This is the third part of an analysis of Brian Deer’s (in photo) claims that Dr Andrew Wakefield fixed the findings in the 1998 Lancet paper 'Ileal-lymphoid-nodular hyperplasia' (Lancet, 1998; vol. 351, p.637), a small clinical case series of 12 patients with autism and bowel disease treated at the Royal Free Hospital in 1996/97. The British Medical Journal editors not only published Deer’s claims on 5 January 2011 [ HERE ], but backed them with the charge of academic fraud against Wakefield. Using the UK General Medical Council (GMC) transcripts and research papers, Parts 1 and 2 questioned the way Deer selected his evidence to support serious allegations of fraud.
To substantiate scientific fraud the accuser must show that the original findings were changed – ‘fixed’, ‘falsified’, ‘fabricated’ in the BMJ’s words – to advance a particular interest or theory that the original findings do not support. These analyses of Deer’s claims test if this is so. Part one examined Deer’s claim that three of the nine children reported with regressive autism did not have an autism diagnosis [ HERE ]. Part two examined the claim that, “despite the paper claiming that all 12 children were ‘previously normal,’ five had documented pre-existing developmental concerns” and could not have regressive autism [ HERE ]. Deer used GP records and local consultants’ letters from the early 1990s that were not available to Wakefield and his team when investigating the Lancet paper children in the mid-1990s, and which only became available 16 or so years later to the GMC fitness to practice hearing against Dr Wakefield and Professors John Walker Smith and Simon Murch sitting between 2007 and 2010. Of the volumes of documents used as evidence in the GMC hearing, the Royal Free team in the mid-1990s would have only the following: the General Practitioner’s (GP) referral letters summarising each child’s case; four consultants’ letters which two GPs appended to their referral letters; and the results of their own clinical investigations at the Royal Free. The two analyses of Deer’s evidence showed that he selected extracts from the GMC transcripts that appeared to support his claims but failed to examine statements from the GMC hearing that countered his claims. That medical clinicians disagree over the interpretation of symptoms – as the GMC hearing showed in abundance – does not constitute fraud. In other words the first two analyses of Deer’s evidence found no evidence of Wakefield’s alleged fraud.
The third and present part examines Deer’s claim that “In nine cases, unremarkable colonic histopathology results—noting no or minimal fluctuations in inflammatory cell populations—were changed after a medical school ‘research review’ to ‘non-specific colitis’”. Seen in the context of an article claiming Wakefield fixed his results, this claim suggests further fixing took place by colleagues, not just Wakefield, attending the review. However, as discussed below, the review was a means whereby clinicians reached agreement on the different interpretations of the children’s bowel condition. This claim is examined in the light of the published GMC transcripts of the hearing against the three doctors. Child 11’s case is the exception – a private patient from the US, his case was discussed only in passing and so cannot be independently verified from the GMC transcripts. So nine cases becomes eight. We can judge the accuracy, selectivity and interpretive license Deer applies to his evidence against Wakefield. His claims that Wakefield fixed his results are summarised in a list of bullet-points at the end of his BMJ article.
Deer’s article provides little in evidence to back up his third claim. Of the nine children, only child 3 is singled out as “a prime example” of a bowel disease case. No evidence is provided in the body of the article on the other children’s bowel condition. We examine what Deer says about this child before commenting on the other seven children. The procedure is simple: Deer’s quotes and his endnotes referencing GMC and other evidence are given in italics followed by other statements of evidence from the GMC ignored by Deer. We end up with two lists of statements from the GMC: one from Deer supporting his allegations and the other from GMC witness statements countering his claims – both lists containing matching statements that contradict each other. Of course disputation is the stuff of scientific debate and the driver of scientific progress. But at no point can disputed statements count as evidence of fraud. They cannot stand as the last word on whether Wakefield fixed his results when other statements by medical clinicians and experts to the GMC hearing undermine the thrust of his allegations of fraud – a charge the GMC did not level against Wakefield or the other two doctors.
This is what Deer says of child 3.
Quote 1 “What was unquestionably true was that child 3 had serious bowel trouble: intractable, lifelong, constipation. This was the most consistent feature among the 12 children’s symptoms and signs but, being the opposite of an expected finding in inflammatory bowel disease, it was nowhere mentioned in the paper.
108 a) Ajjegowda Shantha. GP records and evidence to the panel. Day 5. (b) John Walker-Smith. Evidence to the panel. Day 91.
