by Martin Hewitt
Three months after Brian Deer's first BMJ article (HERE) claiming Dr Andrew Wakefield fabricated the findings of the Lancet paper ('Ileal-lymphoid-nodular hyperplasia', Lancet, 1998; p.351), it is now time to revisit the claims and test them against the record, in particular the UK General Medical Council transcripts of the fitness to practice panel against Dr Andrew Wakefield and Professors John Walker-Smith and Simon Murch. We can then judge the accuracy, selectivity and interpretive license Deer applies to his evidence against Wakefield.
This is the first of a series of articles examining the sources Deer uses to back his claims. We start by examining Deer’s evidence for claiming that three of the nine children (child 6, 7 and 12) reported with regressive autism did not have an autism diagnosis. Future articles will examine the evidence for what is said about the other eight children.
The BMJ editors who commissioned Deer’s article continue to support the accuracy of his allegations. Do the editors consider that the GMC judgement of Wakefield means he now has no case to answer? If so, we should remember that Deer uses transcript evidence for a charge the GMC did not lay against Wakefield, namely that he fabricated his research findings.
What Deer says about each of the three children is examined. Each of Deer's claims followed by references word for word is placed are quoted. The claims are followed by other extracts from the transcripts and other sources underneath, together with an explanation of their significance. All emphases below are added to the original.
First, a general point about Deer’s evidence. As you see below, Deer uses testimony and documentation from local GPs and consultants that were not in most cases available to the Royal Free team when the Lancet paper was written. In the UK GP referral letters summarise the patient's cases without including documentation. Of the Lancet 11 (the twelfth was from the US and not included in the GMC case), only GPs for child 1 and 8 enclosed some documentation and child 2’s GP enclosed one letter from a child psychiatrist with the referral letter.
CHILD 6 (born April 1992; MMR 15 June 1993; Royal Free admission: 27 October 1996)
Deer claims that three – including child 6 – of the nine children described in the Lancet paper as having regressive autism were neither autistic nor regressive.
“But only one—child 2—clearly had regressive autism.(41) Three of nine so described clearly did not. None of these three (children 6, 7 & 12) even had autism diagnoses, either at admission or on discharge from the Royal Free….Both [brothers 6 & 7] had histories of fits and bowel problems(42) recorded before their MMR vaccinations.”(43) (44)
41. Michael Rutter. Evidence to the panel. Day 37. Day 39
42. Dr N. GP records. Day 6.
43. Dr N. GP note. 18 March 1993. Day 6. The record says child 6 received MMR on 15 June 1993.
44. John Trounce. Report. June 1995. Day 6. Child 7 received MMR on 24 November 1995.
Deer writes: “Child 6 aged 4 years at admission, had Asperger’s syndrome,(45) which is distinct from autism under DSM-IV, is not regressive,(46) and was confirmed on discharge.”(47)
45. John Walker-Smith. Letter to Andrew Wakefield. 4 October 1996. Day 41.
46 Filipek PA. Autistic Spectrum Disorders. In Swaiman KF, Ashwal S. Pediatric neurology:. 3rd ed. Mosby, 1999.
47. Panel findings of fact. p.25
Deer, with no medical qualifications to his name, contends that autism and Asperger’s are distinct, relying on the 1994 Diagnostic and Statistical Manual of Mental Disorders IV (DSM-IV) where the two conditions are given separate entries in the section on pervasive developmental disorders.
DSM-IV describes differences between Asperger’s and autism in one of the triad of autistic features, namely language acquisition, but similarities in the other two, in abnormal social interaction and restricted activities and interests. Overlapping features for impaired social interaction in autism and Asperger’s are marked by the use of almost identical paragraphs in the two entries in DSM-IV (pages 70 & 80). It adds that “Differentiation of the two conditions can be problematic in some cases” (p.83). So no clear differentiation of Asperger’s from autism. But Deer treats the two conditions as unquestionably distinct.
Deer also supports his claim by referring to a source published after the 1998 Lancet paper, and so not available to the authors, namely the chapter referenced in (46) above. The more recent 2006 version of this chapter clearly refers to an overlap – not difference – in the criteria for autism and Asperger’s: "The diagnostic term [Asperger's] was included for the first time in DSM-IV (1994), and the criteria for the qualitative impairments in social interaction and restrictive and repetitive patterns of behaviours and activities are identical to those for autistic disorder....The validity of Asperger's disorder as a definitive diagnostic entity separate from higher-functioning (verbal) children with autism remains controversial" (p.909 in Hirtz DG, Wagner A & Filipek PA, 'Autistic spectrum disorders', in KF Swaiman, S Ashwal & DM Ferriero (2006) Pediatric Neurology: principles and practice, 4th ed., Philadelphia: Mosby).
As for Asperger’s being non-regressive, the 2006 chapter does not mention non-regression in relation to Asperger’s. DSM-IV states that there are no delays in early language acquisition and cognitive development (two of the triad features) in the Asperger's child. However, evidence of developmental delay is not the same as regression following normal development.
CHILD 7 (born February 1994; MMR November 1995; admitted Royal Free 2 September 1996)
Deer writes: ["But only one—child 2—clearly had regressive autism.(41) Three of nine so described clearly did not. None of these three (child 6, 7 & 12) even had autism diagnoses, either at admission or on discharge from the Royal Free….] Child 7 was admitted at nearly 3 years of age without a diagnosis,(48) and a post-discharge letter from senior paediatric registrar and Lancet coauthor David Casson(49) summarised: “He is not thought to have features of autism.”(50)
48. Panel findings of fact. p 93. “On 15 January 1997 Child 7 attended an outpatient consultation with you [Walker-Smith] during which you elicited a history of intermittent episodes of passage of blood associated with constipation and alternating diarrhoea with mucous. You did not undertake an abdominal x-ray to confirm whether or not constipation was the primary cause of Child 7’s symptoms.”
