The Wakefield MMR verdict

January 28th, 2010 by Ben Goldacre in bad science, MMR | 135 Comments »

Here’s a very brief piece I bashed out for the Guardian newsdesk today on the Wakefield finding, the further reading below will be more helpful if you’re interested in the story.

Ben Goldacre, The Guardian, Thursday 28 January 2009

In medicine, “untoward incident inquiries” tend to look for systems failures, rather than one individual to blame.

It’s certainly clear that Andrew Wakefield and his co-defendants failed to meet the high standards required of doctors in research. The GMC have found he was “misleading” “dishonest” and “irresponsible” in the way he described where the children in the 1998 paper came from, by implying that they were routine clinic referrals. As the GMC have also found, these children were subjected to a programme of unpleasant and invasive tests which were not performed in their own clinical interest, but rather for research purposes, and these tests were conducted without ethics committee approval.

These tests were hardly trivial: they included colonoscopy, where the child is sedated, and a long tube with a camera and a light passed through the anus and deep into the bowell; lumbar puncture, where a needle is placed into the spine to get cerebrospinal fluid; barium meals and more. It’s plainly undesirable for doctors to go around conducting tests like these on children for their own research interests without very careful external scrutiny.

But there is the wider context: Wakefield was at the centre of a media storm about the MMR vaccine, and is now being blamed by journalists as if he were the only one at fault. In reality, the media are equally guilty.

Even if it had been immaculately well conducted – and it certainly wasn’t – Wakefield’s “case series report” of 12 children’s clinical anecdotes would never have justified the conclusion that MMR causes autism, despite what journalists claimed: it simply didn’t have big enough numbers to do so. But the media repeatedly reported the concerns of this one man, generally without giving methodological details of the research, either because they found it too complicated, inexplicably, or because to do so would have undermined their story.

As the years passed by, media coverage deteriorated further. Claims by researchers who never published scientific papers to back up their claims were reported in the newspapers as important new scientific breakthroughs, while at the very same time, evidence showing no link between MMR and autism, fully published in peer reviewed academic journals, was simply ignored. This was cynical, and unforgivable. Then, after Tony Blair refused to say if his son had received the vaccine, the commentators rolled in. Experts from Carol Vorderman to Fiona Philips from GMTV have all shared their concerns about MMR with the nation. Less than a third of all broadsheet reports on MMR in 2002 mentioned that the overwhelming evidence showed no link between MMR and autism.

The MMR scare has now petered out. It would be nice if we could say this was because the media had learnt their lessons, and recognised the importance of scientific evidence, rather than one bloke’s hunch. Instead it has terminated because of the behaviour of one man, Andrew Wakefield, which undermined the emotional narrative of their story. The media have developed no insight into their own role, and for this reason, there will be another MMR.


I talk about this stuff at much greater length in the last chapter of the book:

And more importantly, because it’s free, here:

The Cardiff ESRC report is good (and I don’t want to incite copyright theft or anything, but for god’s sake, someone put it on Scribd, this is good publicly funded research, but Cardiff and ESRC change their web address for it every five minutes, I don’t think they really understand how the internet works):

Lastly, I recommend reading the GMC’s findings for yourself, if you’re interested in this stuff, as the details of what they got up to, the ethics committee dodginess, and the mis-statements around who the children were, is pretty bad. The document annoyingly hasn’t appeared on the GMC website – even though this is a massive news story today – but someone has resourcefully uploaded it for free public access here:

If you like what I do, and you want me to do more, you can: buy my books Bad Science and Bad Pharma, give them to your friends, put them on your reading list, employ me to do a talk, or tweet this article to your friends. Thanks! ++++++++++++++++++++++++++++++++++++++++++

135 Responses

  1. quasilobachevski said,

    February 9, 2010 at 1:11 am

    Gah. I meant ‘on tenterhooks’, of course.

  2. Alasdair.Kay said,

    February 9, 2010 at 3:20 pm


    At a brief skim it looks to me like their control group of “unvaccinated” is the same kids as the test “vaccinated” kids only in the ~15 months prior to their vaccination – which given that the mean age of onset is >4 years would present a problem. They don’t really explain how they do the “age adjustment” to account for this.