109 Murch S, Thomson M, Walker-Smith J. Author’s reply [letter]. Lancet 1998; 351:908
110 Squires RH and Colletti RB, Indications for pediatric gastrointestinal endoscopy: A medical position statement of the North American Society for Pediatric Gastroenterology and Nutrition. Journal of Pediatric Gastroenterology and Nutrition 1996; 23: 107-110.
111 Ian Booth. Evidence to the panel. Day 41. “Looking for inflammatory bowel disease would be a most unusual way of approaching a patient with severe, long-standing constipation.”
Deer’s central contention is that, although the children, including child 3, may have displayed bowel symptoms, none had inflammatory bowel disease (IBD) as claimed in the Lancet paper, and that at worst they had constipation which Deer claims was not consistent with IBD. However the lack of mention of constipation in the Lancet paper is evidence of its initial and provisional status as a small case series – which the Lancet called an ‘Early Report’ – which presents initial findings that subsequent more rigorous research sustains or refutes. As the research developed after 1998, further papers by Royal Free team members referred to constipation in the context of IBD, supporting contradictory findings that challenge what Deer says about the children’s bowel ‘trouble’:
Regarding endnote 110, the protocol was issued in 1996, prior to the above research, and considers the use of endoscopy in general paediatric gastroenterology. It does not address the use of endoscopy in diagnosing gastroenterological problems in autistic children.
Regarding 111, the above research into the presence of constipation in the context of IBD suggests a different view to Professor Booth's. Further, the following exchange took place at the GMC hearing between Walker-Smith’s counsel and Dr V Miller, defence expert witness in gastroenterology: "Q Staying with constipation, do you come across constipation as a feature of inflammatory bowel disease? A Absolutely, yes” (Day 101-38).
So the 1998 paper was the first of several papers on autism and bowel disease from members of the Royal Free team which subsequently went on to develop more rigorous findings, including the co-presence of constipation and IBD and the significance of the children’s histopathology.
Deer goes on to argue that, following the initial histology reports, the bowel findings were changed by a medical school “review” – a statement suggesting that, if the changes amounted to ‘fixing the data’, the responsibility lay with participating colleagues seeking agreement on conflicting histopathology readings. Responsibility for these changes cannot rest on Wakefield alone, contrary to what the BMJ editors’ contend in their accompanying editorial to Deer's piece.
Quote 2: as I revealed in the BMJ last April, the hospital’s pathology service found the children’s colons to be largely normal, but a medical school “review” changed the results.
115 Deer B. Wakefield’s “autistic enterocolitis” under the microscope. BMJ 2010; 340:838-41.
What Deer’s article does not reveal are the purpose, rigour and exemplary professionalism of these reviews that “changed results”. Without more exact procedures to determine medical findings associated with autism and bowel disease, early findings must inevitably rely on individual interpretations, with their greater risk of error and inexactness. Yet these same moves towards accuracy, when subject to critical peer review among scientists and clinicians, play an important role in reaching greater accuracy in interpretation. In the case of histological readings of bowel samples, different histopathologists can arrive at different readings. Given the critical importance of these findings for the outcome of the patient’s clinical care, a process of checking and debating different views is imperative. With this in mind Professor Walker Smith instituted two new practices to review the biopsies of children treated by his team, including, but not limited to, the Lancet 12: first, weekly clinical reviews of the histological findings attended by general histopathologists who had produced the initial reports, an expert in bowel histopathology, the appropriate clinicians and medical scientists; secondly, a blinded review of the initial histology findings conducted by the expert bowel histopathologist (the GMC panel did not examine the blinded reviews). The transcripts record the added rigour the weekly reviews imposed, in subjecting initial reports, constrained by time and resources as witnesses indicated (see eg Day 33-3), to close scrutiny and a deliberative process of reaching consensus on their significance. Work pressures on the hospital pathology department meant that the Lancet children were one small group among 700 or so cases per week received from different departments (Day 3 – 36).
Apart from child 3, Deer provides no evidence in the body of his article to support the claim that “unremarkable colonic histopathology results” were changed for the other eight children. The nine children can only be identified from the table attached to the article, where under a column headed ‘non-specific colitis’ he compares what was said in the Lancet paper for each of the 12 children with what he claims the records show, placing a ‘Yes’ or ‘No’ against each child to substantiate or refute the Lancet data. 'No’s' are placed against child 3, 4, 5, 7, 8, 9, 10, 11 and 12. Whereas the Lancet said that these children exhibited findings of non-specific IBD, Deer claims these children showed “unremarkable” histopathology results that were subsequently changed.
