic.”
That was in 1994. Then if we can go on to your own publications at page 232. Is this another editorial by you?
A No, we rotate the writing of editorials between the editors, but, certainly, I take responsibility for everything that comes out under The Lancet’s name.
Q This is an editorial in 1996, so we will be coming closer to the relevant period, about eighteen months before publication. This was inspired by a particular incident in the New England Journal involving two experts who had written an editorial which, we now know, had some connection with the drug company involved in research. If we go down to the last paragraph:
“Readers deserve to know the past and present commercial associations of authors relevant to the research in question. The recent events of the New England Journal of Medicine show how difficult a policy of full disclosure ban can be in practice. The Lancet’s simple test, by now means perfect, is this...”
And you repeat the test which I took you through originally:
“Would a non disclosed commercial interest, should it be revealed later, prove embarrassing to an author? We rely on the conscience and judgment conscious of the author to draw our attention to such a personal conflict. In the end, this must be the person with whom the responsibility lies.”
Lastly, the Vancouver group to which you referred. We have the 1997 requirements at page 197. Those are the ones you referred to your own editorial. These are the uniform requirements for manuscripts submitted via medical journals and we can see the membership of that organisation in the left hand corner and your own name there as the editor of The Lancet, sparing your blushes, because this group involved most of the senior editors of the senior medical journals, general medical journals. If we can turn to Conflict of Interest on page 202:
“Conflict of interest for a given manuscript exists when a participant in the peer review and publication process – author, reviewer, or editor – has ties to activities that could inappropriately influence his or her judgment, whether or not judgment is in fact affected. Financial relationships with industry for example through employment, consultancies, stock ownership, honoraria, expert testimony either directly or through immediate family are usually considered to be the most important conflicts of interest. However, conflicts can occur for other reasons, such as personal relationships, academic competition, and intellectual passion.
Public trust in the peer review process and the credibility of published articles depend in part on how well conflict of interest is handled during writing peer review, and editorial decision making. Bias can often be identified and eliminated by careful attention to the scientific methods and conclusions of the work. Financial relationships and their effects are less easily detected than other conflicts of interest. Participants in peer review and publication should disclose their conflicting interests, and the information should be made available so that others can judge their effects for themselves. Because readers may be less able to detect bias in review articles and editorials than in reports of original research, some journals do not accept reviews and editorials from authors with a conflict of interest.”
We are concerned with authors and there is then a short section on them before it goes on to reviewers and editorials. Authors:
“When they submit a manuscript, whether an article or a letter, authors are responsible for recognising and disclosing financial and other conflicts of interest that might bias their work. They should acknowledge in the manuscript all financial support for the work and other financial or personal connections to the work.”
If we turn back to page 199 of the same document which relates to Acknowledgments, we just see a reference again:
“At an appropriate place in the article (the title page footnote or an appendix to the text; see the journal’s requirements) one or more statements should specify...”
The third and fourth are:
“Acknowledgements of financial and material support, which should specify the nature of the support; and (4) relationships that may pose conflicts of interest.”
That is something of a canter through those, Dr Horton, but those are the documents you referred to when you did your own editorial. Do they fairly sum up the position, in your view, in relation to the general requirements at that time?
A Yes, they do. Perhaps I could just add a little footnote of context. Some authors had written very passionately that these kinds of guidelines were, in the words of one particular, very respected epidemiologist in Boston, Ken Rothman, that these requirements were the new McCarthyism in science, that this kind of naming and shaming, as he was arguing, was damaging trust in the research process. So editors were engaged at that time, through the early/mid 1990s in a pretty ferocious debate with the research community, of which some might say we were a part, and this debate focused on the very nature of what conflicts of interest were and how potentially damaging they were in understanding research when it was published.
We took the view as editors, as you have underlined in its most extreme format from Jerry Kassirer at the New England Journal of Medicine, we took the view that disclosure, although not a panacea, was a necessary requirement; disclosure to the editors. There is then a subsequent issue about whether that conflict of interest should be published. What we were asking for at that time was disclosure to editors so that there could then be a discussion with the authors about whether that potential conflict of interest should be disclosed or not, so we were inviting a discussion to take place.
Q If an author had genuine doubts as to whether some thing was a matter which might be interpreted by readers as a conflict and he was uncertain as to whether that was a legitimate concern or not, what would you have expected him to have done about it?
A We would have expected him or her to tell us about the potential conflict, to have argued robustly that this in fact, in his or her view, might not have been a conflict and therefore should not have been disclosed, but that we would arrive at a common position together, having discussed it – not just between one author and ourselves but a collection of authors. All authors on the by line of a paper share a responsibility for the contents of that paper so there should be a discussion amongst the authors and between the editors and those authors about whether a conflict of interest might have a material impact on the interpretation of their work.
Q I think it is plain from the documents I have taken you to, but, generally speaking, the sort of issues which raise the possibility of conflicts of interest which should be declared in the medical community, are of what kind?
A Of what kind, sorry?
Q Are they generally associated with a connection with the pharmaceutical industry or with the medical devices producers, that sort of issue?
A If we are talking about financial conflicts of interest, then sure we are talking about sources of funding. But, there are these other conflicts as well that might come about in relation to one’s association with a particular institution, with a particular side of an argument. In a litigation case they are much more nebulous and difficult to pin down because you cannot measure a physical amount of money like you can with a funding source, but they are no less important. We were asking for all those potential conflicts of interest to be disclosed to editors and lead to some sort of discussion.
Q Although you differentiate between them, might there be situations where it may be a combination of both; in other words, a direct financial interest and, as you say, a more nebulous interest?
A Very much so and, going back into the 1990s, I can think of examples where papers were rejected because of conflicts of interest, both financial and non financial.
Q If you can put your mind back to 1998, had you at that time ever come across, as you can recall, the Legal Aid Board as a source of funding for medical research?
A Personally, I had not, no.
Q When you were considering the paper, with which we are concerned, for publication, did you have any reason to suppose that there was any funding from an outside source in relation to the research that was being done?
A No, not at all. In the paper as published there was no source of funding disclosed and, as far as my recollection is concerned in the discussions about that paper, there was no source of funding. I think there is a reference in the paper to the Trustees of the Royal Free Hospital, but I do not remember any external source of funding that came to light being disclosed.
Q There is indeed a reference to that source. Would that source have struck you as in way a source that you should go behind or question any further?
A No, it is very common for hospitals and institutions to have funding mechanisms for research that is done within those institutions; special funds that have been created to support preliminary work, work of young investigators who have not been able to secure funding from more traditional sources, such as the MRC or Wellcome Trust.
Q I think it is right, taking you on in the story for a considerable period of years to 2004, that a series of allegations were made, is that correct, in relation to various matters in The Lancet paper?
A That is right.
Q Can you tell us by whom they were made?
A By a freelance journalist, Mr Brian Deer.
THE CHAIRMAN: What was the name, did you say?
A Brian Deer.
MS SMITH: Did he attend a meeting with you in the presence of an MP?
A Yes, he first called to tell me – I think we had a couple of conversations – to tell me that he had some information that was highly relevant to the February 1998 Lancet paper by Dr Wakefield and colleagues and would I be interested in talking with him about that. I said “Yes” I would. In February he came to this meeting, which was very strange because it is not usual we have a Member of Parliament turn up to the Lancet’s offices to join in a meeting of that sort but that is what happened, Evan Harris turned up to take part in that meeting.
Q Were allegations – I will deal with them all because you set them out very clearly in The Lancet – did they include allegations in relation to funding issues?
A Yes, they did.
Q Was that the first you heard of there being an issue?
A That is right. That was the first time that I was made aware of the connection, both with the Legal Aid Board and the specific funding of the work that was reported in The Lancet.
Q How did you handle it, Dr Horton, obviously you listened to what they had to say. What did you do thereafter?
A Well the presentation by Brian Deer took the form of him standing up before a group of editors and laying out a series of allegations, not just relating to the Legal Aid Board funding of the work but also including the way the work had been handled by the ethics committee at the Royal Free Hospital – two specific allegations, one, that the work had not actually received ethics committee approval and, second, the approval that was given for a piece of work was in some sense a fabrication that the work that took place and was reported in The Lancet was done under cover of another ethics committee approval process for an entirely different piece of work which was an extra ordinary serious allegation.
The second set of allegations concerned the nature of referral of the families to the Royal Free. It was described in the paper as a consecutive series of patients and Brian Deer argued that, indeed, it was not a consecutive series, it never could have been a consecutive series, so that that claim in the paper was materially incorrect. The third part was in relation to the Legal Aid Board funding.
These were three sets of allegations which went to the heart of the credibility of the paper and were clear allegations of research misconduct. We ended up having a disagreement about the terms of our discussion. He had come to The Lancet specifically to get an on the record quote from me for an article that he was writing for the Sunday Times. My understanding was that he had hoped for his article to have been published the previous weekend, but his editors were slightly less confident about the basis of his claims than he himself was, and so were seeking to have an on the record quote, one or more quotes, from others to was substantiate the concerns he was laying out.
On the other hand, having presented those kinds of allegations as an editor of a research journal, as somebody who is a co-signatory on the ICMG guidelines, a participant in the Committee on publication ethics, a whole bunch of organisations which we are members of and somewhat bound by, these allegations of research and misconduct triggered a formal process to try and find out whether there was indeed misconduct, specifically whether there had been a fabrication or falsification of the research as he was outlining.
Q As far as you were concerned, when you heard these serious allegations being made been investigative journalist, what was your main concern?
A My main concern was to take those allegations and go straight to the institution where the work had been conducted and get answers to them.
Q How did it progress from then on? Did you have a meeting with the co-authors first or go to the institution first?
A There were several stages, and this is what led to a disagreement between Brian Deer and myself over this matter. He had come to get a quote for a newspaper article. I said “No, I cannot give you a quote for a newspaper article because what you told me now triggers a set of responses for me which are really quite formal. The first thing I need to do is go and talk to the authors who you are particularly pinpointing as at fault.” The three authors were Dr Wakefield, Professor Walker-Smith and Dr Murch. He was unhappy about that, to say the least, and what began as a perfectly amicable presentation and discussion turned rather tense and conflictual and difficult as it became clear I was not giving to give an on-the-record quote.
He had been unable, at that point, to get quotes from the three protagonists for his newspaper article. My goal was, what I asked him to do, was to photocopy the relevant evidence that he had so that I could take that to these three individuals and put the allegations that he had to them. The articles were photocopied by my assistant as I recall.
Q When you say articles were photocopied, you mean the documents that Brian Deer presented to you?
A Yes, he had various documents relating the Ethics Committee approval, I think newsletters from Dawbarns Solicitors showing the relationship of the litigation to the research, and the Legal Aid Board related documents. We contacted Dr Wakefield, Professor Walker-Smith and Dr Murch and asked the three of them to come to The Lancet’s offices so that we could discuss these allegations. We were also joined in that discussion by another author of the 1998 paper, Dr Peter Harvey, a paediatric neurologist. The meeting took place, the allegations were put, we had a discussion about those, there were some agreement, there was some disagreement. The next step was to take those allegations to the Royal Free Hospital itself. What one is trying to do here is to invite the institution to conduct its own investigation and to give us a report on whether it thinks those allegations are true or false and for us to also try and assess the evidence, as best we can, and then to make a public report of that investigation.
Q When you saw the three doctors, first of all before you went to the Royal Free --
A The four doctors, Dr Harvey as well.
Q Did you receive any information from Dr Wakefield orally at that meeting in
relation to the Legal Aid Board funding issue?
A Yes. He was perfectly open about that. He said that he had received Legal Aid Board funding for a study, that some of the children - he could not exactly recall the number, I think four or five - who had taken part in the Legal Aid Board funded study were also in The Lancet 1998 report. All three of them were very firm in denying the allegations over Ethics Committee approval, the allegations that there was no Ethics Committee approval or even a fabrication of Ethics Committee approval, and all three denied the allegation that there was something wrong with the referral of the children to the hospital as reported in The Lancet.
Q You have told us you followed that up with a meeting at the Royal Free.
A The vice-Dean at the Royal Free at the time was Professor Humphrey Hodgson. I contacted him, went up to the Free and met with him to discuss the allegations and then an investigation took place.
Q I am going to take you on to your eventual findings as you published them in The Lancet in a moment but I want to ask you, first of all, Dr Wakefield told you that there was an overlap between the children who were part of the Legal Aid Board study and the children who were written up in the paper, is this correct?
A That is correct.
Q I think you said four or five?
A That is correct. He said he had to contact Richard Barr to clarify the exact number of children in the overlap.
Q What was your view as to whether or not that was a matter which should have been disclosed when the article was submitted for publication?
A I was very clear to Dr Wakefield in that meeting that that was something that should certainly have been disclosed to us, not only the overlap but also the connection with the Legal Aid Board and the specific funding for this work. His position was that he did not consider it a conflict of interest and indeed the funding that he had received was not for the specific work that was published in The Lancet but for a separate piece of work that I think, at the time in 2004, had not yet been published.
Q With reference to the description that you previously given of dual commitment, can you explain what it was that concerned you about the information that Dr Wakefield gave you that some of the children were indeed part of both The Lancet study and the Legal Aid Board study.
A When you look at the newsletters that Brian Deer had provided to us, I was very concerned about what was written in those newsletters. I do not know if you have them in any of the evidence here to look at. There were two newsletters as I recall. In one Dr Wakefield was specifically mentioned. His views were - I think the words quoted in the newsletter were - he had deeply depressing views about the possibility of a link between the vaccine and cases of autism. This was a newsletter I think with a date of 1996 on it, two years before publication in The Lancet. Here was a written public record connection between the lead author of a supposedly independent paper published in a medical journal and a solicitor taking part in litigation on behalf of families.
This is litigation that, as I understood it, or the legal aid board funding, was supposed to be an independent investigation. No judgment was supposed to have been made about whether the vaccine was linked to these conditions. If you read the newsletters it looked to me very much as if a judgment had been made. The newsletters talked about achieving success in the litigation. There was a comparison with the pertussis vaccine scare of some years before. I think in the first newspaper the wording was something like “We are in a better position now than we were with respect to the pertussis litigation.” In a second newsletter the research was described “as aiming to progress the cases” I believe was the wording. Problems with the vaccine were cited and they even went so far, in the second newsletter, to talk about preliminary indications of the research which made a strong link between the vaccine and autism. This was all back in 1996/1997, some time before The Lancet paper. In my position as an editor, I was clear I should have been told about those connections so that we could have evaluated that information when we discussed the paper, not only ourselves by our peer reviewers.
What was even more concerning, in some ways, was when these allegations were laid out to the authors in that Lancet meeting, Dr Murch and Professor Walker-Smith made clear that they were unaware of the connection between Dr Wakefield and the Legal Aid Board funding. I can recall clearly to this day quite literally Dr Murch’s jaw dropping with horror when these facts were laid out, facts that were uncontested at the time by Dr Wakefield. We felt very misled, to be frank.
Q You have dealt, in some detail, with the association of Dr Wakefield with the litigation. You know, in fact, and there is no secret about it, there was money given by the Legal Aid Board for connected research at least.
