Misdirected criticism by someone from whom I would never have expected it

It is with some trepidation and more than a little regret that I begin writing this piece. The reason for my hesitation is that, by doing so, no matter what I say I’ll be inserting myself into what appears to be a disagreement among people all of whom I admire very much. I don’t really want to do it, but I feel obligated, because the issues brought up in the disagreement are important, and reasonable people can disagree–sometimes strongly–about them. I also believe that someone whom I admire greatly has made a regrettable mistake.

Over the last week or so, I’ve been blogging a lot about the recent revelations in the BMJ by investigative journalist Brian Deer of the true depths of Andrew Wakefield’s perfidy. New revelations of shady business plans to promote woo-based diagnostic tests for “autistic enterocolitis” were added to familiar revelations of massive undisclosed conflicts of interest, breaches of research ethics, and falsification of data to produce an even more disturbing picture of Wakefield’s activities than ever before. Not only is he a crappy, dishonest scientist, but the evidence presented by Brian Deer is very strong that he’s also a dishonest fraud as well. Over the last seven years, Deer has done amazing, yeoman work in digging into the toxic waste dump that is the Wakefield’s dishonesty, fraud, and bad science, excavating the evidence, and putting it all together to produce this picture. I have nothing but the utmost admiration and respect for him.

Which is why I was more than a bit puzzled by Deer’s blog post yesterday for The Guardian entitled The medical establishment shielded Andrew Wakefield from fraud claims. Emblazoned under the title is this subtitle: “Brian Deer spent years investigating Andrew Wakefield’s MMR and autism research, which he now alleges was fraudulent. Here he argues that doctors closed ranks behind one of their own.” Before I explain what Deer has done that, in my opinion, is a mistake, let me just point out that, actually, it would not have surprised me if this were in fact the case to some extent, given how the Royal Free Hospital and UCL aided and abetted the Wakefield publicity machine after the release of his 1998 Lancet paper, as described in Deer’s BMJ article this week. Unfortunately, even though it wouldn’t surprise me if it were true that parts of the medical profession actually did close rank to some extent around Wakefield, at least until the General Medical Council hearings began, the physicians chosen as examples by Deer to illustrate his point were poorly chosen and and are unconvincing. And I’m not just saying this because I greatly admire them.

First, Deer criticizes Dr. Paul Offit:

But a Philadelphia-based commentator was not impressed by the BMJ’s intervention. “It doesn’t matter that [Wakefield] was fraudulent,” Dr Paul Offit, a vaccine inventor and author in Pennsylvania, was quoted in the Philadelphia Inquirer the next day as saying. “It only matters that he was wrong.”

I wasn’t surprised. From his establishment vantage-point, this was the third time Dr Offit had popped up to opine on the issue. Twice previously he’d been quoted as saying that my findings were “irrelevant” (although he’d been happy enough to use them in his books). Science had spoken, his argument went. There was no link between the vaccine and autism. It was experts like him who should rule on this matter, he seemed to imply, not some oik reporter nailing the guilty men.

I will admit that, on the surface, Dr. Offit’s words seem somewhat troubling in that they do appear to dismiss the importance of Deer’s findings. On the other hand, having corresponded with Dr. Offit a few times over the last couple of years and read two of his books, I find it hard to believe that this is what he really meant. Personally, I’d really like to see all of what Dr. Offit said in context, because the Philadelphia Inquirer article quoted only quoted one sentence about this issue, and that was it. Moreover, later in the article Dr. Offit is also quoted as saying:

Wakefield “is not a scientist,” Offit said. “He is fanatical.”

This hardly sounds like any sort of bashing of Brian Deer or closing ranks around Wakefield. Rather, from my perspective, Dr. Offit’s statement probably represents a fundamental–and innocent–difference between the mindset of a scientist, like Dr. Offit, and that of an investigative journalist, like Brian Deer. This difference in mindset, in my estimation, has led Deer to incorrectly view comments such as the one by Dr. Offit in the wrong light, as if they somehow represent the arrogant denigration of his years of hard work.