    They don’t appear to have any kids over ~15 months old who are not vaccinated (98.5% vaccinated before the age of 3).

    Could be wrong though, I was just skimming it and couldn’t see any description of the unvaccinated group aside from “The children were assigned to the nonvaccinated group until they received the MMR vaccine. From that date, they were followed in the vaccinated group.”

    I assume that since Poisson regression requires occurrence of “events” to be independent that the fact that both groups are the same people would negate that even if you ignore that there isn’t an actual control group.

  3. Alasdair.Kay said,

    February 9, 2010 at 3:43 pm

    Actually, I see they hint the vaccine uptake to be about 85% so there would be about 15% of children who are unvaccinated. Their numbers could still be skewed, and lack independence, by including the transition from one group to the other on vaccination though.

  4. skyesteve said,

    February 9, 2010 at 5:30 pm

    I maybe be wrong here – I don’t have the Wakefield and Co paper in front of me – but if memory serves me well that paper carried with it a statement along the lines of “We did not prove an association between measles, mumps and rubella vaccine and the syndrome described”.
    Was it not afterwards that Dr Wakefield suggested more strongly there was a link and urged a switch to single vaccines whilst his fellow authors said “We emphatically endorse current vaccination policy until further data are available.”
    As I said, I could be wrong and I’m happy to stand corrected…

  5. pv said,

    February 9, 2010 at 6:39 pm

    Whatever other studies might or might show, how does any of this validate Wakefield’s falsified nonsense as anything other than falsified nonsense. There is still no evidence to link MMR with autism, and Wakefield has still been found guilty of gross misconduct and dishonesty.

  6. pv said,

    February 9, 2010 at 6:40 pm

    “… might or might not show…” obviously!

  7. quasilobachevski said,

    February 9, 2010 at 10:59 pm


    Whatever other studies might or might show, how does any of this validate Wakefield’s falsified nonsense as anything other than falsified nonsense. There is still no evidence to link MMR with autism, and Wakefield has still been found guilty of gross misconduct and dishonesty.

    To make a scientific (rather than, say, ad hominem) argument that there’s no link between the MMR vaccine and autism, it’s not enough to discredit Wakefield or his methods. From reading Ben’s book and this blog, the Danish study seems to be one of the most definitive and often-cited pieces of evidence in this direction. So any suggestion that it’s flawed seems interesting.


    I don’t think your criticism of the Danish study holds up. The authors do more than ‘hint’ at the numbers involved. In the results section of the abstract, so in the first column of the first page, they say:

    Of the 537,303 children in the cohort … 440,655 (82.0 percent) had received the MMR vaccine.

    So it seems pretty clear that they had a decent-sized control group (about 18%).


    I’m still on tenterhooks!

  8. nevajism said,

    February 10, 2010 at 3:59 am

    Sorry- I was busy with a few things- ok the paper im going to analyse is actually this one

    “Thimerosal and the Occurrence of Autism: Negative Ecological Evidence From Danish Population-Based Data”
    Kreesten M. Madsen, MD* et al
    PEDIATRICS Vol. 112 No. 3 September 2003, pp. 604-606

    which I think is pretty much the same study and can be found here:


    ok so reading the title and abstract all seems to be well, there is an increase in autism despite the removal of thimerasol from vaccines in 1992 so no link, QED you might think, but lets look at this a bit closer

    in the methods they say “since 1995 outpatient activities were registered as well”


    “The proportion of outpatient to inpatient activities was about 4 to 6 times as many outpatients as inpatients”

    ok so thay have increased the patient base from 1995 onwards by adding 4 to 6 times more patients, fine proving they have normalised for it, and it does look like they have by looking at the graph (incedence are given in cases per 10,000)
    but look how they calculated that:

    “For each year and age band, we calculated the incidence as the number of people who at that age band and year was diagnosed with autism for the first time divided by the total number of people alive and living in Denmark at that age band and year.”

    so they have effectively given a measure which has NOT been normalised for the change in size of patient base- this is a fatal flaw and yet you wouldnt know it if you just read the abstract and title (which most people do)

    This is equivalent to saying we looked at a disease in bromley in 1995 and there were 100 cases, we then looked in the whole of london in 1997 and there were 200 cases- hence there is more of the disease