A There was. The amount was disputed I think, but there were monies of some nature.
Q There may be a dispute as to the precise nature of the connection but what I wanted to ask you, and I think it is clear from what you have said, was the actual funding, the actual fact of money having been transferred, the extent of your concerns or were your concerns wider in relation to the association?
A Wider. The money was a fact. The fact that Dr Wakefield had been tied up, had formed a link, with the Legal Aid Board at all, whether money had been exchanged or not, was a materially relevant fact that should have been disclosed to the editors. When one looked at the newsletters going back to 1996, to see two years of what appeared to me quite deep involvement with one firm of solicitors in this litigation process certainly was a matter that should have been disclosed to the authors out of the courtesy let alone to the editors.
Q Had it been disclosed to you, had you known, how would you have handled it? Would it simply have been a matter of discussion or might it have affected your view of the paper generally? How would you have handled it?
A It is very difficult going back to 1997/1998 to try and guess how one would have handled it. I have thought about this long and hard. The first answer would be to say it would have been a material piece of information to influence our decision about whether to publish the work or not. That is not quite enough. When I think back to other papers that we considered, and either pursued or rejected on the basis of conflict of interest disclosures, I am sure that in a discussion with my colleagues as editors that information would have led to rejection of the paper.
Can I give you an example of a case where we did reject just to show that this would have been perfectly consistent with what we were doing at the time. It is interesting that one of the papers supplied to us yesterday mentioned calcium channel blockers. Back in the
mid-1990s there was a debate about a particular class of drugs to treat heart disease. There was a great deal of concern that these drugs had been inadequately investigated and may be causing quite serious adverse affects causing heart attacks in patients. We published several of those papers.
We received a paper from a drug company, sponsored by a drug company, with a group of authors, some of whom were tied to that drug company, about one particular calcium-channel blocker which showed that there was no adverse risk to that particular drug. That was very reassuring good news and certainly would have been an important piece of information to add to the debate. Since The Lancet had already published papers very critical of this particular class of drugs in some sense, we would have had a duty to publish the paper giving an opposite point of view but, given the nature of the funding, given the source of the data, given the connection of some of authors with the pharmaceutical company, we took the view that that was a conflict of interest too far and that paper should be rejected. That took place in 1996. I actually travelled to Sweden to meet the authors at a Swedish research centre linked to the pharmaceutical company to discuss their work. I know that we were making decisions at that time against papers based on conflict of interest.
If I try and take the same set of principles that I know were operating and apply them to this particular case, I am sure that would have been the decision at the time, especially given the fact that the interpretation was so tenuous. We were talking about families who reported from memory a link, a temporal link, between vaccination and the onset of symptoms. Given how tenuous that connection was, the nature of a two-year process of being involved in litigation, I think we would have thrown our hands up and said this is not something that we can say is reliable.
Q As you said, it is terribly difficult to put your mind back. Can I ask you just for the sake of information to the Panel to go the other way? Say you had come to the conclusion that you were going to publish it. It was an interesting paper and you were going to publish, but you had been given this information as to Dr Wakefield’s involvement with the litigation. Would you have considered publishing without making some reference to that so that the readership knew about it?
A No, that certainly would had to have been published in some way, shape or form at the end of the paper.
Q You have told us that you considered this matter with the Vice-Dean, Professor Hodgson, from whom in fact we have heard, at the Royal Free. I think you reached conclusions generally about the other allegations as well as this allegation, the Legal Aid Board matter. You published, in fact, your conclusions and the authors’ response to them accompanied by a statement from the Royal Free.
If we look in volume 3, at page 1211, you set out there that on February 18, 2004 serious allegations of research and misconduct had been made. You set out the allegations: firstly, that contrary to a statement in The Lancet paper ethics approval for the investigations conducted on the children reported in the children, some highly invasive, had not been given. Secondly, the study reported in The Lancet was completed under the cover of ethics approval for an entirely different study of 25 children with a new paediatric syndrome: enteritis and disintegrative disorder following measles/rubella vaccination. Thirdly, contrary to the statement in The Lancet that the children were “consecutively referred to the Department of Paediatric Gastroenterology”, children were invited to participate in the study by Dr Wakefield and Professor Walker-Smith thus biasing the selection of children in favour of families reporting an association. Fourthly, that the children who were reported in The Lancet article study were also part of a Legal Aid Board funded pilot project led by Dr Wakefield, a pilot project with the aim of investigating the grounds for pursuing a multi-party legal action on behalf of parents of allegedly vaccine damaged children the existence of which was not disclosed to the editors of The Lancet. Fifthly, that the results eventually reported in The Lancet paper were passed to lawyers and used to justify the multi-party legal action prior to publication, again a fact not disclosed. Sixthly, that Dr Wakefield received £55,000 from the Legal Aid Board to conduct the pilot project and, since there was a substantial overlap of children in both the Legal Aid Board funded pilot project and the Lancet paper, this was a financial conflict of interest which should have been disclosed to the editors and was not.
You then say the editors of The Lancet have seen and reviewed the documentary evidence available in support of the allegations and you have acted in accordance with the guidelines as set out on publication ethics. You say:
“We have presented this evidence to the senior authors of the paper in order to seek their responses .... Dr Horton has shared this information with Professor Hodgson, vice-Dean at the Royal Free ... the institution in which the original work took place.”
Then you said with this notice are accompanying statements from Dr Murch, Professor Walker-Smith and Dr Wakefield answering the allegations of research and publication misconduct together with a statement from the Royal Free. You say:
“Given these four statements, together with an evaluation of the available documents, we consider that ... ”
You then set out your responses to the four allegations. In respect of the first one, you conclude that the evidence does not support the allegation. Mr Miller is asking me to read out the allegations. I am very happy to do so, sir, but they are long statements by the doctors and I was going to invite you to go out and read those to yourselves, although if the defence want me to read it all out, which they tell me they do, I will do so.
MR MILLER: Sir, rather than summarising it – we are racing through some quite important material; we have heard a lot of evidence from Dr Horton which has been very helpful about the process and this is the end of that process in which he is telling the public at large, or the medical public anyway, what the outcome is. I think just for Ms Smith to summarise it is unfair. I would prefer if she could take a little time to deal with his response to the allegations which have been made.
MS SMITH: Sir, I certainly was not attempting to race through it and I was indeed trying to ensure that you had enough to see where those parts which obviously you will be reading in their totality were. If Mr Miller really wants me to read out all the allegations, I am happy to do so and Dr Horton will have to bear with me.
THE CHAIRMAN: I think Mr Miller has made that suggestion and I think it is perfectly right that you should do it. Secondly, I am looking at the time. It is five to 11. I wonder whether you would like to do this after the break, or would you prefer to do it before the break?
MS SMITH: It will take me a long time, sir, so I think it is probably best if we have a break and then I will embark upon it.
THE CHAIRMAN: Indeed. I am also thinking of Dr Horton, who is giving the evidence. We will now adjourn and we will resume at 11.20. Dr Horton, you are still under oath, still in the middle of giving evidence, so please do not discuss the case with anyone.
(The Panel adjourned for a short time)
THE CHAIRMAN: Ms Smith?
MS SMITH: Dr Horton, do you still have page 1211 in bundle 3 open in front of you?
A I do.
Q I had read out the allegations to you and we are now going to the findings under each allegation, which are in the right-hand column:
“Allegation 1
The evidence we have seen indicates that ethics committees approval was given for data collection from clinically indicated investigations in the children with an initially undiagnosed illness and who were described in the 1998 Lancet paper. This illness was at first believed to be enteritis combined with a disintegrative disorder. Subsequent detailed clinical investigations eventually showed this condition to be the syndrome finally reported in The Lancet. This course of events was not described in full in the Lancet paper, although the similarity of the behavioural changes with those of a disintegrative psychosis (Heller’s disease) were commented on in the discussion section of the 1998 Lancet paper. In summary, the evidence does not support this allegation.
Allegation 2
As described under Allegation 1, detailed clinically appropriate investigations led to a re-evaluation of the initial diagnosis of these children, as set out in protocol 172-96. The evidence we have seen indicates that here was no attempt by investigators to conduct the study of children reported in The Lancet in 1998 under cover of an entirely different investigation. In sum, the evidence does not support this allegation.
Allegation 3
Professor Walker-Smith notes that although the referral pattern was unusual – direct contact by patients with Dr Wakefield leading to referral to the Royal Free – the children were indeed consecutively referred. He reports that to the best of his recollection he did not invite any children to participate in the study. Thus, as far as the facts can be ascertained by a review of the case notes and from memory, children reported in the 1998 Lancet paper were consecutively referred to the Royal Free and were not deliberately sought by the authors for inclusion in their study based on parents’ belief about an association between their child’s illness and the MMR vaccine.”
Before I go on to allegations 4 to 6, which relate to the Legal Aid Board involvement, Dr Horton, can I ask you, in relation to those first allegations and your observations, can you tell us the extent of your investigations and what you relied on in coming to those conclusions?
A Yes. I mentioned before that my first port of call at the Royal Free was to discuss these allegations with Professor Hodgson. Professor Hodgson then, as I recall, invited Dr Murch, Professor Walker-Smith and Dr Wakefield into his office – certainly the first two I can recall – we discussed those allegations and I asked Professor Hodgson and the certainly two, possibly three, doctors present to try and ascertain whether those allegations were true or not and the only way that that could be done was by going back to the case notes in the case of referral and by looking at the detailed documentation relating to ethics committee approval. In fact, many of the documents relating to ethics committee approval seemed already to be in the possession of Brian Deer and he has passed those to me, so we seemed to have a common set of documents between the Royal Free and myself.
In relation to the ethics committee, we simply went through those documents together and Professor Hodgson went back and looked at further documents to see if there were further documents that I was not in possession of. His conclusion, looking into those, was that the procedures in place at the Royal Free were such that the work had received ethics committee approval, although the statement in The Lancet was rather compressed, in that the statement did not describe what were obviously quite extensive and detailed discussions about the protocol between the investigators and the ethics committee at the Royal Free, but the allegation both that there was no approval and that what approval had been given for a particular study was actually a fake study, those allegations were not correct.
In relation to the nature of the referral of patients, the steps taken were that Dr Murch and Professor Walker-Smith certainly I can recall went back to their department at the Free and took the case notes under the supervision of Professor Hodgson to look to see what the nature of the referral was. In particular, the allegation was that the children reported in the Lancet paper had been somehow cherry-picked, was the phrase which was being used; they had been selected deliberately to illustrate an association between the MMR vaccine and the syndrome described. So the effort was to identify what was the source of the referral of these patients. There were two parts to this.
The first part was: where did the patients come from, so to speak? The answer to that was that they were referred both from general practitioners and other consultant paediatricians. Remember that Professor Walker-Smith’s unit was an internationally recognised and respected centre for paediatric gastroenterology, so it would be quite usual for him to receive referrals from other consultant paediatricians. Also – and this is what was explained which was new and which I think we described in the statement – there had been direct contact between parents and Dr Wakefield, which was an unusual pattern of referral. The parents may have connected with Dr Wakefield, but then the route of referral was not through Dr Wakefield; it was via the GP or the consultant paediatrician. So there had been some cross talk between Dr Wakefield and the parents, although the route of entry to the Free was by conventional channels.
In addition, Professor Walker-Smith went back and looked at the biopsy book at the Royal Free to see how the samples from children had been obtained. I think these were from ileo colonoscopies. Indeed, the report back from him was that they had been consecutively sampled. My interpretation, to cut a long story short, was that the letter of the law, so to speak, the use of the words “consecutive referral” was correct, but behind that was clearly a much more complex process than was described in the paper. Was that research misconduct? No, because there was no fabrication or falsification. There was merely compression. Should it have been expanded upon more? With hindsight, probably yes.
Q Dr Horton, that is very helpful. You have told us that you had this meeting at the Royal Free. How long was this meeting and this investigation? How long did it take you?
A Let me get my dates right. February 18 I think was a Wednesday in 2004. We met with Dr Wakefield, Dr Murch, Professor Walker-Smith and Dr Harvey that afternoon and, to be honest, I cannot remember whether I went up to the Royal Free on Wednesday afternoon or Thursday morning, but the process went on during the course of Thursday before we were satisfied that we had answered as best we could the allegations.
Q If we look very briefly first of all at allegation 1. You say:
“The evidence we have seen indicates that ethics committees approval was given for data collection from clinically indicated investigations in the children with an initially undiagnosed illness and who were described in the 1998 Lancet paper. This illness was at first believed to be enteritis combined with a disintegrative disorder. Subsequent detailed clinical investigations … ”
As far as that is concerned, you tell us that you had the documentation from the ethics committee which indeed we have. Is that correct?
A That is right.
Q As far as whether the tests were clinically indicated or not, was that something you could form a view about?
A What I was trying to do was to understand the nature of the acquisition of the data as reported in the Lancet paper. Do you want me to describe that?
Q I want to know really what informed that observation. You have said it was for data collection and you have told us you saw the papers from the ethics committee and you say “the clinically indicated observations”. I just wondered on what you based that observation.
A Testimony from Dr Wakefield, Professor Walker-Smith and Dr Murch, the documents relating to the to and fro discussions between the ethics committee and the authors, I think there were also some letters from the Dean of the Medical School at the time and also Professor Hodgson’s own statement that we subsequently published, confirming that the Royal Free was perfectly happy with the nature of the ethics approval process for the study. So those three components together led to us taking the view that as far as the ethics committee allegations were concerned, they were unproven.
Q As far as the allegations in relation to the referral pattern are concerned, you have explained those clearly to us. I think you said Professor Walker-Smith looked back at the biopsy book.
A He looked back at the biopsy book; as I recall, Dr Murch and Professor Walker-Smith together looked at the case notes. I will be honest and say I cannot remember whether Dr Wakefield did or did not look at the case notes himself. I apologise for not being able to remember.
Q When you say:
“Thus, as far as the facts can be ascertained by a review of the case notes and from memory, children reported in the 1998 Lancet paper were consecutively referred to the Royal Free … “
That is a review of the case notes and the memory of the authors, is it?
A Exactly right.
Q If we can go on to allegations 4 to 6:
“Dr Wakefield had two roles in this work. First, he was the lead investigator of a Royal Free study into the nature of a new syndrome with bowel and psychiatric symptoms. Second, he was commissioned through a lawyer to undertake virological investigations as part of a study funded by the Legal Aid Board. At the time of submission and eventual publication of his 1998 Lancet paper, this second study had not been disclosed to the editors of The Lancet and his co-authors. We judge that it should have been so disclosed, irrespective of the number of children overlapping between the pilot project funded by the Legal Aid Board and the Lancet paper. Such a disclosure would have provided important information to editors and peer reviewers about the context in which this work was taking place – a context that would have been vital in making a final decision about publication. We believe that our conflict of interest guidelines at the time should have triggered such a disclosure, including the fact that a significant minority of the children described in the Lancet paper were also part of the Legal Aid Board funded pilot project. These guidelines stated that … ”
Then you give the conflict of interest test which we have already looked at.