Given the significance of this claim for his general claim that Wakefield fixed the Lancet paper results, it is surprising that this evidence is not included in the article itself (apart from child 3’s). Deer confines the evidence to a different paper which readers of his BMJ article may have missed. A link ‘web extra’, to a ‘background document’ with Deer’s data, is found on the left column of the article’s webpage. This goes to a pdf document – each page headed with Deer’s own web address – on a separate part of the BMJ’s website. The ‘web extra’ provides two tables separating out the content from the single table in the article into Table 1, what the Lancet paper says, and Table 2, what Deer says. In Table 2, the column refuting the existence of non-specific colitis has ‘no’s’ and a footnote against each of the nine children. The footnotes cite the first histology reports on the children’s biopsies (small and large bowel biopsies numbered I, II, III, etc) produced by the Royal Free’s general histopathologists, other reports and Deer’s comments. No reference is given to the findings of the later blinded readings by the expert bowel histopathologist and scant reference to the more rigorous analyses of Walker-Smith’s weekly reviews. Important parts of the Royal Free team’s research procedures and findings are missing.
If the initial histology reports written by the Royal Free’s team of general histopathologists, which Deer quotes in his footnotes to Table 2, are compared with the subsequent data from the expert histopathologist in gastroenterology available to the GMC, we see markedly different readings regarding the presence and extent of inflammation in the children’s bowel. We will look in detail at the matching reports for child 3 first and then, to save time, compare short extracts from the matching reports for the other seven children (excluding 11). Bold text is used throughout to emphasise conflicting statements in the matching reports.
Quote 3: [Footnote] 56. Histology report. ‘I. Shows small bowel mucosa with an increase in intra-epithelial small lymphocytes, but no architectural abnormality. II. Shows small bowel mucosa with prominent lymphoid follicles. III-VII. Each show large bowel-type mucosa within normal histological limits.’ The pathology service comments: ‘Mild inflammatory and reactive changes in the small bowel samples, of uncertain significance on morphological grounds alone. No microbes or granulomas identified in any of these samples.’”
The initial histology report for child 3 – in common with initial reports for the other 8 children – identifies signs of inflammation which are considered unproblematic. By contrast child 3’s discharge summary written by Dr David Casson (a member of the Royal Free team and co-author of the Lancet paper) on 16 September 1996 draws on the further histopathology review of the results and stresses the pathology of inflammation in the child’s bowel:
“Overall appearances were normal until the ascending colon. Here one definite apthoid ulcer was seen and towards the caecum there were multiple prominent colonic lymphoid follicles. The terminal ileum also appeared abnormal showing marked lymphonodular hyperplasia though there was no ulceration. Histology possibly demonstrates mild chronic inflammation within the lamina propria of the terminal ileum. It should be noted that it is difficult to estimate whether or not this is within normal limits. … Throughout the large bowel there was a patchy increase in chronic inflammatory cells with an occasional prominent lymphoid follicle with a germinal centre. There was also an occasional focus of acute cryptitis within the ascending colon. There was also mild crypt distortion. …. The patchy distribution of this inflammation and the involvement of the terminal ileum are in keeping with a diagnosis of Crohn’s disease.”
Subsequently the diagnosis of Crohn’s was replaced by ‘non-specific’ or ‘indeterminate colitis’, reflecting the limited knowledge in the paediatric gastroenterology community in the mid-1990s of how to read bowel symptoms in autistic children. Professor Walker-Smith explains his logic at the time: “I believed these children had significant inflammatory bowel disease but it was not Crohn’s disease and was not ulcerative colitis. That is why the actual words, ‘indeterminate colitis’ were used here” (Day 74-42-3).
Quote 4: 60. Histology report. “I. Small bowel type mucosa with a lymphoid follicle. II-VII. Large bowel mucosa, some with attached muscularis mucosae, with no evidence of architectural distortion or increase in inflammatory cells in the lamina propria. Lymphoid follicles with germinal centres are present in many of the biopsies. No cryptitis or crypt abscesses are seen. The surface epithelium appears intact. No granulomas, ova or parasites are seen.” The pathology service comments: “Large bowel series with terminal ileum, with no histopathological abnormality.”
Comparing this initial report with the following exchange between Walker Smith and his QC, we see why he took the subsequent results more seriously and why he placed more credence on the reports of the expert bowel histopathologist, Dr Dhillon:
Q Just take as an example – we could do the exercise and I am going to do it in due course when we go through the individual cases, but in some cases Dr Dhillon appears to have found a description of the pathology there that is not present in the original report.