  9. skyesteve said,

    February 10, 2010 at 7:27 am

    Wasn’t the basis of the Danish conclusions, however, that they had 440,000 children who had had MMR and about 100,000 who hadn’t and the incidence of ASD was not significantly different?
    If the incidence of ASD has risen is this a genuine rise in actual “disease” or is it a rise due to increased awareness and increased likelihood of a diagnosis of ASD being made? Whilst I know it has no scientific basis I’m sure we’re all aware of folk from our childhood who nowadays would be readily diagnosed as ASD but weren’t then.
    Moreover, if MMR is the cause of any rise in ASD (real or perceived) should there not have been a dramatic rise in ASD within a year of it’s widespread introduction? Perhaps I’m wrong but I’m not aware that that’s what happened. Indeed, I have seen some stuff which suggests the rise in ASD diagnosis preceded the introduction of MMR.
    Perhaps there is a genuine rise in the incidence of ASD but by focusing so much on MMR maybe we have let the real culprit slip through our hands. I don’t know – just sepculating (which I know I shouldn’t do here but I still think it’s a valid question).

  10. Alasdair.Kay said,

    February 10, 2010 at 9:48 am


    Apologies – I did say I just skimmed it, but correct me if I’m wrong, it does say that the “nonvaccinated” group includes all the people who were later vaccinated up until the point where they were given the injection? I wouldn’t have thought that you could include members in the control group as “untreated” if they then became “treated”, unless that was consistent for all subjects or unless they had 3 groups, treated, untreated and transfers.

    For example, assuming they developed autism after the injection, how would you ever know that it wasn’t something they would have got anyway had they remained in the non-vaccinated group – at that stage they would be counting as a “nonvaccinated, no autism” and also as a “vaccinated, autism” statistic. I wasn’t sure if when they said “from that date, they were followed in the vaccinated group” meant that their data was removed from the “nonvaccinated”. I’d have thought that since it was a retrospective study they should have just assigned the people to one group or other. I didn’t see any explanation of their age adjustment either. But as I say, just skimmed it…

  11. Alasdair.Kay said,

    February 10, 2010 at 9:53 am


    I think you make a very valid point there – all the fuss over MMR may well have taken the attention off any one of a number of other possibles factors. It’s so difficult to know what the historical incidence was, when accurate techniques for scoring and recognition is such a recent thing. Whilst the questions hang over MMR there will be less opportunity for researchers to get funding to investigate other possible causal factors.

  12. pv said,

    February 10, 2010 at 11:38 am

    To make a scientific (rather than, say, ad hominem) argument that there’s no link between the MMR vaccine and autism, it’s not enough to discredit Wakefield or his methods.

    I actually wrote:
    “There is still no evidence to link MMR with autism…”, which I think at this point in history is true. I did not say there is no link.

    And I hardly think it is ad hominem to write:
    “Wakefield has still been found guilty of gross misconduct and dishonesty”, because it too is true.

    The only point I was making is that Wakefield didn’t find any link in the first place so, even if anyone can find fault with the Danish study, it in no way makes a case for Wakefield’s “claims”.

  13. nevajism said,

    February 10, 2010 at 11:12 pm

    sorry got the link wrong on that paper its here:

    if that doesnt get it just do a search on the title, the pdf is freely available

  14. quasilobachevski said,

    February 11, 2010 at 2:16 am


    The authors of the Pediatrics paper address your criticism on the nose. Specifically, they say the following at the end of the first column on page 605.

    In additional analyses we examined data using inpatients only. This was done to elucidate the contribution of the outpatient registration to the change in incidence. The same trend with an increase in the incidence rates from 1990 until the end of the study period was seen (data not shown).


    In the NEJM paper they do two different analyses. One measures persons, and the other person-years. I think you’re referring to the second analysis. The difference is explained in the first footnote to Table 2, which documents the results of the second analysis. It says

    The distribution of cases of autistic disorder or other autistic-spectrum disorders according to vaccination status differs from that in Table 1 [the first analysis] because, in this analysis, children who were vaccinated after the disorder had been diagnosed were classified according to their vaccination status at the time of the diagnosis (i.e., as unvaccinated).