“The difficulty of adopting a dual role as a clinical investigator and as a participant in the evaluation on behalf of the Legal Aid Board is revealed in Dr Wakefield’s response to Allegation 5. Although it may be correct that ‘this [Lancet] publication . . . added nothing further to the issue of causation than that that was already well known to the lawyers, the perception of a potential conflict of interest remains. Editors and reviewers should have had an opportunity to take his dual role into consideration when assessing this paper for publication.
Finally, although the Legal Aid Board funding referred to a different aspect of Dr Wakefield’s work from that reported in The Lancet, the perception of a conflict of interest nevertheless remains. This funding source should, we judge, have been disclosed to the editors of the journal.”
We have looked at your views on this at some length, Dr Horton, but what I do want to ask you is this. Where you refer to the Legal Aid Board funding referring to a different aspect of Dr Wakefield’s work from that reported in The Lancet, where did that information come from?
A That was reported to us by Dr Wakefield.
Q Then you give a summary:
“The first three allegations of alleged research misconduct have been answered by clarifications provided by the senior authors of this work. The wording in the published paper regarding Ethical Practice Committee approval and patient referral was accurate, yet at the same time summarised obviously lengthy and complex institutional and clinical review and referral procedures. In the light of the public controversy surrounding this work and the allegations made to us, one could argue that more explanation could and should have been provided in the original paper. Although, with hindsight, this seems a reasonable criticism, all research papers published by all journals are inevitably concise accounts of often complicated research protocols. We do not judge that there was any intention to conceal information or deceive editors, reviewers, or readers about the ethical justification for this work and the nature of patient referral. We are pleased to have had the opportunity to clarify the scientific record over the matters raised by these serious allegations.
We regret that aspects of funding for parallel and related work and the existence of ongoing litigation that had been known during clinical evaluation of the children reported in the 1998 Lancet paper were not disclosed to editors.
W also regret that the overlap between children in The Lancet paper and in the Legal Aid Board funded pilot project was not revealed to us. We judge that all this information would have been material to our decision making about the paper’s suitability, credibility, and validity for publication.
In considering what sanctions The Lancet should apply, the COPE guidelines ... ”
that the committee on publications fix.
“... gives eight options in a ranked order of severity. Given the public health importance of MMR vaccination, together with the public interest in this issue, we have decided to pursue a course of full disclosure and transparency concerning these allegations, the authors’ responses, the institution’s judgment, and our evaluation.”
Q Before I go on to the statements by the doctors, which were published in The Lancet and by the Royal Free, can I ask you, have there been other occasions when you have had to investigate allegations made about a research paper and its propriety in general terms?
A Frequently.
Q Is it customary to discuss and take the word of those against whom the allegations are made?
A It is.
Q Can you elaborate on the reasons for that?
A This process of how you investigate research and misconduct has become highly vexed over the past decade. At the time, back in the late 1990s, some of us, in particular Richard Smith at the BMJ, myself and the editor of a journal called Gut, had gathered together to create something called the Committee on Publication Ethics. We had done so because we were very concerned that these allegations of research misconduct were coming up, sometimes proven, sometimes not, but when they did come and were proven, they would be splashed across newspapers and the media and cause a lot of damage to public trust in the research process. We had gone to various organisations, such as the General Medical Council and others, to say, “Can we please create a much more formal investigative process for allegations of research misconduct, akin to the Office for Research Integrity in the United States?” The reaction of the research community in the UK to those requests from editors was less than enthusiastic. The reason for that was that they felt by creating a formal process of investigation for allegations of research misconduct, would in some sense legitimise the issue, would be, to put it, as I recall one person saying, a hammer to crack a nut, that this was a very small problem in research. To create an institutional structure to investigate a small problem would be inappropriate and would draw undue attention to that small problem.
Our view was, small it may be – although in truth we did not know how big or small the problem was, because one could not know, so that was a supposition that was invalid – but, however big the size of the problem was, when things went wrong, the response was always so damaging that we did need a process for investigation; there was not one. In that situation, the only reliance that we could put upon was individual institutions investigating allegations of misconduct at their own institution if you did not have a national process. We had several meetings at the General Medical Council around the corner from here when Sir Donald Irvin was President, with him. He led a group on research misconduct at the GMC to discuss how this should be dealt with. It included the President of the Royal College of Physicians, the Royal College of General Practitioners, representatives of the Royal Society. The ultimate decision was that the responsibility for investigations of allegations of research misconduct should rest with institutions.
That also presents a problem, because that means that every institution has to have its own mechanism for such investigation. In this particular case, we went to the Vice-Dean of the Royal Free, laid out the nature of the problem and asked him to investigate and come back to us, as best he could, with his own judgment of the veracity or not of the allegations. In addition to that, we would look at the documentation as best we could and try and form our view as to whether those allegations were true.
This is an imperfect process, but it is the best that we can do when you do not have a national institution to investigate this. At the Office for Research Integrity, for example, in the United States, the first port of call is the institution to investigate but, if the institution is unable to investigate, there is a federal process for investigating that allegation of misconduct. That is the gap in the process that we do not have. Until recently, only last year, a university wide process has been attempted to be established.
Q I am going to turn on to look at the statements you elicited from the doctors. Sir, so that I do not have everybody jumping up, I am going to be selecting what I read to you simply because these are such long documents that I think, with respect, you will probably find it very difficult to listen to them being read allowed without reading them to yourselves. I am going to try not to be too selective but, as I say, the selection is in the interests of getting it over. I am then going to invite the Committee to go out before Dr Horton is cross examined and read these to themselves. I hope that is satisfactory.
THE CHAIRMAN: I fully understand that suggestion. Mr Coonan, are you happy with that suggestion that has been made?
MR COONAN: I understand the proposal. It depends on the degree of selectivity. I remark in passing, that in order or to advance the prosecution case, if I may say so, Ms Smith was scrupulous in taking the witness through to very large chunks of documents and one hopes with the same balanced approach here. It may be, if there are gaps, that it may be cured by you and your colleagues having a very close scrutiny of these documents before Dr Horton is cross examined.
THE CHAIRMAN: I am sure that we would be actually looking at these documents, as Ms Smith has suggested, before Dr Horton is released, but can I also make a suggestion that – I think Miss Smith has said – if you do find at any stage there is a particular part of a passage that you would wish to draw our attention to, I am sure you would be able to do it.
MR COONAN: I am sure I speak for all three members of the Bar on this side of the table.
THE CHAIRMAN: Thank you, I think that is very helpful. Ms Smith?
MS SMITH: In the light of that, I shall certainly select on the side of reading rather than not, to ensure I am not accused of skipping parts that the defence want. The first is a statement by Dr Simon Murch he says:
“These allegations concerning our 1998 study are extremely serious, and clearly require immediate clarification. I welcome the opportunity to do so. My comment relates to the alleged lack of ethical practice committee approval. I refute the allegation absolutely on the basis of extensive documentary evidence.
The protocol for the 1998 Lancet paper was submitted in September 1996 to what was then the Ethical Practices Sub Committee.”
He gives the title and says it was signed by Dr Wakefield as lead investigator with the named consultants and signed collaborators.
“The application was initiated due to findings of colonoscopy of two children with behavioural disorders which would now be classified within the autistic spectrum, and a history of chronic gastrointestinal symptoms and recognition of a broadly similar clinical history among other referred patients. Specifically, for several years previously we had looked after an autistic child with severe ulcerative colitis who eventually required colectomy (not included in the study) and the second child, colonoscoped (on September 1996) had ileitis of sufficient extent that a diagnosis of probably Crohn’s disease was made. Following this diagnosis, the child had been treated in good faith by our [IBD] into an ongoing (ethically approved) study of polymeric enteral nutrition. He had already made remarkable symptomatic improvement, including apparent cognitive advance. We appeared to be dealing with a condition of significant severity, and had seen clinical improvement. News of this improvement was rapidly disseminated among parents of autistic children.”
He believed that led to many further referrals:
“The child was included in the study, with additional investigations performed after ethics approval was obtained.
The title of this submitted application is a point of contention, and should be clarified. Having taken initial advice from our psychiatric colleagues on the basis of referral letters, it was considered that these children demonstrated a form of autism called disintegrative disorder (Heller’s disease). After full psychiatric assessment of each child seen, it was later concluded that the more accurate description for the submitted paper should be pervasive developmental disorder. The working title had remained disintegrative disorder, while some parents referred to their child as autistic and others did not. The whole area of nomenclature in autistic spectrum disorders was notably at that stage. As we saw more patients, we moved towards a more inclusive label of autism which was used in subsequent correspondence after February 1998 to the ethics committee. Measles and rubella were singled out in the application since these conditions, but not mumps, had been linked to autism and not isolated reports.
Application (172 96) was for permission for in depth analysis of 25 patients referred either by GPs or the vitamin B12 unit. The selection criteria was explicit were the presence of disintegrative disorder, symptoms and signs suggestive of gastrointestinal disease and parental request for investigation. All patients reported met these criteria. The consultant paediatricians responsible for the children’s care decided on the investigations, although was taken from colleagues at other centres.”
He says:
“We determined that these investigations were required clinically, not only to characterise gut inflammation but also to exclude primary neurological diseases. We had in particular taken advice for the neurological investigations, since some of the referrals appeared to have suffered an encephalitic illness, and specifically the inclusion of lumbar puncture was suggested as important for assay of cerebrospinal fluid lactate, to exclude mitochondrial cytopathies that can cause both neurological regression and bowel disease. Several of these cases had not been investigated to exclude a primary cause for their regression. We thought it important to ensure that we were not missing an underlying abnormality.”
He sets out the proposed investigations and the tests that were to be carried out:
“The protocol was referred back at first submission in November 1996 with clarifications and amendments were suggested, and was approved in November 1996. The protocol formed the basis for all children investigated in the 1998 Lancet paper and all were investigated. We had no idea at the time of our Ethical Practices Committee application that lymphoid hyperplasia would prove so common, although it was a prominent part of final report.
It is important to document where the protocol differed from the submission. First, neither nor my fellow endoscopist, Mike Thomson, eventually considered it justified to perform upper gastrointestinal endoscopy in most patients there was no published evidence of upper gastrointestinal pathology, and we were performing these procedures under sedation, as was then our practice. Getting the precise level of sedation is not easy in children which such behavioural difficulties and we felt this was not appropriate at that time, although our policy altered in later years. Second, in the event, we did not continue with this extended protocol for the full 25 patients, again because of the clinical concerns of myself and my colleagues, since we had found no evidence of underlying metabolic abnormality in any case and did not consider lumbar puncture of further cases was indicated. Other children subsequently seen were thus not subjected to this extended protocol, and investigated by testing of inflammatory markers and abdominal X ray, with endoscopies performed if thought clinically indicated, unless there were clear clinical reasons to perform additional tests.
Following the publication of the initial report, professor Walker Smith sought guidance from the Ethical Practices Committee about further investigation of future cases.”
He quotes the letter we have seen asking for the extension of approval:
“In a letter to the ethics committee, further studies were referred to under the title ‘autism and non specific colitis and Lymphoid Nodular Hyperplasia’ since that was the clinical entity that the earlier study had defined.”
He points out that that was reviewed and data collection clinically indicated investigations was approved on the basis of gastrointestinal symptoms and initial assessment and in no way relevant to the 1998 Lancet paper which had been conducted entirely according to the 1996 approval:
“Thus, there was no change in the name of the ethical approval requested for the 1998 paper, as mistakenly alleged.”
He points out there was:
“A local review initiated by the Royal Free in July 1998 which confirmed that the application had been fully considered by the ethics committee, and that assurance had been given that the investigations were clinically indicated. It was apparent that the continuing investigation of those children had been reviewed by the ethics committee in July 1998 and appreciated that investigation of children seen after publication had become less extensive, and usually restricted to gastroenterological testing as thought clinically appropriate.
We contended then, and still contend now, that these were standard and appropriate gastroenterological and the current state of knowledge at that time. Undoubtedly we now perform endoscopy less frequently, but that is based on extensive experience. Similarly, a child with coeliac disease in the 1970s would have had three diagnostic biopsies compared to the one, or even none, now performed.
Thus, I can confirm that the patients presented in The Lancet study were investigated in accordance with the ethics committee approval of December, 1966, and that no attempt was made to seek retrospective approval.”
That was Professor Murch. It was followed by a shorter one by Professor Walker Smith which deals, in particular, with the referral:
“I deny that the allegation that there was systematic bias in the pattern of referral for the children in the 1998 Lancet paper. No kill were invited to participate in this study.
Upon review of the Centre for Paediatric Gastroenterology, Royal Free Hospital, work book ... we confirm that the children who were reported in The Lancet paper of 1998 were the first 12 children consecutively referred to the university department of paediatric gastroenterology with autism and related disorders, who had gastrointestinal symptoms requiring ileo colonoscopy to exclude chronic bowel inflammation. These children were referred to me at the university department of paediatric gastroenterology from July 1996 to February 1997 one from XXX and one from XXX. By the time the paper was accepted for publication, as mentioned in an appendix [to the paper] to January 1998, a further 40 children had been investigated, 39 with the syndrome reported. The children were all investigated by clinical means to determine whether bowel inflammation was present that could then be appropriately treated.
These children were referred to the Royal Free by their general practitioner (ten cases) or consultant paediatrician (two cases). Some parents had heard of Dr Wakefield’s previous work on inflammatory bowel disease and specifically requested referral, but the channel of referral was always as described above. However, the pattern of referral was often that the parents of the child approached Dr Wakefield directly knowing of his work, frequently by telephone. In the case of one patient, in whom it has been alleged that I contacted a consultant in order for a referral to be made, he had been asked by the parents of this child to contact me to explain what investigations were available at the Royal Free for children with autism and bowel problems….
To the best of my recollection, I did not invite any children to participate in our study.
None of the children at the time of referral was known by the team of paediatric gastroenterologists who cared for and investigated these children to be involved in a pilot project commissioned by the Legal Aid Board. At the time of consultation, I was aware that some parents were engaged in legal proceedings. Review of the clinical notes of the 12 children in the 1998 Lancet paper indicate that we became aware at the time of publication that one child was involved in litigation proceedings against the vaccine manufacturers.”
That is Professor Walker-Smith. Dr Wakefield, Allegation 4, and you may want to remind yourself what that was, that was the first relating to the Legal Aid Board.
“Allegation 4 completely misrepresents the facts. These were two quite distinct issues: the first a clinical report of the 12 cases and the second, a hypothesis-testing laboratory study to examine for the presence or absence of the measles virus in autistic children when compared with appropriate controls.
A minority of the children described in the 1998 Lancet report were part of the second study that was funded in part by the Legal Services Commission. The relationship of these two distinct studies to the legal status of the relevant children is set out below. The Professor Walker-Smith has already described the basis for the referral of the children according to clinical need.
At the time that the children reported in the 1998 paper were referred to Professor Walker-Smith for investigation the time material to their sequential investigation and subsequent inclusion in the report, none of the 12 reported children was in fact legally aided ....