A There are differences. There are some children where there are differences, yes.
Q Taking as an example ... Child 4...You have written there....
A It says, “Lymphoid nodular hyperplasia. Dhillon differs – 2/5 Paneth cell metaplasia. Increased eosinophils 5/5 mild increase in inflammatory cells”, And then I have written after that, “routine – normal”, rather as an aide memoir to myself saying that the routine report had been reported as normal....
Q ....I asked you whether the Dhillon findings were the same as the routine, were they in all cases the same as the routine?
A ....They were not the same in all cases, no.
Q Taking that as an example, you have got “Dhillon differs” – you have written that to yourself.
A I have.
Q What did you mean by that?
A He considered that there was more significant histopathological abnormality than the routine report. Obviously the routine report says it is normal and he, for example, says, “Mild increase in inflammatory cells”. I agreed with that, otherwise I would not have written it.
Q And 2/5, what does that mean?
A That is the two out of five paneth cells there was metaplasia
Quote 5: 64 Histology report. “Specimen I consists of fragments of small intestinal mucosa which includes lymphoid follicles but which is without pathological abnormality. Specimens II, III and IV are large bowel mucosa fragments with normal crypt architecture. There is at best a minimal increase in chronic inflammatory cells within the superficial lamina propria. No active inflammation is seen. Specimens III and IV show minor crypt architecture distortion, including occasional bifid forms. Paneth cell metaplasia is not seen. No excess chronic inflammatory cells are seen. A very occasional polymorph is seen within surface crypt epithelium. No ova, granulomas or parasites are seen in any of these biopsies.” The pathology service comments: “Large bowel series; minor changes the significance of which are uncertain but do not amount to the diagnosis of inflammatory bowel disease.” Also of interest: child 5 was reported by endoscopist and Lancet co-author Simon Murch as not having ileal lymphoid hyperplasia, but the published Lancet paper said that he did have this.
Child 5’s discharge summary which the Royal Free sent to the GP is referred to on Day 11 (p.52): “he has apparently been having bouts of diarrhoea on and off since he was 4 years of age. There is no blood or mucous in the stool. He has occasional abdominal pain during which he suddenly clutches his abdomen and flexes his knees. This has a duration of about 10 minutes and then resolves.
The histology report in footnote 64 is discussed in Mr Miller's examination of Professor Walker-Smith (Day 78-45/6), which again provides insight into how he read the histology report in a way significantly different from Deer.
Q I would like your comments on the findings first.
A Specimen I is in accord with the endoscopic appearances. The II, III and IV, “there is minimal increase in chronic inflammatory cells within the superficial lamina propria”, that is a definite abnormality that suggests chronic inflammation, but the adjective “minimal” is used and there is “no active”, that means acute, inflammation. III and IV says “minor crypt architecture distortion including occasional bifid forms”. That clearly indicates that there has been previous damage from inflammation, damage in crypt architecture, and the crypts have responded to that abnormality by, instead of having a single form, they have developed a double or bifid form.
In III and IV the opinion was that there was no excess chronic inflammatory cells and a very occasional polymorph was seen within the surface crypt epithelium. That is certainly abnormal, which is an indicator of some acute inflammation.
Quote 6: 72 Histology report. “I-II. Two pieces of small intestinal type mucosa with essentially normal villous architecture. There is no increase in inflammatory cells in the lamina propria or in intraepithelial lymphocytes. Part of a lymphoid follicle is included. No parasites or granulomas are identified. III-IV. Sections from all sites show large bowel mucosa with no abnormality of crypt architecture or significant increase in inflammatory cells in the lamina propria. Some of the biopsies contain lymphoid follicles. No granulomas or parasites are seen.” The pathology service comments: “Small bowel biopsy and large bowel series without significant histological abnormality.”
The colonoscopy report showed “Slight evidence of vascular abnormality in rectum and sigmoid but otherwise essentially normal. The terminal ileum however demonstrated a marked degree of lymphonodular hyperplasia” (Day 1 – 57). Commenting on his post-admission letter to Child 7’s GP, when examined by GMC QC Ms Smith, Walker-Smith points out: “It was an interesting letter from my perspective because clearly something had been found that was considered significant”, namely the finding of evidence of bowel inflammation and indeterminate colitis” (Day 6 – 31). Again Walker-Smith, drawing on his long-standing experience and expertise in gastroenterology, placed greater significance on the presence of lymphonodular hyperplasia.