    From this I infer that, in the first analysis, children who were diagnosed with autism were counted as vaccinated or unvaccinated in the obvious way.

    (The statistical analysis section doesn’t say this exactly, and seems really confusing.)


    Very sorry to have misrepresented you. Of course you’re right that Wakefield never established a link.

  15. Martin said,

    February 11, 2010 at 4:34 am

    If there were a link between MMR and autism, would the reduction in MMR vacinations in the UK during the previous decade lead to a reduction in cases of diagnosed autism? Or am I being overly simplistic?

  16. Alasdair.Kay said,

    February 11, 2010 at 9:14 am


    Good question – unfortunately not because accurate diagnosis of autism and ASD is still a developing process. Autism diagnosis rates may have increased over the last 20 years – but the prevalence of autism or ASD may not have.

    More cases of autism are being diagnosed, but this could just be that we are able to recognise ASD now where before it went undiagnosed. It’s hard to get a precise prevalence rate because ASD varies so much – Asperger’s Syndome or high functioning autism may never get diagnosed if the person is able to live a relatively normal life and the ASD isn’t accompanied with any learning difficulties. More accurate ability to diagnose autism really improved from the 90’s, whilst widespread use of MMR was introduced in 1988 (so the effects would theoretically be seen early 90’s). It’s like which came first, the chicken or the egg (and is it even a chicken’s egg anyway)?

  17. nevajism said,

    February 16, 2010 at 12:00 am

    Back to the piediatrics paper-
    you have to admit that if this is the hard evidence for anything it is pretty unclear- first they show a graph of data which they admit has not been normalised for patient population size, then as a throw away comment thay say that they looked in-patients only and it shows the “same trend”- but we dont know what the trend is because the data has been presented in a statistically flawed manner and they dont present the in-patients only data at all which we realy want to see- how did this make it through any kind of peer review? – and how can this be the hard evidence that organic mercury in vaccines dosent cause autism — anyone who wants to see the other side of the story should look at this talk from David Ayoub, M.D:

  18. nevajism said,

    February 16, 2010 at 3:02 pm

    Guess Im now banned from the other thread after ben sent that stream of insults he then blocked my reply (nice one ben!)

    On the subject of systematic reviews and MMR’s (nicely combining the two subjects in one) here is a paper outlining some of the problems with the Cochrane review on MMR:

    C.G. Miller (2006) Questions on the Independence and Reliability of Cochrane Reviews, with a Focus on Measles-Mumps-Rubella Vaccine, Journal of American Physicians and Surgeons 11(4) pp111-115.

  19. Ben Goldacre said,

    February 16, 2010 at 3:12 pm

    you mean here where you prevented people from having an interesting discussion about systematic reviews?

    nothing has been censored, i can find no evidence of you posting anything to that thread anywhere in my email or the web interface of this website.

    you’re an idiot, and a waste of time, pure and simple. and now you’re quoting JPANDS. beautiful that you’ve found each other, they’ll probably print an academic article from you if you send them something, give it a go.

  20. nevajism said,

    February 16, 2010 at 3:45 pm

    why all the insults Ben? have I touched a raw nerve?
    I tend to look at what something says rather than where it comes from – authoritory does not always mean truth. I once spent quite a bit of lab time showing that a paper in PNAS was wrong, but the rebutal only merited a paper in biophysical J.

  21. nevajism said,

    February 16, 2010 at 3:47 pm

    is it anoying for you to find another science nerd who does not agree with you ben?

  22. pv said,

    February 16, 2010 at 4:00 pm

    nevajism, you refer to a book called Silenced Witnesses (Vols 1 & 2) and you described Brian Deer and a complainant (which you can’t spell). That puts you in a deep hole of idiocy to start with. Everything else you write only makes it deeper. You just keep on digging…

  23. nevajism said,

    February 16, 2010 at 4:07 pm

    just because you and Ben call me an idiot doesn’t make it so – I assume you have heard of the term ad hominem?