Whether parents perceived an association with MMR vaccine or not, whether parents had approached lawyers with the intent to seek legal redress or whether children were in receipt of legal aid funding or not had no bearing whatsoever on their selection for clinical investigation or inclusion in The Lancet paper. Since these allegations were made I have returned to the parents to determine the facts. At the time the children underwent colonoscopy (i.e. the time at which their pathology as reported in The Lancet in 1998, was defected and reported ...) one child had a legal aid certificate. The authors had no knowledge of this fact until now.
In support of this and in view of these allegations, parents of children in the Lancet report have provided written signed statements that (i) they contacted me for help given their child’s gastrointestinal symptoms, (ii) their referral to the Department of Paediatric Gastroenterology at the Royal Free was through their child’s doctor, (iii) that at no time did I encourage them to seek legal redress through the courts in the MMR class action and that their child formed part of the initial study of 12 in The Lancet ...
Independently, I was commissioned through a solicitor, Richard Barr, to undertake quite separate virological studies on ten children. This is entirely in line with other university-based studies that have been similarly funded by the Legal Services Commission and reported, for example, in the BMJ. The list of children provided for me by Richard Barr was based on his knowledge of an overlap between parents referred to the Royal Free and those whose parents had made contact with Richard Barr. I could not have constructed such a list since I had no knowledge of the litigation cohort or the legal of status of children within it. I was specifically concerned with addressing the scientific question in relation to measles virus - a perfectly legitimate question in view of the nature of the intestinal disease and the sequence of events in the children. Measles virus infection of the intestine was a specific interest of mine.
Once again it is important to emphasise that I had no specific knowledge of the legal status of the ten children on the list other than that as described above. Investigations, in light of the current allegations, indicate that four of these children (exact number to be confirmed by Richard Barr) were among those reported in the 1998 Lancet paper. The virological studies on these children have been submitted for publication.”
He said if and when they were published acknowledgement would be given to the Legal Services Commission.
“Allegation 5 is an inaccurate misrepresentation of the facts.”
That is the allegation that suggests that the results reported were used to justify multi-party action.
“The results eventually reported in the 1998 Lancet paper were in the public domain long before their publication in February 1998 having been presented at several national and international scientific meetings. They were readily available for interested parties to scrutinise and interpret. The findings were not actively made available to the media until after publication but other than this there was no attempt to conceal them.
Such was the level of concern about the clinical and scientific findings in this group of children that I and Professor Walker-Smith reported them to a meeting in October 1997, convened by the Hon Tessa Jowell MP, then Minister of Health, attended by the Chief Medical Officer.. and in the presence of Richard Barr of Dawbarns solicitors and representatives of interested parent groups. Barr was there as a lawyer responsibly concerned by the sheer numbers of parents reporting to him of developmental regression and gastrointestinal symptoms in their children following MMR.
It is important to emphasise that the only aspects of the 1998 Lancet paper that could have been used to justify a multi-party action as in the foregoing accusation is the parents’ perception of a temporal relationship between MMR vaccine and the onset of symptoms. This perception is well known to the lawyers long before we were even aware of the role of the lawyers or the proposed multi-party action, certainly before our publication of our article the Lancet. This publication added nothing further to the issue of causation than that which was already well known to the lawyers. The accusation is therefore specious. My own report to the Legal Services Commission was served in 1999.
With respect to allegation 6...”
This is the one in relation to the actual funding of the money to conduct the pilot project.
“... these were two separate matters. One, a report of the clinical investigations and the other a study commissioned quite independently through Richard Barr. The latter study was designed in order to explore the issue of possible causation. These studies were concerned with viral detection in the diseased intestinal tissue of ten potentially affected children and the approach is entirely in line with other university-based studies that have been funded by the Legal Services Commission and reported in the BMJ. Funds received from the Legal Aid Board were paid into, and properly administered, a research account with the special trustees.
I have stated above that the origin of the list of children was provided to me by Richard Barr. My involvement was limited to the legitimate concern: was measles virus present in the intestinal tissues of these children?
As outlined above, I can confirm that the publication of the relevant virological studies is still awaited. An interim submission of a report of this study (rejected) contained an explicit acknowledgment of the Legal Aid funding; this will be made available as necessary.
If and when the relevant virological studies are finally published, due acknowledgement will be made.
For none of these or any subsequent children has legal status influenced the need for investigation or interpretation of the findings. Where it is known that the children are in receipt of a Legal Aid certificate or where studies receive funding from the LSC, this will be included in any relevant publication.
The clinical and pathological findings in these children stand as reported. They have now been confirmed independently by a reputable physicians and pathologists. On the basis of the molecular detection of measles virus in diseased intestine of these children this issue too merits further study.
I regret the difficulties that this issue has caused my colleagues over the last week and I am grateful to them for their advice and support. I am enormously grateful for the timely manner in which Richard Horton has dealt with this issue and for his clarification of the issues surrounding perception and reality where conflict of interest may be concerned.
My colleagues and I have acted at all times in the best medical interests of these children and will continue to do so.”
Those were the three statements that you published in accordance with your way of managing these matters. It is accompanied by a statement by the Royal Free, which we have read before. This is Professor Hodgson’s, but I will remind you of it:
“We are entirely satisfied that the investigations performed on the children had been subjected to appropriate and rigorous ethical scrutiny. Because the nature of the conditions affecting child behaviour and gastrointestinal symptoms was unknown and require elucidation, the investigation was properly submitted and fully discussed with the Ethical Practices Committee.”
It is set out who that Committee was, and was independent of the University Hospital.
“The Committee, after clarifying a number of issues including that the children’s investigations were defined by the clinical symptomatology and diagnostic requirements, and having taken expert advice, approved the protocol submitted.
The clinical management and investigation of these children was performed at the Free by a dedicated team of consultant paediatric gastroenterologists, in full consultation with and agreement of the parents of the affected children. The investigations were those thought appropriate in the light of the severity of the children’s symptoms according to the clinician’s judgment at the time.
Had the advice of the institution been sought at the time concerning conflict of interest, they would undoubtedly have advised that any potential conflict should be declared so that others could judge whether such conflicts were real.”
I think it is right that published in the same volume of The Lancet article was a form of a retraction of the paper. I say “a form”, because it was entitled Retraction of an Interpretation.
A I think it was subsequent, was it not?
Q Page 1207, March 6.
A Okay. Right.
Q What you did was you published a retraction and you also published an editorial and the editorial is page 1207.
A Can I just clarify? The statements that you have just read out might have been published in print on March 6 but they were actually released to the media in February. They were actually sent out on the Friday afternoon. We were contacted by Brian Deer on the Wednesday. He came into the offices on the Wednesday and we released those statements on the Friday. We then put them in the print journal on March 6 at the time of the Retraction of Interpretation, just to be clear about the timing.
Q If we go to 1207, this is simply a photocopy of the same edition of The Lancet. Is that correct?
A Correct.
Q We see an editorial on page 1207 and on page 1210 a Retraction of an Interpretation.
A Correct.
Q Can I ask you, did you have to seek legal advice in relation to the wording of those documents?
A Yes, we did.
Q Why was that?
A We are owned by a publisher who has an in-house lawyer and she is regularly consulted over matters of potential libel. Whenever we have an issue we think might be a matter of public interest but might also lead to litigation, we always pass things by her. She certainly took a look at these and also consulted a firm of lawyers external to The Lancet.
Q Did you have particular reason for fearing that litigation might ensue?
A It is difficult for me to remember the course of events exactly. In mid-February we released the initial statements. They were certainly looked at by our lawyers. Then afterwards, when I was interviewed to comment on these statements, I made certain statements which then triggered a few letters from lawyers saying that I might be sued as a result of the things that I had said in the public domain. That kind of ramped up the legal concerns within The Lancet. Then suddenly, you know how it goes, you just completely lose control of the situation. You get one aggressive letter, exchanged with another aggressive letter and I simply became a spectator as I watched these escalating aggressive be exchanged.
Q Who was indicating that you might be sued?
A We had a letter from Peter Carter-Ruck, solicitors on behalf of Dr Wakefield, I believe. I remember being phoned at home by The Sunday Times and being threatened with some kind of action as well.
Q If we look, first of all, at the retraction, which is at page 1210, can you explain this to us? This is very specifically entitled “Retraction of an Interpretation”. Is this a recognised process to retract a paper or to retract the interpretation on a paper in certain circumstances?
A Yes. If you go back and look at the International Committee of Medical Journal Editors guidelines, retraction is a recognised form of reaction to an allegation of research misconduct which might completely jeopardise the findings of a paper.
In this particular case it was a little more complex because nobody was saying that these children did not exist, nobody was saying that they did not have the findings that were reported in the paper or that the parents’ testimony was not indeed genuine. The specific concern was that the basis of the interpretation, that is to say that there was a potential environmental trigger, MMR vaccination, which should be taken seriously and which should be the basis for further investigation. It was that which, in our judgment, and the judgment of this group of authors, was invalidated by the disclosure of a conflict of interest.
Q We see the authors who signed this: Professor Murch and then we have Anthony, Casson, Malik, Berelowitz, Dhillon, Thomson, Valentine, Davies and Professor Walker-Smith. The remaining three authors were a Dr Linnell, and there is little note saying we were unable to contact John Linnell. The others were Dr Wakefield and Dr Harvey. Who was it who ensured that the doctors signed up to this? Was this anything to do with The Lancet who signed up for it or was it done by the doctors internally?
A The first point of contact was a telephone call, if my memory serves me right, between myself and Dr Simon Murch, as he then was, discussing how appropriate it would be to have a retraction of some form, in the end a Retraction of an Interpretation. He then went and secured as many of the authors as he could to sign that Retraction of an Interpretation and then submitted it to us. As I remember, also the Royal Free Hospital was involved in the formulation of this. Although Professor Hodgson is not a signatory, because it is a retraction of part of a piece of work conducted at the Royal Free, they had a very serious interest in what this retraction of interpretation would say. In the end, as I recall, I rang him directly, because it was proving difficult to negotiate this, and said we need to sort this out. There is a desire on the part of the majority of the authors to retract this interpretation, can we have the wording from you.
Q We see the wording:
“This statement refers to the Early Report ‘Ileal-lymphoid-nodular hyperplasia ... published in 1998. It is made by ten of the 12 original authors who could be contacted. It should be noted that this statement does not necessary reflect the views of the other co-authors.
The main thrust of this paper was the first description of an unexpected intestinal lesion in the children reported. Further evidence has been forthcoming in studies from the Royal Free Centre for Paediatric Gastroenterology and other groups to support and extend these findings. While much uncertainty remains about the nature of these changes we believe it is important that such work continues as autistic children could potentially be helped by a recognition and treatment of gastrointestinal problems.
We wish to make it clear that in this paper no causal link was established by MMR vaccine and autism as the data were insufficient. However, the possibility of such a link was raised and consequent events have had major implications for public health. In view of this, we consider now is the appropriate time that we should together formally retract the interpretation placed upon these findings in the paper according to precedent.”
If we turn back to 1207, this is your editorial and I am certainly not, you will be pleased to hear, read it all out but I am going to go to a couple of paragraphs in it. You say at the beginning:
“This week, The Lancet prints a partial retraction - a retraction of an interpretation - from the majority of authors of a paper published by Andrew Wakefield and colleagues. Wakefield and one co-author, Peter Harvey, have not signed this. We hope to publish their response shortly. The original report made clear that the authors did not prove an association between measles, mumps and rubella vaccine in a newly described syndrome but the authors did raise the possibility of a link on the basis of parental and medical histories and they suggested that further investigations are needed to examine this syndrome and its possible relation to the vaccine. This interpretation of their data, together with a suggestion made by Wakefield during a separate press conference held at the Royal Free that there was a case for splitting the MMR into its component parts, triggered a collapse in confidence in the UK’s MMR vaccination programme. It is the interpretation expressed about a connection between the vaccine and the new syndrome that is now being retracted. Today’s retraction comes after debate following the release of new information two weeks ago about the circumstances surrounding the publication of this work. An enormous amount of effort has gone into reviewing and analysing the events before and after publication in the 1998 article. It is now time to look forward.”
You then deal – with no disrespect to your editorial, but I am skipping this – with the background as to research into autism, into research integrity and to public engagements. I shall be inviting the Panel to read this when they retire. We go on in the right-hand column, two-thirds of the way down:
“The reason that today’s retraction is partial and not total is that the discovery of a possible link between bowel disease and autism is a serious scientific idea, as recognised by the MRC, and one that deserves further investigation. Although dismissing the entire 1998 Lancet paper as poor science gives a clear and correct message to the public about the status of any claim regarding the safety of MMR in scientific and clinical terms it is both wrong and damaging. The autism-bowel disease link was considered part of a series of physiological observations judged by the MRCC to be ‘interesting and in principle worth investigating’. Subsequent research has yielded conflicting findings.”
You give the references for those and say that this work should be supported. Then going on, if we may, to page 1209, the left-hand column:
“Publishing controversial new ideas
It seems obvious now that had we appreciated the full context in which the work reported in the 1998 Lancet paper …was done, publication would not have taken place in the way that it did. These are difficult judgments to make in hindsight. For example, our sensitivity to potential conflicts of it is very much higher today than it was in 1998. What we will not do is become profoundly conservative in our decision making about original ideas. A forum to raise new and sometimes unpopular thinking, even on the basis of what might at first appear flimsy evidence, is important – and often vitally so for clinical medicine and public health. How we discuss this new thinking then becomes the central question to answer, not whether we should publish it or not.”
As I say, Dr Horton, I will invite the Panel to read the rest of that editorial to themselves in due course. As far as you are concerned, did that editorial sum up, obviously with the benefit of hindsight at that stage, your views as to the decision to publish this paper?
A Yes, it did.
Q Thank you very much, Dr Horton. Those are all the questions I have to ask you.
THE CHAIRMAN: That is the end of your examination in chief, Dr Horton. You are now going to be cross-examined by Mr Coonan for Dr Wakefield.
Cross-examined by MR COONAN
Q Dr Horton, there will be a little reading, you and I together, following roughly the same format. Before we get to that, I would like to ask you, please, just to step back to the time when the decision was made to publish the paper. This exercise is really a summary of the features which you described to the Panel were in play at the time, because you have given us a wealth of information and detail and the Panel are going to have to grapple with all of that in due course. By the sound of it, when you were looking at this paper, when it was submitted to you, it was subjected to a pretty rigorous process, was it not?
A It was.
Q I do not know whether the process which was in play at The Lancet in 1997, that part of 1997, was the same as in other journals at all, but how does it compare with other journals? More rigorous, do you think, or less rigorous?
A At the time, I would say exactly the same amount of rigour as our major competitors, the New England Journal of Medicine and JEMA.
Q You have told us that as a feature of that process, there were three reviewers: one I think has been identified and as we understand it will be giving evidence before this Panel, an epidemiologist, but you cannot remember now the identity of the other people.
A I cannot, I am afraid, no. I remember one, because Professor Candy used to be in Birmingham and I was a Birmingham graduate. So I remember his name, because he was in the department of paediatrics in Birmingham when I was a student there.
Q This was a paper which needed an epidemiologist to look at it, did it not?
A Yes, indeed it did.