Quote 7: 76 Histology report. “I. These are both fragments of poorly orientated, but normal small bowel mucosa. A lymphoid reactive centre is seen in each sample. II-IV. These are all pieces of normal colonic type mucosa containing occasional lymphoid aggregates. Minimal inflammatory changes may be the result of operative artefact.”
Following GP treatment for constipation, child 8 was admitted to her local hospital after she became unsettled during the night, was screaming for a week, had diarrhoea (“6 bowel motions/day) and complained of abdominal pain. On arriving at the Royal Free, the question for Professor Walker-Smith was whether the histology report indicated the presence of inflammation in the bowel. In his examination by Ms Smith, he rejects the general histopathologist’s suggestion that signs of inflammation were artificial due to problems in extracting the biopsy (an ‘operative artefact’), rather than evidence of real inflammation, because there was insuficient time for the small bowel specimens to become inflamed after removal (Day 80 – 37). Walker-Smith concludes that it is “Not Crohn’s or ulcerative colitis, no, but the fact reported there is “minimal inflammatory changes”.
When Mr Miller QC questions Dr Miller, the defence expert witness in gastroenterology (Day 103-6), about the presence of inflammation in child 8’s bowel, he replies, “No, it is not entirely normal. It is not, frankly, grossly abnormal, but a lymphoid reaction is implying, as we have discussed in previous patients, that there is some overactivity of the lymphoid tissue, thereby implying some degree of inflammation.”
Quote 8: 80 Histology report. “I. Small bowel mucosa showing no histological abnormality. II-VII. Large bowel mucosa showing prominent lymphoid follicles but no histological abnormality.”
Again at variance with the general histopathologist’s original report above, the colonoscopist’s report found “a marked increase in the size and number of prominent lymphoid nodules”, which Walker-Smith relayed to the GP, adding that, “The colon was endoscopically normal except for an area at the hepatic flexure which was slightly erythematous. Histologically there was an increase in chronic inflammatory cells throughout the colon with a moderate increase in intra-epithelial lymphocytes.... Our diagnosis is indeterminate colitis with lymphoid nodular hyperplasia” (Day 4-13).
Later Dr Miller, when examined by Mr Miller QC, explained that large bowel mucosa showing prominent lymphoid follicles was “...an unusual feature because lymphoid hyperplasia...is normally in the terminal ileum” (Day 103 - 16).
Quote 9 : 84 Histology report. “I. The specimen consists of small bowel and have sampled a Peyer’s patch. Where present, the overlying villae appear unremarkable. The lymphoid tissue shows reactive changes. Parasites and granulomas are not seen. II-VI. All these biopsies show large bowel mucosa with occasional isolated bifid glands. The inflammatory population is within normal limits. Parasites and granulomas are not seen.” The pathology service comments: “No significant histological abnormality.” A supplementary report was requested, following a weekly histology meeting with clinicians, and possibly with Wakefield. This said: “These biopsies have been reviewed following a clinicopathological meeting. The ileal biopsy shows confluent lymphoid aggregates within otherwise unremarkable small intestine. The large bowel biopsies show a very subtle scattering of chronic inflammatory cells within the lamina propria. The superficial lamina propria contains focal nuclear debris and the surface epithelium appears slightly degenerate. No active inflammation is seen. More levels have been cut and no granulomas have been identified.” The pathology service comments: “Minor abnormalities. ? Significance.” Child 10’s biopsies were also examined at another centre. Following the supply of samples to the University Hospital of Wales, Cardiff, Huw Jenkins, consultant paediatric gastroenterologist, wrote: “I’ve now had a chance to review 10’s intestinal biopsies kindly sent down from the Royal Free hospital, and although there are lymphoid follicles present in the small intestine these are often regarded as a normal finding, and certainly our pathologists here would suggest that the colonic biopsies were within normal limits. Certainly they do not feel he has good evidence of gut inflammation in the biopsies.”
The long footnote records a process of uncovering problematic signs in child 10’s bowel which are not initially understood. First, the general histopathologist’s report identifies reactive changes in the lymphoid tissue, but no significant histological abnormality. Secondly, the biopsies were reviewed, which confirmed the initial findings, but uncertainty remains over their significance. Thirdly, but absent from the footnote, Dr Murch’s colonoscopy report (Day 25 - 59) states unambiguously, “This colonoscopy was definitely abnormal, in probably a more striking example of the pattern seen in the cohort of the autistic children. The rectum showed definite mild abnormality, with a slightly granular mucosa and abnormal vascular pattern. Prominent lymphoid follicles could be seen throughout the colon, with no other mucosal abnormality. The caecum showed an eryhthematous, granular mucosa around a swollen ileo-caecal valve, while the terminal ileum showed minor inflammatory change and striking lymphoid hyperplasia distally. I suspect that the biopsies will show unequivocal abnormality!”