    (sorry about the spelling – I’ll run things through a spell checker in future if it makes you happy)

  24. Ben Goldacre said,

    February 16, 2010 at 4:32 pm

    hi nevajism

    i guess if we were at a party i’d just ignore you. if you were being as annoying as u have been in my house (which u kind of are) i guess i might think about asking you to leave

    now, on this thread, i’m happy to leave you being dumb about vaccines, because we expect fools on those threads. what i found objectionable was when you drifted and started to be tedious on another thread. that was like jumping on a table and shouting at a party, and demanding that everyone looked at you and attended to you and your thoughts, instead of having their own chats. that’s why i think you’re unpleasant. you’re very free to continue being a fool here. we have no obligation to like you.

  25. pv said,

    February 17, 2010 at 9:51 am

    nevajism said,
    February 16, 2010 at 4:07 pm

    <i<just because you and Ben call me an idiot doesn’t make it so – I assume you have heard of the term ad hominem?

    You also need to brush up on your reading and comprehension skills. It’s not because anyone says says you are an idiot, it’s because you make it so evidently obvious. And you seem to be wearing your pointy “big D” hat with more than a modicum of pride.
    You cite books written by a known fantasist and fraudster as if they have some merit. You make claims about Brian Deer and the GMC which are patently false (and have been refuted by both Deer and the GMC), presumably because they fit in with your prejudices. So far you haven’t acknowledged the error of your claims. And you expect to be taken seriously.

    I’d say you’ve dug yourself an escape-proof hole there. Keep digging!

  26. quasilobachevski said,

    February 19, 2010 at 3:20 am


    I agree that it would be preferable if they had given the figures. But let me just summarise how the debate so far has gone.

    1. You said there was a problem with the NEJM paper.

    2. You changed your mind and said the problem was with the Pediatrics paper.

    3. I pointed out that the authors address your complaint.

    4. You blustered and moved on to something else.

    Have you ever heard Ben talk about the rhetoric of denialism? You’re a classic example. When one of your arguments gets knocked down, you move on to the next with no sign of self-criticism or apology for wasting everyone’s time.

    Funnily enough, I can’t be bothered to watch your video. I have better things to do.

    Oh, and one other thing. Ayn Rand? Ron Paul!? Ahahahahahahahaha! Ahahahahahahahhahaha! Thanks, I needed a laugh.

  27. jweirmccall said,

    February 19, 2010 at 8:55 pm


    “I once spent quite a bit of lab time showing that a paper in PNAS was wrong”

    The lady doth protest too much, methinks. I love how your evidence against vaccines come from people who stated the following regarding evidence:

    ‘Physicians must beware of accepting the concept of a standard of care that is itself evidence-based, threatening the autonomy of physicians and subjugating the patient’s interest to that of the collective.”

    They also share your love for Ayn Rand quoting them in their editorials. You are very apt bedfellows indeed.

    Ps. Having problems posting at the moment, so sorry if this has appeared before

  28. nevajism said,

    February 28, 2010 at 2:32 pm


    I thought I’d wait a while before replying just to see if you could muster a good argument- sadly I was disappointed.

    “The lady doth protest too much, methinks”
    Funny that I was thinking the same thing – I seem to have triggered a lot of protestations from Ben and his acolytes

    quasilobachevski – I thought I addressed your points pretty directly- are you SERIOUSLY telling me that you think this is a valid piece of work?

    pv – so a person who has made accusations against Brian deer and the GMC is wrong because they (the accused) have said he is a mentally deranged fantasist? – Well they would say that wouldn’t they. What did you expect them to say? Rather than take sides in that way I suggest you read what all parties have said and then use your own logical powers of reasoning.

    On the subject of Wakefield- may I suggest that the sudden resurgence of interest in his case was in a desperate attempt to try to block this paper:

    which had passed the peer review process but now seems to have been withdrawn at the request of the editor- the results are pretty damning wouldn’t you say? Does anyone want to point out a specific problem with this paper? I cant see why it would be withdrawn other than for political reasons, even if you say there weren’t enough monkeys in the experiment, the least that should happen is that this work should trigger a larger scale study to confirm or refute its findings, this is how good science works, is it not?

    – I wait for further accusations of being a dunce, madman, idiot – please use your imaginations and see what other words you can come up with to try and refute me with personal insults- its always the best standby when you haven’t got a real argument to fall back on.