Q I think the response of the reviewers – and for these purposes the epidemiologist – was pretty positive.
A You know, it is very important just to set the record straight about this paper, because I think there has been a lot of nonsense written and spoken about it. One particular piece of nonsense is that this was a study which set out to prove a cause or an association between the vaccine and this syndrome. In no way did it do so. So to judge the paper as some kind of attempt to prove cause and effect was utterly wrong.
Q Thank you very much. I am sorry, I rather interrupted you. Do carry on.
A No, that is fine.
Q That is an important factor, is it not, to dispel a lot of the swirl of myth which has gathered around this paper over the years?
A Very much so. It is a key fact about the paper.
Q It has infected the public view of this paper.
A Entirely misleadingly.
Q And has infected a view of it even in what may be otherwise described as otherwise responsible quarters. Is that fair?
A Extremely fair.
Q The reviews I think, just coming back to a point I was putting to you, the reviews which were received by the editors were as far as you can recall all positive?
A On the basis of what the paper was simply saying, here is a case series reporting a new syndrome and then, when you went back and asked parents, there was an association temporarily with vaccination. But the thrust was the new syndrome and, on that basis, it was entirely appropriate to consider it seriously for publication, correct.
Q Quite apart from the layer of rigour which was deployed by the reviewers and the epidemiologist, we also have the layers of editorial scrutiny by yourself and your fellow editors.
A That is right.
Q By the sound of it, judging by what you told us yesterday, that would have occurred on more than one occasion.
A Correct.
Q So the end product of those layers of scrutiny was that this was a serious scientific idea or hypothesis or finding which deserved to be published.
A We felt so at the time. In the subsequent nine years ---
Q We will come to that in due course. I am just dealing with the initial process. Again, I think it is, if I may say so, terribly important that we do not telescope matters which occurred later with those at the time. Do you agree with that?
A I completely agree with you. I have struggled for nine years to get back to that position every time I have tried to discuss the subject.
Q It is difficult, is it not?
A It is very hard.
Q If we do not make a conscious attempt to divide the two, danger lurks, does it not?
A And indeed I think misleading messages are given to the public.
Q Absolutely. So as a case series, as you have described this paper, this was really a sort of unexceptional model for a paper to be published.
A The model was certainly unexceptional. The point of controversy obviously was the parents’ testimony.
Q I hope you will forgive me. It possibly encapsulates this point rather well, if I may say so, if I could just read a passage from your witness statement and see whether you would like to comment.
*“Our view at the time of publication of the 1998 paper informed by the advice of our peer reviewers was that the paper contained important new information that would be of interest to a general medical readership.”
Is that all right so far?
A Correct.
Q You go on:
“It is unexceptional to publish a case series in this way and to include speculation on the causes of the syndrome or symptoms described.”
That is a fair comment, is it not?
A Correct.
Q Then:
“These kinds of responses may be used to construct hypotheses that require further study.”
Agreed?
A Correct.
Q
“Examples include the first reports of HIV and VCJD. This method is routine in developing hypotheses.”
You agree with that, obviously.
A Indeed. In fact, I would say it is an obligation of investigators to seek those kinds of clues.
Q Then you set out a number of examples:
“In The Lancet in 2000 … ”
And you give the reference –
“ … two respected epidemiologists, David Grimes and Ken Schulz wrote … ”
And you set out a phrase –
“ … descriptive studies such as a case series … ”
Pausing there, because that is what they are.
A That is right.
Q
“Descriptive studies are often the first foray into a new area of medicine. The results of these studies show how many people develop a disease or condition over time, describe the characteristics of the disease and those affected and generate hypotheses about the cause of the disease.”
*document not available to shorthand writers for checking
You say that is a direct quote from The Lancet. You thought it appropriate to include that in your witness statement. That is a rather good exposition, is it not, of the underlying drive to publish?
A Yes. I wanted to quote David Grimes and Ken Schulz in the witness statement because whenever I say anything about this, it always sounds rather self-serving, so using somebody else’s words may be partly self-serving, but at least they are not my words.
Q In any event, you commend that, do you, to the Panel as a pretty good working illustration of the underlying process?
A This is absolutely standard research and publication practice when you are trying to identify a new syndrome.
Q Against the background of the review process and the underlying philosophy of a publication like this which we have just looked at, let us just look for a moment at the paper itself. The Panel have seen so many references, but let us just turn it up again. It is at page 787 in volume 2, the paper itself. I just take you again just to remind ourselves, because it may prove to be an important marker for the Panel in due course. At the top of the page in the left-hand column, the authors collectively assert:
“We did not prove an association between measles, mumps, and rubella vaccine and the syndrome described.”
Then this:
“Virological studies are underway that may help to resolve this issue.”
Then it goes on:
“If there is a causal link between measles, mumps and rubella and this syndrome, a rising incidence might be anticipated after the introduction of this vaccine in the UK in 1988.”
Then it says:
“Published evidence is inadequate to show whether there is a change in incidence or a link with measles, mumps and rubella vaccine.”
That assertion in that way was clearly a responsible assertion to make, was it not?
A It was indeed.
Q The body of the paper was considered to be clearly, because it was passed for publication, again a responsible putting together of this hypothesis?
A Correct.
Q But you at The Lancet have to be, if I may say so, even in 1997, ultra cautious. Is that fair?
A Yes, very fair.
Q You – and I use that word collectively – thought it appropriate to provide some degree of balance by commissioning the commentary which we see in the same edition.
A That is right.
Q The commentary – this is Chen and DeStefano – the commentary device was not an uncommon device to use in 1997, was it?
A No, it was not.
Q It is a device which is available to editors in the interests of providing, if you like, that degree of fine balance to the wider medical readership.
A That is correct.
Q Therefore, as you say, a common device to papers which might – simply might – generate a degree or a flurry of correspondence or even controversy to a greater or lesser degree, being ultra cautious, you provide that degree of balance.
A That is correct.
Q You also accurately took the step of describing this paper, as we can see on the first page, page 783, in the top left-hand corner, as “an early report”.
A Correct.
Q Emphasising the fact that this was very much a finding on a limited number of patients of this syndrome and was indicating that the evidence will then have to be judged to decide whether or not to take this further. Is that fair?
A It is fair.
Q It may be that I shall have to explore this with Professor Candy and Dr Bignall when they give evidence and if you are not in a position to deal with it, do say, but if one goes back to page 787, you see at the top; we dwelt on it for a minute:
“Virological studies are under way that may help to resolve this issue.”
You gave evidence yesterday about an accompanying paper which The Lancet rejected, just simply as a matter of fact rejected that, and we see it is in the log book. That was a virological paper, was it not?
A To the best of my recollection, that is exactly right.
Q Do you remember that it was concerned with laboratory work in connection with the establishment, or otherwise, of the measles antigen and bile pathology?
A Yes, indeed. Might I add an additional comment about the paper.
Q This paper?
A Not the virological paper, the paper we published. One issue that has come up is, well, why did we include the parents’ testimony about the temporal association because it would have been quite easy for us to have erased that aspect of the process.
Q You anticipated my next question.
A I will shut up and you carry on.
Q No, that is very helpful. Will you please carry on?
A It would, with hindsight, have been easy to have taken that aspect of the paper out and just simply reported the syndrome with no parental testimony added, and we discussed this actually in our manuscript meetings: should we ask the authors to remove this part of the paper because it could be misinterpreted. We felt at the time that we had two views about this. The first view was, in the wake of the debate about BSE and Variant CJD, the Chief Medical Officer, Donald Acheson, had been criticised for not disclosing information that he had privately to the public. We thought, here we are in a situation where we have given information, it is admittedly weak from a causal point of view but, nevertheless, it is information that pointed to some possible clue that might require further investigation. Should we censor that or allow it to be published with appropriate caveats and caution? Our view was that it should be published.
Q I am grateful for that. All that comes down to is that, after proper and mature assessment, the decision by The Lancet to publish the paper in the form in which we see it, was a responsible and, as you thought it, informed basis for publication?
A That was my feeling at the time, yes.
Q But that any assessment by others which caused you to ponder hard about this matter, was really a risk that others might misinterpret what the paper was saying?
A That is correct.
Q Ultimately, and do please to do not treat this as a criticism, but ultimately the decision to publish this with the bit in about the temporal link was the decision of The Lancet?
A Yes, it was, of course, our decision to publish this paper with that component in, absolutely. I hope that you will not forget the press conference which, from my point of view, was rather critical.
Q I am coming to that. I am dealing with matters of decision of publication. Again, you do appreciate I have to do this in chronological segments in order for the Panel to get a clear view of what happened each time.
MR COONAN: I note the time. I am now going on to a significant topic which is going to take a little time. I wondered if you might think it appropriate to rise now.
THE CHAIRMAN: Yes, indeed. Ms Smith has also asked us to read the papers 1207 to 1210, so we will also be looking at those papers as well. What we will now do is, I think we will go into camera now to do that reading of a few pages that Ms Smith has advised and us and then we will break for lunch and resume at 2 o’clock. I think Mr Miller, do you want to say something?
MR MILLER: I am going to ask Dr Horton to look at two others extracts from The Lancet between 1998 and 2004, which I am quite happy to hand up now. He has not dealt with this aspect yet, but I am in your hands. I can either hand them in so you can skim that first, because I will definitely be asking him questions about it.
THE CHAIRMAN: It will be easier for us to go through the documents we have been already advised, and Ms Smith has taken some of the issues to those documents, so we would actually go through those documents now. I think we will take those documents, maybe, at the right time.
MS SMITH: Perhaps I could take up Mr Miller’s kind offer as far as Dr Horton is concerned, so he can look at them while he is having his sandwich which may assist him later.
THE CHAIRMAN: Maybe Dr Horton could be given those documents. The Legal Assessor will already have the documents. I will leave that for the Panel secretary to sort out, Mr Miller, and pass on the appropriate documents to Dr Horton to use some of that time to do his reading. We will now go into camera and will read those documents and resume at 2pm.
STRANGERS WITHDREW, BY DIRECTION OF THE CHAIR
AND THE PANEL RESUMED IN CAMERA
(Luncheon adjournment)
STRANGERS HAVING BEEN READMITTED
MR COONAN: Dr Horton, can we move on to the next topic? If you have in front of you bundle 3 at page 1211. I am going to try and avoid jumping round too much, but the focus for the moment will be in the series of pages in bundle 3. The material here was published in The Lancet, in effect after the allegations which have been made by the journalist. One of the facets of the allegations against Dr Wakefield which we can see are, principally, to be seen in allegations 4, 5 and 6, do you agree?
A Yes indeed.
Q We see that in the left hand column and the Panel have been taken through the nature of those as reported. Then in the right hand column, The Lancet’s response to those allegations, based in part at least to those allegations based, in part at least, upon what Dr Wakefield had told you?
A Correct, in part.
Q If I could attempt to summarise the kernel of this allegation, so far as it affects Dr Wakefield, it is this, is it not, that there was a failure to declare Legal Aid Board funding in respect of The Lancet paper, in other words in respect of The Lancet study?
A Correct. I would slightly qualify that, because our conflict of interest, the ICMJ statement and ours, ask for recognition and disclosure of any conflicts of interest. That has a slight latitude to it.
Q I am going to come back to this because this is an important topic.
A It is.
Q If I fail in any way, remind me. Dr Wakefield’s response, which we see that you have captured in the right hand column towards the bottom, the Panel have been taken through it, was to this effect, was it not and I am going to summarise it to help it along. What he was saying was that there had been Legal Aid Board funding, that is point 1. Point 2, that that funding was in connection with a separate, my emphasis, virological study funded by the Legal Aid Board and that no legal aid funding had been in fact been expended on matters covered by The Lancet paper?
A Correct.
Q Therefore, and in addition, that The Lancet paper itself was not concerned with virology?
A That is correct.
Q If one looks at the paper and analyses it very carefully, there is no virology featuring in the paper?
A That is correct.
Q But, he accepted that, as a matter of fact, some of the children in the Legal Aid Board study, funded by legal aid, the virological study, some of those children had in fact featured in The Lancet paper as written?
A Four or five, he said.
Q I am sorry to labour the point, but I think it is terribly important conceptually that everybody understands as we move forward, as I will in a minute, what it is that we are talking about, but, so far, you are with me?
A I am with you.
Q That account which I have summarised was provided to you by Dr Wakefield, in effect orally. He spoke to you and disclosed that to you when you met him?
A That is right.
Q In other words, and again I hope you will be able to agree with me, that he gave full disclosure of those facts at that meeting?
A Yes, after I had laid out the claim that he had that funding, indeed.
Q Absolutely. Otherwise he would not know what he had to respond to.
A Yes.
Q Your response, on hearing on the one hand the allegations I have just summarised and, two, Dr Wakefield’s response which I have just summarised, was to say, and again I am going to summarise and I hope it is accurate, “Well, even if there was no direct Legal Aid Board funding in respect of the children within The Lancet paper study – even if – it still raised a question of conflict of interest”. Is that fair?
A That is absolutely fair.
Q The way you put it, to be precise about it, was that, even if that be right, what he was saying, it still raised a question of a perception of conflict of interest. Is that right?
A Correct, a perception that we felt should have disclosed to us, to his co authors as.
Q The question, and in saying that, supply a degree of emphasis to that phrase, “a perception of a conflict of interest”, not least because that is the phrase you used in the literature, is that right?
A That is right.
Q Can we look at it. We will pick it up, first, on page 1212 at the top of the left hand column. Dr Horton, I am picking this narrative up in the light of Dr Wakefield’s explanations to you which you have captured in the preceding column on 1211.
A Yes.
Q You say:
“The difficulty of adopting a dual role as a clinical investigator and as a participant in an evaluation on behalf of the Legal Aid Board is revealed in Dr Wakefield’s response to allegation 5.”
Then you quote a short passage from his response which, if I may say so, is accurately reported. You then say:
“... the perception of a potential conflict of interest remains. Editors and reviewers should have had an opportunity to take his dual role into consideration when assessing the paper for publication.”
Then you say:
“Finally, although the Legal Aid Board funding referred to a different aspect of Dr Wakefield’s work from that reported in The Lancet, the perception of a concept of interest nevertheless remains.”
That last extract, Dr Horton, you are in effect picking up and reproducing Dr Wakefield’s position, namely that the funding, the Legal Aid Board funding, applied to a different study?
A Correct.
Q Again you say, “But even so, it is a question of perception”.
A It is. Our feeling was that he should have shared that information with the editors who would then have shared that information with the reviewers.
Q I understand that was the view then. Can we move on to complete the snapshot of events in 2004 by looking at 1215? This is Dr Wakefield’s statement in relation to the allegations by the journalist. These are his exact words which he wrote. I look at the last paragraph on the left-hand side:
“I regret the difficulties that this issue has caused my colleagues. I am grateful to them for their advice and support. I am enormously grateful for the timely manner in which Richard Horton has dealt with this issue and for his clarification of the issues surrounding perception and reality where conflict of interest may be concerned.”