When these concerns were placed side by side with the University Hospital of Wales report, which viewed the biopsies as unproblematic, the Royal Free clinicians had to decide whether child 10’s inflammation was pathological and required treatment. Presumably because of their experience and duty of care, the team decided that treatment was warranted.
Quote 10: 92 Histology report. “I-IV. Pieces of large bowel mucosa including lymphoid follicles with germinal centres. There is no architectural abnormality and no increase in inflammatory cells in the lamina propria. No organisms or granulomas are seen.”
Again in the hands of an expert gastroenterologist, the unremarkable is viewed with more concern. Dr Murch finds “Appearances almost normal to caecum. Again there were minor changes in the rectum and caecum (slight changes in vascularity and prominent lymphoid follicles). The ileo-caecal valve could not be identified”. Pieces of large bowel mucosa including lymphoid follicles with germinal centres...” (Day 25 - 63). Examined by Mr Miller QC, Professor Walker-Smith summed up his conclusion that “the finding of lymphoid follicles, especially with germinal centres, in the large bowel mucosa is an abnormal finding” (Day 82 - 21).
This study of the GMC transcripts shows that Deer has constructed his allegations of fraud against Wakefield in two ways. First he redefines what constitutes fraud. Academic fraud is no longer confined to researchers who deliberately change their original data in publications promoting scientific preconceptions the original data do not support. According to this conventional view, the accuser would have to compare only the researchers’ own original data with the published data. For Deer fraud is committed and substantiated if there is, besides the original research data, additional data the researchers did not have access to that is different from the published data. We have seen that Deer either did not have access to or did not use all the original data generated by the Royal Free team, such as the results of the blinded histopathology reviews and the weekly reviews. But we have also seen that Deer makes considerable use of additional information, such as GP and local consultant records which the Royal Free team would not be aware of at the time the case series was conducted. For example, the first and second analyses of Deer’s article showed that he drew on GP and local consultant records, most not available to the team, that appeared to show that three of the children were not autistic and that five (including child 11) had pre-existing developmental concerns before and not after the MMR. The analyses used other statements, some from the same witnesses, to show that the children were indeed autistic and that the pre-existing conditions were not relevant to the conditions that did emerge after MMR (except for one child who had the measles jab before his MMR). Of course these different views may be matters for further debate – though relating to records taken in some cases 20 plus years ago. But the fact that debate is possible means these are not matters of indisputable truth, against which differing findings can be viewed as fraudulent.
The second way in which Deer constructs his case for fraud is by selecting statements from the GMC hearing in such a way that differences of interpretation and matters of legitimate scientific debate are foreclosed to produce the impression that the selected statements stand as the only account available. For example, when we compare Deer’s evidence with other statements from the GMC transcripts, we see that there were differences in the conclusions clinicians reached on the significance of the children’s bowel symptoms, specifically regarding the evidence of inflammation. But rather than acknowledge the normal scope of interpretative variation in histopathology findings, Deer provides the reader only with the ‘unremarkable’ accounts of the initial histopathology reports in seeking to undermine the histological evidence of ‘non-specific colitis’ in nine of the 12 children summarised in the Lancet paper.
Deer’s reconstructions of scientific fraud – of the definition and data determining fraud – are endorsed by the BMJ editors in an editorial introducing his article. The editors allege that “Wakefield altered numerous facts about the patients’ medical histories in order to support his claim to have identified a new syndrome”. Yet not one example of an original fact altered to fit a preconceived syndrome – so constituting fraud – is given by Deer or the BMJ. Discrepancies between the Lancet published data and other data refer in the main to data the Lancet team did not see because they were not part of the original findings. Where data were changed, this was part of the routine processes of reviewing data to seek greater factual accuracy and of reaching a consensus on different interpretations. Not only Deer but more surprisingly the BMJ editors represent the routine but complex processes of making clinical judgments as fixing the data - a travesty of the clinical process.
Deer’s binary view of the truth – where this is no room for interpretation, debate over gradations of meaning, and for reaching consensus – obscures the true nature of clinical investigation and the process of medical advance.