  29. nevajism said,

    February 28, 2010 at 2:59 pm

    should mention the title of that NeuroToxicology paper, just in case you dont follow the link:

    Delayed acquisition of neonatal reflexes in newborn primates receiving a thimerosal-containing Hepatitis B vaccine: Influence of gestational age and birth weight
    Laura Hewitson Lisa A. Houser, Carol Stott, Gene Sackett, Jaime L. Tomko a, David Atwood, Lisa Blue, E. Railey White and Andrew J. Wakefield

  30. nevajism said,

    July 12, 2010 at 12:39 pm

    this is a very interesting recent lecture from wakefield outlining the above study:

    the other 9 sections are also on youtube

  31. edwardm said,

    January 1, 2011 at 6:25 pm

    nevajism – Get a grip. You’re just posting random papers now which test for possible links between thimerosal-containing vaccines and autism. Are you that stupid that you think this is evidence that MMR causes autism when the MMR vaccine has never had thimerosal as an ingredient.

  32. oakwoodbank said,

    January 8, 2011 at 10:24 am

    It can reasonably be assumed that some vaccines do cause damage to children or we would not have the Vaccine Damage Payments Act 1979. [1] The list of diseases to which this act applies includes measles and rubella as well as diphtheria, tetanus and whooping cough.

    There does not appear to have been a consolidated paper that attempts to show how vaccines might cause brain damage by a review of the scientific literature except for the one that I have published on various web sites including scribd. [2] It has been published for open access. This paper looks at the causative factors mainly from the effects of the whooping cough vaccine (pertussis).

    However, the U.S. Department of Health & Human Services, Centers for Disease Control and Prevention, National Immunization Program, promulgates that the risks from MMR vaccine can be permanent brain damage. [3]

    Epidemiologists will probably say, and quite rightly, that the benefits of vaccination outweigh the occasional damage that is caused by some of them. The number of payments for vaccine damage in the UK is in excess of 1,000 and compared with the number of children who have benefitted, that number is miniscule bit not unimportant.

    Over a thousand children will be living diminished lives as a result and their parents will be anguished that they were, in part, responsible for their child’s disability.

    For this lifetime of inequality and the loss of; normal education, a job and a life that contains the expectancies that most of us have envisaged and possibly achieved, the maximum award is, If you are severely disabled as a result of a vaccination, a one-off tax-free payment of GBP120,000.

    It is this shameful situation that ought to cause concern rather than a continuing denouncement of Wakefield. He may have got his understanding of the relationship between MMR and Autism wrong but that does not prove that there isn’t one.


    2 Challoner A. Brain Damage Caused by Vaccination. 2009.


  33. trombone said,

    November 19, 2011 at 12:30 am

    Probably history now, but from memory Wakefield et al’s research didn’t actually say they had found a causal link, indeed I think it pointed up that the results were inconclusive and that more research needed to be done.

    If my recollection above is correct, then all of his co-author colleagues who later renounced the study, were in effect renouncing their own conclusion that the study was inconclusive!!! Exactly what the establishment were trying to claim!

    This is important when you recall that the press/establishment repeatedly claimed the research (and Wakefield himself) were claiming to have found a link. Bad interpretation on the part of the press? If true, it also implies that those who renounced their own work might have been concerned about their own personal situations.

    Can somebody correct me please? – I don’t like to be wrong!

    Btw, While I accept he may have made mistakes(I am not a medic etc) I was under the impression good science was about asking the sort of questions that Wakefield did, and that all research was good in some way -ie a “no causal link” answer can be just as informative as a positive link. Further, while scientists, journalists etc should try to undertake the best possible work they can, sometimes even the bad science tells us something about say, the peer review process, how to better reduce errors, Accreditation flaws etc, and by identifying those flaws we can address them. Provided it’s not hectoring, criticism is always constructive.

    Of course a corollary of the above para is that the atmosphere between Wakefield, the press and the establishment could have warped the direction of research and hence the wider picture

  34. Alex said,

    August 21, 2013 at 11:22 pm

    Not sure if you picked up on this recent article Ben, any comment? I’m still unsure as to which ‘side’ is trying to pull the most wool!

  35. Alex said,

    August 21, 2013 at 11:32 pm

    Sorry, That link should have been