Again, it is the use of this word “perception” of conflict of interest. Can we stand back for a minute? I just want to set the scene for where I am going next. Sometimes, when matters of direct funding are concerned, and let us give a blatant example, somebody does a clinical trial for a particular drug company and it turns out that, one, he has shares in the company; two, he has been receiving fees from the company, and so on. Those are obvious cases of an actual conflict of interest?
A Yes. If that sponsor is responsible for the drug under test, yes.
Q I do not want to be unduly journalistic about this, but that is pretty barn door stuff, is it not?
A It is.
Q The other side of it concerns this question of perception, because by that one is talking about the perception by others of the fact that there may be a conflict of interest. Is that fair?
A Correct. To take that idea on, it is perception that in the mind of the reader may lead to some kind of bias.
Q Absolutely, but the essence of it is that the question has to be asked through the mind of the reader.
A That is exactly right.
Q It is a question that has to be asked through the mind of the reader as to whether or not there might be a conflict of interest.
A Yes. We are asking the author to think about that, to recognise it and then to disclose it.
Q When you were saying, as you did, to Dr Wakefield, leaving aside questions of direct funding because he provided an explanation about that, did he not?
A He did.
Q We are then into the territory of perception of conflict of interest. What you were seeking to capture by that, if I may suggest, was two facets of this: first of all, that it was discovered that he had been involved - and I use that word generally - in litigation per se. As you have indicated this morning, there appears, on the face of it, and on the face of the documents which became available to you, a connection in a proposed litigation process, and had been for some two years, in accordance with those documents, is that right?
A Yes, the two newsletters. That is right.
Q We can identify that as one factor under the umbrella of perception of conflict of interest, can we not?
A Yes.
Q The second facet, may I suggest, is the fact that he, as he admitted, used Legal Aid Board funding on the separate virology study but that there had been, as a matter of fact, a cross-over or overlap of patients in the two.
A Exactly. From where I sat looking at this, one human being was asking a lot of that one human being to materially separate in their brain these two aspects in a group of children that overlapped.
Q So the argument is followed through, these aspects fall under the umbrella of a perception of conflict of interest. In other words, the question that has to be asked when it is posed is a question through the mouth of a potential reader.
A You are asking the question of the author to put him or herself in the position of the reader and say: Do I think that that person looking at my work might think that there was a potential bias?
Q Absolutely right. I do not disagree with any of that formulation. Therefore, the key to this, in terms of perception of conflict of interest, is to attempt to pose that question that you have just formulated, with which I agree, in the context of 1998. I see you smiling.
A That is absolutely right. That is the challenge here.
Q May I suggest, you wrote a book about all of these matters, did you not?
A I tried.
Q May I say I have read it; it was a good read. This is precisely one of the matters you raise in the book, is it not?
A Yes.
Q The challenge of applying that sort of formulated question you put back to 1998.
A Yes, that is right.
Q If we go back to 1215, last paragraph, of course Dr Wakefield’s words:
“I am enormously grateful [to you] for clarification of the issues surrounding the perception and reality where conflict of interest may be concerned.”
It looks from that as if you were seeking to explain to him the difference between perception and reality.
A I can remember the conversation in my office, exactly.
Q He did not appear to appreciate the difference, did he?
A We disagreed about the interpretation of the difference, yes.
Q In other words - again I do not want to be over robust about this - you found yourself having to spell out the difference to him, is that fair?
A Yes, I agree.
Q I think a further comment which I think you made on this challenge can be found at page 1209. This was part of this long editorial beginning at 1207 and, for present purposes, I am going to take you to one short passage at 1209. You say this under the heading of “Publishing Controversial New Ideas”:
“It seems obvious now ... ”
Pausing for a minute, you are there talking in 2004, six years down the line?
A Correct.
Q
“... that had we appreciated the full context in which the work reported in the 1998 Lancet by Wakefield and colleagues was done publication would not have taken place in the way it did. These are difficult judgments to make in hindsight. For example, our sensitivity to potential conflict of interests is very much higher today than it was in 1998.”
You draw attention to a number of footnotes. You referred to two footnotes and there you actually, if I may say so, put your finger on the problem, is that fair?
A Yes, absolutely. It is very difficult with hindsight.
Q Very difficult with hindsight and again one has simply got to put oneself in a position of how such a question about potential conflict of interest should be answered in 1998?
A I think that is right. When we looked back at the quality and the nature of the debate in the 1990s, however, we felt that the guidance that we provided in the journal, and also importantly the guidance provided by the International Committee of Medical Journal Editors, was nevertheless clear and unambiguous and should have led to disclosure.
Q I fully appreciate that has been your view. It is set out in these documents we have seen and I will take you and the Panel to them so we can see it and we do not shirk it. That is obviously a view expressed in 2004 by, dare I say it, the editor of The Lancet who may have had cause to feel somewhat upset that these allegations were being made involving the decision at least to publish.
A I was less upset about the allegations being made regarding the decision to publish. Remember that I had lived for six years with allegations being made about the decision to publish so I had become somewhat used to them, even perhaps immune to them. What I was more anxious about was a new material fact about a conflict of interest which I was unaware of. I may add to that, these two others in the room were also surprised about that, two of the other co-authors. When the Royal Free investigated it, they too came to the conclusion, yes, with hindsight, that conflict should have been disclosed.
Q First of all, you keep using the word “hindsight” and the Panel will have to judge the value or lack of it in employing hindsight to make these judgments.
A Be careful about the use of the word hindsight. The guidance in 1997 in the ICMJE regarding the disclosure of conflict of interest was unambiguous: it was recognise and disclose to the editors - no ambiguity there.
Q I am going to turn to it in a moment. I will take time over this because it is clearly an important matter. The question of the material which was in play in 1998 is, of course, quite important. It may be timely to turn up - this is my word - the rubric which used by The Lancet directly on a monthly basis to remind authors of the obligations they had. It may be helpful, before we embark on the next phase of this, to turn up volume 2 at page 616. You have told us this morning that this was published on the 4 October 1997 so we can take it that it was in operation at the time that The Lancet were considering these matters.
A Correct.
Q In the left-hand column we see under conflict of interest “and funding”. You see the two elements there. I am going to look at it again with you, if you do not mind. We have looked at it already but I make no apology for looking at it again with you.
“Conflict of interest and funding. The conflict of interest test is a simple one. Is there anything, for example a shareholding in or receipt of a grant or consultancy fees from a pharmaceutical company or a contract from a medical devices’ manufacturer that would embarrass you if it were to emerge after publication and you have not declared it.”
It says at the end:
“All sources of funding must be disclosed.”
I appreciate what you say about The Lancet signing up to the International Committee of Medical Journal Editors and we will look at that in a minute. As a straight forward rubric for the declaration of conflict of interest and funding, let us look for the minute, bearing in mind the distinction, which I think you and I have drawn, between direct funding for a piece of work for publication and, on the other hand, a perception of a conflict of interest, in other words asking that formulated question that you posed earlier. Do you follow the distinction? Against the background of that distinction that we had come to, can I look at the elements of this advice.
A Yes.
Q First of all, there is clearly an emphasis on funding, is there not, in this rubric?
A Correct.
Q It really has within it a question, which is to be asked directly by the author, of whether or not he or she has any direct funding stake in the piece of work which is being done.
A Correct.
Q That piece of advice as stated - no more than that - does not refer, does it, to perceptions of conflict of interest?
A It does not use the word “perception”, no.
Q It does not, is it were, go into any advisory detail as to what a perception of conflict of interest may consist of.
A It does not.
Q It does not deal with the problem of dual roles or dual commitments.
A It does not go into that level of detail, no. That is why the statement needs to be read in conjunction with a fuller definition of conflict of interest published in the ICMJE guidelines where, if we look at that, it confines conflict of interest much more fully than is defined there.
Q It does and we will look at it in a minute. Clearly somebody who was reliant upon this, who reads this on a monthly basis or six monthly basis and so on and so forth and had not gone to the website, would receive this - and it is my word and not intended to be loaded or pejorative in any way - limited advice.
A Correct.
Q A focus on the matters captured in that rubric in 1998 was also featured I think in The Lancet. You will now have to turn to volume 7, page 232. This was an editorial, I think you said this morning, not in fact written by you, although that does not matter, it was written in 1996, so two years before, and in the bottom right-hand corner, you or your colleague says:
“The Lancet’s simple test, by no means perfect, is this: would a non-disclosed commercial interest, should it be revealed later, prove embarrassing to an author?”
Really, that was the flavour, was it not, of the advice which we have just looked at in 1998?
A That is right. The reason why the word “commercial’ is inserted there, rather than in the earlier advice which we have been talking about, is because this was an editorial commenting on a very specific case.
Q Whilst we are in the same volume ---
A It might be worth just reading the penultimate sentence there, because to my mind that is quite important. I agree with you that these are fine issues of definition and subjectivity, but here, when we have:
“We rely on the conscience and judgment of the author to draw our attention to such a personal conflict.”
It is very hard to legislate for every possible instance of conflict of interest in a set of guidelines for authors. There are probably as many types of conflict of interest, subtly different from one another, as there are authors. What one is trying to do is to very broadly define the bounds of conflict of interest. Actually that statement, “We rely on the conscience and judgment of the author” is very important to the way we are inviting authors to consider the issue of conflict of interest.
Q As you have said, it is ultimately a subjective question that the potential author has to ask.
A It is a subjective question, given the time and place when he or she is asking it.
Q We are dealing with 1997.
A We are.
Q Of course, the questions that you, as an editor, expect authors to be posing to themselves would have been really a reflection of the general culture operating in the research community at that time.
A Precisely.
Q By the sound of it, you are in effect I think accepting that there would have been a breadth of view as to what was required to be disclosed at that time when authors operate their subjective judgment
A We went through some of those views this morning, that is right.
Q The important thing is, not just from your standpoint as an editor, but equally in the research community at that time, there would have been quite a broad sweep of opinion as to what your rubric actually would have triggered off or should have triggered off. Is that fair?
A Yes. At one extreme, some people thought this was, in a phrase, “The new McCarthyism of science”. Others would have felt entirely the opposite.
Q When you said in 2004, as we have seen already, that one’s sensitivity to potential conflict of interest is greater now, that is entirely right, is it not, because the sensitivity of those in the research community has moved on, has developed, has it not, over a period of time?
A Correct.
Q The international committee which you were referring to and I promised that we would look at it, because I knew you were keen to see it – and why not? – is at page 202 of volume 7. Just so that the Panel fully appreciate the point you are making, what is actually commented on in this right-hand side is accessible, or would have been accessible, to an author if they had taken the steps to go and look and seek out this information through the route which is actually set out in The Lancet.
A Which is why we put this in a box at the very top of the section on instructions for authors.
Q I appreciate that, but they have to take the steps to seek it out.
A They would. We gave the web address of course.
Q Whereas those folk who just looked at what was in The Lancet itself would just have your advice on page 616.
A That is correct.
Q In effect – and again, I do not think I need go through it in full – this right-hand column is dealing here with the question of potential conflict of interest, is it not?
A This is about potential as well as real conflict of interest, yes, including the perception of conflict of interest.
Q Absolutely, but for my purposes, I just wanted to focus on the question of perception. I take it as read that there will be the question of direct conflict, immediate conflict will be apparent to anybody who is going to submit an article to The Lancet.
A We would hope so.
Q If you could just turn on, please, to page 209 in volume 7, where this morning you looked with Ms Smith at the body of this document, it is a BMJ editorial in 1994. There is a passage which was not cited – this is not a criticism; it just was not highlighted – it is on page 209, the third paragraph up. Perhaps I could read it:
“It is financial conflicts of interest that cause the most concern. The New England Journal of Medicine, which has led the way with its policies on conflict of interest, concentrates on financial conflicts of interest on the grounds that they are widespread, optional, and seductive. Thomson says that policies concentrate on financial gain because it is more objective and easier to regulate by impartial rules. These arguments have much to recommend them, but we want to try to have a policy that covers all conflicts of interest. Other sources of conflict are personal, political, academic, and religious, and we believe that these may be just as potent as financial conflicts.”
Then over the page, the author there says:
“We plan as soon as possible to include the source of funding for a research study in all scientific papers … “
Then in the second paragraph:
“We are moving the policy along by always recording the source of funding for research, asking people to sign a document, and sometimes disclosing conflicts … ”
And so on. One gets a picture – is this fair – from looking at these earlier documents that there was a policy in the course of development?
A I agree.
Q With any policy development, you may be leading, as responsible editors, but the research community are very frequently somewhat lagging behind.
A That is fair. That is entirely fair and I think actually it is the duty of this small group of journals to try and lead that kind of policy development, but to make sure at every stage that we are debating it and discussing it in our journals.
Q For you, it is a difficult task, is it not, to try and get a balance right between setting out advice to would-be authors, at the same time trying to encourage people to bring forward research papers on the one hand, but tipping it too much the other way and frightening them off?
A Yes. You do not want to have a culture of suspicion in science. You want to have a culture of trust and vigilance, but if we go too far, then we can actually damage the whole culture of research.
Q That is quite a difficult balance, is it not?
A Yes. It was then and it remains so now.
Q As you have said, it is difficult to bring the research community along on the leash.
A Often, you know, we create conflicts with the research community.
Q I think your assessment was at about the relevant time – and I am talking about 1997/1998 – that very often or, dare I say, frequently, these conflicts of interest were simply not reported.
A There are two levels of not reported. Not disclosed to the editor and then not published and we were trying to work on both of those issues.
Q In terms of the first category, not disclosed, they simply frequently were not disclosed?
A Despite the guidance that was issued in the ICMG guidelines up 1997, yes.
Q You had cause to comment on that fact in your book.
A That is exactly right.
Q Dr Horton, I think I can ask you to put volume 7 to one side. Were you supplied with a further document from the BMJ?
A Yes, indeed.
Q Have you had an opportunity of reading that?
A I have.
MR COONAN: Sir, at this stage, I would invite the Panel to receive a file – not a very large file – of some literature which Ms Smith had yesterday and Dr Horton has also had an opportunity of looking at. Could I ask that the Panel receive this? Rather than trying to insert them in the main bundle, these are produced on behalf of Dr Wakefield and they should therefore, if I may suggest, have a D number. I think it may be D2. (Same distributed and marked as D2) (To the witness) Dr Horton, you have probably understood the purpose of putting these documents in front of you.
A Yes, indeed.
Q But the Panel will not at this stage know, so can I just set the scene? We have been looking together at the, as I call it, 1998 rubric and we have looked at some of the literature leading up to that. I now want to look, with your assistance, please, at how the rubric has developed in the ensuing period to see where we have got to and in particular to see where we have got to by 2004. Some of these documents of course you will be intimately familiar with, since they were published obviously in The Lancet and I suspect – rather more than suspect – that you are familiar too with the other documents.
A Yes. I am familiar with all of the documents you have supplied.
Q Then we can deal with it quite quickly. The first document was published in 1998 in the NEJM and is an important study which I think has been cited by you in subsequent articles. Is that right?
A Yes, indeed.
Q Rather than me laboriously taking you through it line by line, if you have looked at it and refreshed your memory, it might be helpful if you were to tell the Panel what is your summary of it, if you are happy with that.