49. Repeated, multiple comparisons between Casson’s documentation shows an exceptional degree of concordance with clinical records generated by consultants.
50. David Casson. Letter to Sian Bennett. May 1997. Day 102
Deer claims, as though problematic, that child 7 was admitted to the Royal Free without a diagnosis. But the reasons for the referral to Walker-Smith, or any other expert consultant, would be because the GP was unable to diagnosis the symptoms and sought expert advice on diagnosis and treatment. Interestingly the panel’s view in (48) that Walker-Smith should have X-rayed the child, ignored, without giving a reason, the evidence of Professor Walker-Smith’s expert witness on paediatric gastroenterology, Dr Miller. He said. ‘I do not think that x-ray is mandatory in that sort of situation (Day102, p.46).
Again the transcripts show that at admission to RFH, the GP says that "[child 7] himself probably does not have autism although this is not certain at present but he does have convulsions which I believe may make him eligible for your study. He also suffers with bowel problems similar to his brother who is autistic" (GP's letter to Walker-Smith, December 96, Day 105, p.46).
Further uncertainties about child 7’s behaviour prior to admission are seen in the following exchange between the prosecuting counsel and the GP: “Q Was it your view at the time – and I do appreciate you are talking as a general practitioner – that he was autistic? A I think ‘autism’ is probably too strong a word. I felt he had a social disorder of some sort, yes” (Day 6, p.16).
On discharge, Dr Casson’s letter, May 1997, to Dr Bennett, the consultant community paediatrician who saw child 7 (see footnote 50), says, “He is not thought to have features of autism.... It was decided in view of the findings in his brother, to investigate [child 7] further.” (Day 102, p. 50). In other words, a sufficient reason for admission was the possibility that he had similar conditions to his brother, child 6.
During 1997 child 7 continued to see the psychiatrist member of the Royal Free team. The GP’s referral letter, September 1997, to Dr Gillian Baird, developmental neuropaediatrician, Guys’ Hospital, recalls that "[Mrs 7] has also been seen by the Psychiatrist up there [Royal Free] who tells her that [Child 7] is autistic" (Day 6, p.21). Dr Baird’s report, September 1998, read to the hearing, said, “As has been previously suggested, this pattern is that of an autistic disability...his problems are best described as being due to a Pervasive Developmental Disorder in the Autistic Spectrum" (Day 6, p.23).
CHILD 12 (born December 1990 ; MMR 6 March 1992; Admitted to Royal Free 18 July 1996 and 5 January 1997)
Deer writes: “[None of these three (child 6, 7 & 12) even had autism diagnoses, either at admission or on discharge from the Royal Free …] [Child 12] was aged 6 at admission and had previously been assessed for possible Asperger’s syndrome at Guy’s Hospital, London, by a renowned developmental paediatrician.”(52) (53)
52. Andrew Wakefield. Letter to Mrs 12. 19 July 1996. Day 28
53. Gillian Baird, at the Newcomen Centre, Guy’s hospital. September 1996. Day 36
At admission, child 12 had an Asperger’s diagnosis, as Deer says. However, Deer’s view is that Asperger’s is ‘distinct from autism’ and is not part of the autism spectrum (see above).
But looking at the record Deer doesn’t quote, we see that on 6 January 1997 the admission record says, “Admitted for investigation of autism and bowel problems” (Day 36, p.58).
On discharge, Dr Berelowitz's report was, according to Professor Rutter’s evidence, “in keeping with the Newcomen Centre's much fuller appraisal.” He added, “Asperger's Syndrome, as we have discussed, is a milder variety on the autism spectrum” (Day 36, p.63). Rutter was the prosecution expert witness on autism.
The implications of checking the original sources are that the 1998 Lancet research was justified in including Asperger's children in the case series of 12 autistic children. Deer fails to produce definitive evidence that the three children did not have a form of regressive autism. There is nothing in the evidence we have seen so far that warrants the charge that Wakefield fabricated this finding in relation to the three children given in the Lancet paper.
We see clearly how Deer works. First he selects extracts from the transcripts to build up his case, ignoring other extracts that might cast doubt on his case such as the above extracts. Secondly he applies his reasoning rigidly to the evidence as though medical diagnosis is about 100% certainty, for example that there is a clear indisputable difference between autism and Asperger’s. On this basis no doctor’s practice would survive this standard of certainty. Deer targets Wakefield. But were any other doctors’ or scientist’s practice to be subjected to this standard, it would fare as unjustly as Wakefield’s. This is not only a travesty of how medicine works in reality; it is a disservice to medical science and clinical practice. Medicine is not an exact science or an exact clinical practice.
The question remains as to why the editors of the internationally renowned BMJ commissioned, peer reviewed and published Deer’s article. They continue to support its accuracy.
We will leave the final word to Dr Victor Miller, Professor Walker-Smith’s expert witness, when asked by Walker-Smith’s counsel: are you allowed to have hunches? His reply: “I think medicine is often about serendipity” (102, p.15). Deer, the BMJ editors and the GMC should take this to heart.