A Yes.
Q If I think there are bits which need to be added, I can deal with it. In a word or two, what was this all about?
A The context of this is that the 1990s had seen a very, very serious adverse and conflicting debate around the safety of a particular class of drugs for heart disease, calcium-channel blockers. Finally the debate was coming to be resolved by the late 1990s, but what this study did – and I should say that this was a pivotal study in increasing our concern about the influence of conflicts of interest in the interpretation of clinical trial data particularly. I would take you to page 8 of this document, where I think it summarises the conclusions very well. On page 8, at the beginning of the section entitled “Guidelines for Disclosing Conflicts of Interest”, the first sentence of that summarises the general view about the way medical journals had handled these matters:
“The medical profession has failed to develop and enforce strict guidelines for disclosing conflicts of interest.”
The bottom line message of this paper, so to speak, was that only two of 70 articles in this particular study of a very controversial area had disclosed the authors’ potential conflicts of interest and they go on to say, “Clearly , our current disclosure policies are inadequate.”
Q Just pausing there for a moment, again, if I could add in material which is also contained in this paper, the authors had looked at all the English language medical literature published in March 1995 through to September 1996 for articles examining the controversy about the safety of calcium-channel antagonists. So that was the focus of the study. There is obviously a wealth of literature on this topic, or was at that time – more now probably – and that really was, as you have said, a pivotal finding that only two of the 70 articles included in the studies disclosed the authors’ potential conflicts of interest.
A Yes, indeed. I know because this was an article which we discussed at great length at subsequent ICMJE meetings.
Q As the authors say in the next sentence, Dr Horton, “Clearly, our current disclosure polices are inadequate.”
A That is right. The situation was we thought, we believed we had very clear guidance about recognising and disclosing conflict of interest that were unambiguous. The problem was that it seemed that authors were not paying a blind bit of attention to it.
Q Absolutely. That was your experience and the experience clearly revealed by the authors of this study.
A Correct.
Q It may be, if I could put it this way, I hope fairly, that something may be believed to be unambiguous, but it is not actually treated like that by the authors.
A That is right. The way we looked at it was, clearly we felt we had a guideline which was unambiguous. People were not paying much attention to it. That did not mean that people should not pay attention to it. People should have been aware of it, they should have acted on the basis of that guideline, but the fact was – and this was an important piece of evidence in that judgment – people were not paying attention to that guidance.
Q Again, perhaps – and I hope this is a fair comment – a reflection of the culture in the community at that time?
A A reflection of the culture but, if I may slightly editorialise, should not be used as an excuse for dismissing those guidelines.
Q Nobody is seeking to dismiss the guidelines, that is not part of my exploration of those matters with you at all?
A I understand.
Q I am just exploring with you how the community viewed them as a whole. Dr Horton, can I take you down to the penultimate on page 8. Perhaps I could read this:
“The extent to which the pharmaceutical industry influences clinicians’ and researchers’ opinions cannot be determined by the results of our study. We believe that the authors we surveyed expressed their own opinions and were not influenced by financial relationships with pharmaceutical manufacturers. However, it is our opinion that scientific authors are naive about public perceptions concerning such relationships. We wonder how the public would interpret the debate over calcium channel antagonists if it knew that most of the authors participating in the debate had undisclosed financial ties with pharmaceutical manufacturers. The medical profession needs to develop a strong policy governing conflict of interest.”
That again is a powerful opinion by the authors of this study, is it not. Would you agree?
A It is a powerful opinion and it is one that I shared.
Q I would like to move on to tab 2 in your file. We are now in The Lancet in September 2001, so, in effect, things have moved on now between three and four years. I am not going to take you laboriously through the whole of this editorial. It is there, obviously, for the Panel to read and, if you wish to draw attention to any particular part of it, I invite you to do so. First, did you write this?
A No, I did not. This was written jointly, if such a thing can be believed, by the International Committee of Medical Journal Editors. The context, it is useful, since you referred to the Stelfox et al paper in the NEJM, that paper, together with the debate around the whole pharmaceutical industry funding of clinical trial research, triggered a big public debate. As we met in years after 1998, that was a major topic for our discussion. In the end we thought we had to go a step further with our guidance that we had in 1997/1998, and so we drew up the guidance that you see on page 855 of this editorial where we try to substantially strengthen what we were asking of authors. We were trying to really draw a lot of publicity to improve the process of disclosure.
Q We will look at it in a moment, but, essentially, I think what you are saying is that from 1997/1998, through to 2001, a significant debate occasioned and supported by the Stelfox paper that we have just looked at and, in effect, to devise a rather more unambiguous advice centre stage, is that right, for researchers to avoid the Stelfox data arising?
A Yes, in particular, explicitly to ask authors what role the funding source had in the production of that piece of work.
Q Can we look at how it is put. This is 2001, would it be fair to say, therefore, that this represented something of a sea change in the way in which matters were being spelt out?
A A step change, I would say.
Q A step change. I am content with that.
A I might add that this statement received a huge amount of publicity in the new media at the time because the way it was conveyed was editors getting tough about commercial conflicts of interest.
Q Let us look at the step change. This, of course, is in The Lancet and is intended that authors should read it and comply with it.
A Correct.
Q Forgive me if I jump around, I do not mean to mislead. In the interests of brevity, I pick it up on the fourth line:
“Conflict of interest exists where an author (or the author’s institution), reviewer, or editor has financial or personal relationships with other persons or organisations that inappropriately influence (bias) his or her actions. The potential (my emphasis) of such relationships to create bias varies from negligible to extremely great; the existence of such relationships does not necessarily represent true conflict of interest, therefore. (Relationships that do not bias judgment are sometimes known as dual commitments, competing interests, or competing loyalties). The potential (my emphasis) for conflict of interest can exist whether or not an individual believes that the relationship affects his or her scientific judgment. Financial relationships (such as employment, consultancies, stock ownership, honoraria, paid expert testimony) are the most easily identifiable conflicts of interest and the most likely to undermine the credibility of the journal, the authors and of science itself. Conflicts can occur for other reasons, however, such as personal and family relationships, academic competition and intellectual passion. All participants in the peer review and publication process must disclose all relationships that could be viewed as (my emphasis) as presenting a potential conflict of interest.”
A Correct.
Q Again, it would be otiose of me to make any further comment other than to say that one can see immediately why you say that this represents a step change in the advice given to authors.
A Correct.
Q If we look at comments by you under the “Potential conflicts of interest related to project support”, which I do not think we need take you to. Finally, as it were, to tie the ribbon round it, we see at the end on page 856, that you are one of the senior authors who put your name to that step change?
A Yes, indeed.
Q If I now take you to tab 3, there is a short document headed, “The Lancet’s policy on conflicts of interest”, January 4 2003. I should ask you, and I think it is self evident, that you jointly wrote this with Astrid James?
A Correct.
Q We see in the first few lines, in effect what you have just been saying. Is that right, the summary on the first ten or so lines?
A Yes. What we were doing there was simply taking the ICMJE guidance, which we had already signed up to, but reaffirming it in this January 2003 comment.
Q The second page in the panel on the left hand side, advice directly to authors:
“Conflict of interest statement includes:
Any financial arrangement (employment, consultancy, stock ownership, honoraria, paid expert testimony, patent applications, travel grants) that could...”
again the word “could”.
“... bias your submitted work.
Any personal relationship with other people or organisations that could bias your submitted work.”
Other matters are again set out. As you say, reaffirmation of the position, the step change, that had been reached by 2001?
A That is correct. I can see, I totally see where you are going with this. It does not negate the fact that the guidance in 1997 was equally clear about recognising and disclosing. I do just want to emphasise that.
Q We will come to that, but you see there is a method in my approach?
A I do.
Q Because you are intimately involved with the formulation of the rubric in 1998. Perhaps, if I may say so, it is not unnatural that you should feel it is unambiguous. It is a fair point, is it not?
A I know what you are saying.
Q I have to do this in order to assist the Panel for them to make a decision about this.
A Yes.
Q In tab 4, there are two documents which relate to February 2004. The first is at tab 4, the second at tab 5. In tab 4 there is a commentary, again by you and others – we can see your name at the bottom of it on the second page – and there is an editorial on the conflict of interest. Again, I want, literally, to pick out a number of features of this. Again, if you want to draw attention to any other aspect, please do. You say in the first paragraph on the left hand side:
“Conflicts of interest (or competing interests as they are also called) are ubiquitous. Editors, authors and reviewers of manuscripts all have conflicting interests that can lead to bias in what we publish. What we can do, though, is to reduce the potential for bias by having systems for managing conflict of interest in the journal.”
Then you refer to the January 2003 edition, to which I have taken you:
“We have described our systems for managing potential or actual conflicts of interest for authors.”
Later on in this in this article, you say at the top of the next column:
“Disclosure sets the minimum standard that is slowly spreading beyond medical journals.”
You set out how the Nature Publishing Group also has been affected by the way in which (my phrase) the wings of change of change were blowing through the whole area, is that right?
A Yes. This particular comment is dealing less with research and more with review type material.
Q It is the same principle?
A It is the same set of principles, just applied to a different type of article that we would publish.
Q You demonstrate at the bottom of the right hand column, you make a comment:
“Editors or reviewers may be uneasy or uncertain about actual or perceived bias. For example, we rejected a spontaneously submitted review in 2001. Despite excellent reviewers’ comments because one of the two authors was an employee of a relevant manufacturer. A clinical adviser who regularly attends our manuscript discussion meeting was concerned about how the review would be perceived by readers, and so were some editors, leading to the decision to reject. We do not expect readers to have to work out whether a review is biased and would rather risk rejecting the occasional objectively balanced paper than ask readers to invest their time deciding for themselves.”
Pausing there, all this is a reflection of this debate, is it not, that was taking place in the middle and late 1990s?
A Yes, indeed it was. Maybe at some point I could come back to try and assist the Panel in explaining how we arrived at our decision in early 2004.
Q I am about to turn to that.
A You are. Great. Because we have nicely gone into a slightly abstract chronology of the debate around conflict of interest, which is helpful, but slightly takes us away from the facts of the matter in relation to the 1998 Lancet article.
MR COONAN: Dr Horton, I am laying the ground for getting there and I know you are anxious to get there, but we will.
THE CHAIRMAN: I am sorry, I am just looking at the time as well. You are now moving on to a very important issue and I just wonder whether this would be the appropriate time to have this break and then we are refreshed when you move on to this important point.
MR COONAN: Certainly.
THE CHAIRMAN: We will now adjourn. It is 3.10 pm and we will resume at 3.30 pm. Dr Horton, you are still under oath and still in the middle of giving evidence, so please do not discuss this with anybody. We will resume at 3.30 pm.
(The Panel adjourned for a short while)
MR COONAN: Can we turn to tab 5 in the same file? I hope, on the reproduced copies, we can see, right at the top of the page, in handwriting, February 2004?
A Yes.
Q We can see that, at the top of the page, because of the date this would be operative at the time when these matters came to a head in 2004.
A Correct.
Q The Panel can follow this on the third page, by looking at the advice given by The Lancet to would-be authors in terms of conflict of interest and source of funding, and perhaps I can read it:
“A conflict of interest exists when an author or an author’s institution has financial or personal relationships with other people or organisations that inappropriately influence (bias) his or her actions. Financial relationships are easily identifiable, but conflicts can also occur because of personal relationships, academic competition or intellectual passion. A conflict can be actual or potential and full disclosure to the editor is the safest course. All submissions to The Lancet must include disclosure of all relationships that ...
Then my emphasis –
“... could be viewed as presenting a potential conflict of interest.”
Then there is the citation from The Lancet in 2001, that is the step change article that we looked at. You go on:
“The Editor may use such information as the basis for editorial decisions and will publish such disclosures if they are believed to be important to readers in judging the manuscript.”
Then it continues with a section dealing with conflict of interest statements for authors:
“At the end of the text, under a subheading ‘Conflict of interest statement’, all authors must disclose any financial and personal relationships with other people or organisations that could inappropriately influence (bias) their work. Examples of financial conflicts include employment, consultancies, stock ownership, honoraria, paid expert testimony, patent applications and travel grants all within three years of beginning the work submitted. If there are no conflicts of interest, authors should state that there are none.”
Pausing there, obviously the text has undergone some change, has it not, even since 2001?
A Yes, there are some modifications.
Q In the right-hand column, under the heading “Role of the funding source”:
“All sources of funding should be declared as an acknowledgement at the end of the text. At the end of the Methods section, under a subheading ‘Role of the funding source’ authors must describe the role of the study sponsors, if any, in study design; in the collection, analysis and interpretation of data; in the writing of the report; and in the decision to submit for publication. If there is no Methods section, the role of the funding source should be stated as an acknowledgement.”
We can take it that was the advice available in 2004. The last milestone on this journey is 2007, so we are going to jump to today. Tab 6, again this is The Lancet, February 3 this year. In the right-hand column. Forgive me, but I will need to read it into the transcript:
“A conflict of interests exists when an author or the author’s institution has financial or personal relationships with other people or organisations that inappropriately influence (bias) his or her actions. Financial relationships are easily identifiable but conflicts can also occur because of personal relationships, academic competition or intellectual passion. A conflict can be actual or potential and full disclosure to The Editor is the safest course. Failure to disclose conflicts may lead to publication of a Department of Error.”
That is a well known, if I can put it this way, “sin bin” for authors at the bottom of the article. Is that right?
A Very well put.
Q It says:
“All submissions to The Lancet must include disclosure of all relationships that could be viewed as presenting a potential conflict of interest.”
Just pausing there, that is really a formulation of the question that you formulated during the course of your evidence this afternoon, is it not?
A That is right, exactly.
Q Then under “Conflict of interest statements for authors”, I take you down to the fourth line:
“Conflict of interest statement, all authors must disclose any financial and personal relationships with other people or organisations that could inappropriately influence (bias) their work .... employment consultancies, stock ownership, honoraria, paid expert testimony, patents or [a new one] patent applications and travel grants all within three years of beginning the work submitted. If there are no conflicts of interest, authors should state that there are none.”
Then over the page, finally:
“All sources of funding should be declared as an acknowledgement at the end of the text.”
I think this text has not changed. Just looking at that, we see, quite apart from textual changes over the period of time in question that we have just looked at - I hope it is a fair birds-eye view of this - obviously some substantive changes as well in the period 1997 to 2007. That is right, is it not?
A Correct. I think it is a fair bird’s eye view. I think you have led us a little bit away from where we started.
Q I have not finished the process yet. Of course, one has to bear in mind to come not just your evidence but that of Dr Wakefield. He has a viewpoint here too.
A Of course.
Q Can I step back chronologically and go to tab 7? This is an editorial in the BMJ in August 1998. This was the document which I think you had an opportunity of looking at today earlier on. It is dated August 1998 and it has some significance because although not exactly at the time when the paper was submitted to The Lancet nonetheless we can take it, can we not, that the sentiments expressed in it are apposite to the period when the paper was submitted to The Lancet?
A Correct.
Q There would not have been any real difference at all. This is by Richard Smith. I think, if we go to the second page of this, the second paragraph on the left-hand side, beginning “These two papers”, I am going to pause for a minute. To avoid us having to laboriously read what goes before that, the reference to these papers is a reference to Stelfox and also to another paper called Barnes on the issue of the health affects of passive smoking published in JAMA. We can see that from the text in the editorial. Take it from me that is what it refers to.
A Yes.
Q What Richard Smith said was this:
“These two papers and their predecessors begin to build a solid case that conflict of interest has an impact on the conclusions reached by papers in medical journals. They also showed convincingly that medical journals are failing to get authors to declare conflicts of interest.”
I drop down a number of passages under the heading “What should the BMJ be doing,” in the fifth line:
“We send authors of all original papers, editorials and review articles and of selected letters a form in which we define what we mean by conflict of interest and ask them to sign to say whether they have one. We have gone for a broad definition that extends beyond financial interest to personal, political, academic and religious ones. With original papers we give the source of funding and disclose what authors have told us ...
Our impression, supported by the two recent papers, is that many authors are willing to sign that they do not have a conflict of interest when by our definition they do. We have two hypotheses to explain this. Firstly, authors think that an admission of a conflict of interest implies wickedness. We do not think so. Secondly, authors are confident that they have not been influenced by a conflict of interest and so do not tell us they have one. Our response is that bias works in subtle ways and that none of us is blessed with knowledge of our own motivations and mental mechanisms. We are thus proposing some changes to see if we can do better. They will be phased in from now.”
In that editorial in 1998 - you can see now I have come right back to the time - those are sentiments that you would not disagree with?
A No, indeed. I think it is a very accurate reflection of how we felt. May I just add one codicil from that BMJ editorial you cite? Although the two papers which are, in some sense, the trigger for the editorial are both from 1998, the Stelfox and Barnes paper, there are many references cited, in particular 12 to 16, which are from 1994, 1996, which illustrate that this debate, this escalation of concern about conflicts of interest, was going on right from the early mid-1990s. Although the 2001 ICMJE guidelines which I described as a step change were indeed a step change, in fact there was a continuous ramping up of concern during the early mid-1990s to 2001. I do not want the impression to be conveyed that suddenly in 2001 we all woke up: we did not. This was something that had been building for seven years.
Q Again, I am not suggesting otherwise. I think you painted a picture of the editors of these responsible journals seeking to develop, over a period of time, the rubrics, as I have called it, to apply to would-be authors but that there was a big gap between good intentions and actual performance by the authors. Is that fair?
A That is a fair generalisation of the position, yes. That is a fair generalisation but -and again you might be coming onto this.
Q I am. The last piece of historical literature that I am going to ask you to comment on, I now just want to illustrate - and I preface this by saying it is not a scientific illustration and I am not going to put in front of you any particular piece of paper but I am going to ask you to comment as to whether you are surprised by this or not. As an exercise which anybody can do, it does not have to be a scientist - if I can do it anybody can do it - if you look at The Lancet publications in 1998 and just take the first 26 weeks of the publications in 1998, January 1 to may 31, volume 351, you get a total of 1,000 contributions made by 3,567 authors.
A Are these research articles? I doubt it because that is far too many.
Q All contributions.
A Including news pieces and letters?
Q Yes, including letters.
A OK.
Q What I have found is that there are declarations of conflict of interest from only five authors in two letters. That can be checked and I am more than happy to be corrected mathematically. I do not pretend it is a piece of exact science but errors and omissions accepted you understand the thrust of what I am suggesting.
A Yes. I would simply say that remember there are two parts to this: disclosure to the editors and then a decision by the editors, in conjunction with the authors, at the time in 1998, about whether that conflict should be disclosed and published. I have not checked your figures but let us assume they are correct. In no way does that reflect whether there was a disclosure of a conflict of interest to the editors. The requirement that we sought was there should be a disclosure to the editors so there could then be a discussion about whether it should be published or not.
Q I entirely accept that. It is simply put by way of an illustration of, and I will give you the second statistic in a minute, how there has been this significant change in the way both authors and editors view this difficult topic. The second statistic, again which can be checked.
MS SMITH: I am unhappy with Mr Coonan putting statistics. He appears to be giving evidence. Unless he is going to put the date he derived this in front of the witness, we cannot expect the witness to give an informed answer without checking the information against it. As he has already identified, there are all kinds of considerations other than simply doing an arithmetical exercise. Anyway no-one has checked the arithmetical exercise. I am concerned that Mr Coonan is now simply giving evidence to a witness who cannot say whether it is right or wrong.
THE LEGAL ASSESSOR: It seems that Mr Coonan has looked at the material contained in the bundles and has drawn his statistics from that which he has put to this witness. The witness has made his comment. If he has got it wrong and there are more references, there is re-examination and Miss Smith as two juniors who can check the material.
MS SMITH: Of course that would be the case if they were in the bundles but this is not documentation that is in the bundles; it is simply something Mr Coonan has got from somewhere and we do not know where.
THE LEGAL ASSESSOR: Mr Coonan, I understood you were making reference to the various parts of The Lancet we have here.
MR COONAN: No, it goes beyond that. I am asking the editor of this journal whether he is in a position to confirm or comment on this material that I have put to him. It is available to him. If what I am suggesting is wrong, then Dr Horton can come back and tell us.
A You are saying you want me to go away this evening and look through the first 26 issues of 1998 and check your 3,000 and 1,000 numbers? I can do that but I might not be in a terribly fit state to give evidence tomorrow morning if I have to.
Q It does take quite a few cups of cocoa to get through it.
A I am sure you did it yourself.
Q It is a very simple exercise. If Dr Horton does not want to deal with it, he does not want to deal with it. It is a perfectly proper exercise. I am putting some figures based on documents that he is the editor of. I am perfectly entitled to do that. I am not bound by what is in front of the Panel.
THE LEGAL ASSESSOR: As I understand the answer, the witness has said that if the figures put to him are correct, even so he has given his comments and caveats on the evidence. If anyone wishes to dispute the statistics, it is up to them. The witness has not accepted that you can necessarily draw the conclusions that Mr Coonan wants. I do not think there is any point in wasting any more time on this.
THE WITNESS: Just to underline, what you have done is give one half of an equation and that equation cannot be solved without the other half of the equation being disclosed.
THE CHAIRMAN: The witness also said, and correct me if I am wrong, that you are not sure whether the figures are right or not.
A No. If you are asking me to go away and check them.
THE CHAIRMAN: We are not.
A Then it seems to me that, without an independent assessment, it would be hard to judge.
MR COONAN: I do not want to cause unnecessary burdens to Dr Horton. I put what are the results of burning the midnight oil. He has made his comment and there it is. If Dr Horton, when he has a moment, as he will do during the rest of this case, tells me that those figures are completely wrong then I stand corrected.
A Can I invite counsel to submit to me the details of his analysis and the methods by which he came to his conclusions and the precise numbers, then I will be able to go away and check them.
Q That is a fair point.
THE CHAIRMAN: It is a fair point, but I am not sure if that is necessary.
MR COONAN: That is also a fair point.
A Again, I think this gets away from the point of this particular case. I wonder if I might have an opportunity to assist the Panel in trying to explain the mechanism by which we judged and, when all of the evidence was presented to us in early 2004, why we felt this was a discloseable conflict of the interest.
Q I know you have been eager to get there, so let us go there now. I will put this particular point to one side, but if we need to develop it later, then it is not an impossible route to address. Dr Horton, I have taken you through the literature to demonstrate the background leading up to the formulation, if you like, of the 1998 rubric in The Lancet which applied in 1998 – 1997 as well – and looked at how things have developed expressly for authors or would-be authors of The Lancet. It is quite apparent that when you discovered that there had in fact been Legal Aid funding at least for a study confirmed by Dr Wakefield – and we have been through that – there is no doubt that you expressed yourself on the basis that there should have been disclosure, not so much of the direct funding, because on Dr Wakefield’s account, there was not any. Is that right?
A Correct.
Q But of the perceived conflicts of interest that in your judgment as the editor had arisen and should have been disclosed.
A Correct.
Q We can see how you put it in fairness to you by looking at page 1211 in bundle 3. In the right-hand column, towards the bottom – we have looked at these passages already – in that last paragraph you summarise what it was that Dr Wakefield was saying about his role in this other study and you say:
“At the time of submission and eventual publication … this second study had not been disclosed to the editors of The Lancet and his co-authors. We judge that it should have been so disclosed, irrespective of the number of children overlapping between the pilot project funded by the Legal Aid Board and the Lancet paper.”
Then at the bottom of that paragraph, this is how you put it:
“We believe that our conflict of interest guidelines at the time should have triggered such a disclosure, including the fact that a significant minority of the children described in the Lancet paper were also part of the Legal Aid Board funded pilot project.”
Then, just to capture really your observations on this topic, on page 1212, in the left-hand column, the penultimate paragraph, you say:
“We regret that aspects of funding for parallel and related work and the existence of ongoing litigation that had been known during clinical evaluation of the children reported in the 1998 Lancet paper were not disclosed to editors. We also regret that the overlap between children in the Lancet paper and in the Legal Aid Board funded pilot project was not revealed to us.”
I hope that is a fair gathering together of how you put it in 2004. Again, at the risk of, as it were, overblowing this – I do not intend to, but it is important, or may prove to be an important factor – those utterances again were judgments made by you first of all in 2004. Is that right?
A Yes.
Q And they were made against the background of, as you described them, perceived conflicts of interest.
A Yes. Let us just put ourselves back to 1997, early 1998, as we were trying to do in 2004, to imagine what the situation was for Dr Wakefield as the lead author of this paper. May we do that?
Q What I am concerned about, though, doctor – I just enter a word of caution here that I might need to seek guidance from the Legal Assessor – is that you understand the difference between arguing a case and explaining the judgment you actually made at the time?
A Yes. I was simply trying to explain the judgment that we made at the time. I certainly do not want to argue a separate case from the one that we arrived at in early 2004, but I will be happy to be corrected on anything.
Q So far, I cannot object, if that is the basis on which you are seeking to come to an explanation of why you made that judgment. You made the judgment and there it is.
A I just want to be clear about how we came to that judgment because what we were trying to do was to take the available facts as they had been disclosed to us in early 2004 and to take ourselves back to 1997/early 1998 and think, if one tried to put oneself in Dr Wakefield’s position, how would one have thought about this potential conflict of interest and made a decision about whether or not to disclose or whether to disclose? It seemed to us – not only us, but to his co-authors and indeed to the Royal Free – that the case was compelling to disclose. The reasons for this were several fold. First, here was a study submitted to The Lancet, looking at a group of children with a new syndrome, but including testimony from the parents about a possible temporal linkage to vaccination. At the same time, Dr Wakefield was pursuing a separate study – nobody is saying they were the same – a separate study investigating whether measles vaccination was indeed tied causally to this syndrome. So the same person was involved in two studies, one for The Lancet, one for the Legal Aid Board. Irrespective of whether the one for The Lancet was funded by the Legal Aid Board or not, in the mind of the lead investigator, he was taking part in two studies that linked the vaccine in some way, shape or form to the syndrome.
It seemed to us that there should have been some cross talk in his mind about those two studies: first, that we should have been told about the Legal Aid Board study, second, we were also aware from the newsletters that we had seen that he was giving advice to Richard Barr, who seemed to be an important individual in litigation, advice where he was actually quoted in one of the newsletters talking about his views on the possible link between the vaccine and autism. If he is giving advice about a possible linkage to a solicitor involved in the litigation, that seemed to me, irrespective of the interesting discussion we have just had in the last half hour or so, that seemed to me in 1997/1998 something that absolutely should have been told to the editors of The Lancet. It was a fact absolutely material to understanding the pathology going on in these children. Absolutely material. Also, something he was perfectly – and I absolutely concede this – willing to disclose that four or five children happened to be shared between the two studies. Again, if you have individuals taking part in two separate studies in that shared way, that is something that should also have been disclosed as part of the study that was eventually published in The Lancet. So it seemed to us absolutely clear – there was no ambiguity for us – that that information should have been conveyed to the editors of the journal.
Q Dr Horton, that is your reasoning and indeed what you have said is in effect nothing new, because you had indicated even earlier in your evidence that the perception of a conflict of interest covered those elements which you have just again set out. So there is nothing new in effect to what you are saying. What it boils down to – and I have to be very careful here for you, with respect, not to trespass on the function of the Panel – the issue obviously in due course, which I am sure you recognise, will be: what were the underlying facts? That will be a matter for this Panel to determine and not, again with respect, yourself.
A I understand, of course.
Q Secondly, how Dr Wakefield himself approached this problem at the time in 1997, as opposed to on the other hand how you, quite respectfully and properly, viewed how he should have responded.
A Correct.
Q It may be quite one thing – and I say this with the greatest respect – for the editor of The Lancet to take a view that somebody should have done something and it may be quite another to judge in fact by an objective outsider whether in fact he should. Do you follow?
A I understand the point you are making, yes.
Q Because it may be that a host of other factors may have to come into play which, with the best will in the world, may not have been available to you at the time. Is that fair?
A It is quite possible. I do not know. I just wanted to draw us back to the real case, rather than the abstract debate which was taking place in journals over a 15-year period.
Q Those factors may well come back and have a relevance to the Panel later. That is why I have been using you, in I hope the best possible way, to demonstrate the point which I have to make. Ultimately, as you can appreciate, although that is the view you came to, the Panel will have to judge the extent to which Dr Wakefield fell into the category of people who did not actually apparently comply with the rubrics which other editors apart from yourself thought were unambiguous. Do you see the point?
A Yes, I do see the point.
Q That is ultimately an issue for this Panel.
A I understand that.
MR COONAN: Thank you very much, Dr Horton. That completes my cross-examination.
THE CHAIRMAN: It is now 4.10, Mr Miller. It has been a pretty intense day, not only for the Panel, but I am also considering Dr Horton.
MR MILLER: If I only had ten minutes of cross-examination, then I would ask to deal with it, but I have considerably more than that. I also have to put in some more documentation. It has been quite a dense afternoon.
THE LEGAL ASSESSOR: If you have documentation you would like the Panel to see, then perhaps that could be put in now.
MR MILLER: There is a letter and the response from the editor of The Lancet. They are numbered and can be inserted in the bundle. I can hand those up now. (Same distributed)
THE CHAIRMAN: Would it be appropriate for the Panel to read these now?
MR MILLER: If you would, sir. I could use the time certainly, as I do not want to cover the same ground as Mr Coonan, so I could consider the points he has made between now and 4.30, so the time will not be wasted.
THE CHAIRMAN: Thank you. We will go into camera for the next ten to 15 minutes to read these papers which have been circulated and then we will adjourn until tomorrow morning. Dr Horton, can I remind you again that you are still under oath and are going to remain under oath overnight, so please do not discuss this case with anyone.
THE WITNESS: Understood. Thank you.
THE CHAIRMAN: We will now go into camera for the next ten to 15 minutes and then we will adjourn until 9.30 tomorrow morning.
(The Panel adjourned until 9.30 on Wednesday 8 August 2007)
No comments:
Post a Comment