I can’t read Dr. Offit’s mind, but I just don’t see such a dismissal of Brian Deer’s work in his words. Indeed, I can even see where Dr. Offit is coming from, as I’ll explain later. What I suspect that Dr. Offit means is that in an ideal world science should be the deciding factor in such debates, an assertion that I agree with completely. In such an ideal world, it wouldn’t be necessary to worry about whether or not Wakefield committed fraud, because the science would matter far more than the scientist. (Yes, applying that word to someone like Andrew Wakefield does stick in my throat as I type this.) However, because we do not live in an ideal world, that Wakefield apparently committed research fraud matters very much indeed, not the least of which because fraud is something that lay people understand far better than long discourses on why Wakefield’s science was incorrect and didn’t show that the MMR vaccine causes or contributes to a syndrome of regressive autism and enterocolitis. It’s far easier to convince someone saying something like this,”The MMR vaccine never caused autism or enterocolitis. Rather, Andrew Wakefield falsified data to make it seem as though MMR caused these problems, and he did it to create a market for products he had previously patented,” than it is to point out the scientific flaws in Wakefield’s results. More importantly, fraud undermines the scientific enterprise because, unlike errors and science that is later shown to be wrong (as in the decline effect), it produces an intentional, systematic error into the scientific enterprise.

Consider it this way. To a scientist, it really is the data that matter (or should be), not the scientist gathering the data. Science is a means to an end: to find out how nature works, to plumb its mysteries and discover rules by which it works and that we can use to make predictions. Science, to its credit, also tends to work on the honor system. We scientists tend not to consider perhaps as much as we should the possibility that an investigator is lying or falsifying data. We tend to assume honesty as the default, at least when scientists report their results. The downside of that attitude is that scientific fraud is far too easy to get away with and all too often never discovered. The other downside is that often scientists have a hard time believing it when clear evidence of fraud is presented, as it has been for Wakefield. Naively, our heads in the clouds, we tend to disdain the down and dirty business of proving fraud, assuming that science will work it all out. In other words, science and scientists tend not to deal with fraud very effectively at all.

Contrast this to an investigative journalist, like Brian Deer. A good investigative journalist has to be a crusader, and there’s no doubt that Brian Deer is an excellent investigative journalist. Here’s where the different purposes and world views of scientists and journalists come in. In contrast to a scientist, Deer’s primary goal is not to find out how nature works but rather to use his investigative skills to develop a story and, in particular, expose wrongdoing, the better hidden and worst the wrongdoing the better the story. Just look at the history on his website, which includes not just the Wakefield story. He’s taken on pharmaceutical companies galore, for instance, exposing various scams. These skills have served him well in taking on Andrew Wakefield, whose fraud goes beyond that of even many pharma scandals from the last decade.

To this end, Deer has to be every bit as tenacious as the most dedicated scientist–more so, in many cases–and there is no doubt that Brian Deer has been admirably tenacious in pursuing Andrew Wakefield. However, that tenacity is exercised in the pursuit of a good story and in exposing wrongdoing. There is also a marked contrast in that the investigative journalists that I’ve met, rather than assuming honesty as the default, as scientists tend to do when reading scientific reports, tend to assume dishonesty as the default because, well, they investigate a lot of dishonest people trying to hide wrongdoing from them. I can’t help but wonder if this difference in world view and purpose is a large part of the reason for this misunderstanding that has led Deer to write such a misfire of a criticism.

Be that as it may, from my observations, scientists do tend to be much less concerned about exposing wrongdoing in science except insofar as when it is necessary, not because of a lack of a sense of justice, but more often due to a naive faith that science can handle anything, even conscious fraud by an investigator. Unfortunately, that is not always the case. Athough it can judge whether research is competently conducted and whether the results follow from the data presented, peer review, for example, is notoriously bad at detecting falsified data. Consequently, when Dr. Offit says that Deer’s findings are “irrelevant” (I’d really like to see the context in which he said that, too) or that it “doesn’t matter” that Wakefield’s work was fraudulent, he is doing no more than expressing the view of a scientist looking at science at work, expressing the belief that, in the end, science will prevail and that the data matter more than the scientist collecting them. Unfortunately, Deer appears to interpret this belief as somehow denigrating his efforts:

Over the past few years, many of the MMR vaccine’s medical champions have queued up to take a pop at my investigation. They’ve never found any fault in its accuracy or originality. But it was as if they felt that somehow it wasn’t right.

What about Ben Goldacre? I don’t always agree with Dr. Goldacre, but I can say, particularly after having met him in person three months ago, that without a doubt we agree about far, far more things than we disagree about. One of the things we (sort of) disagree about is one of the sore points that Deer has with him:

“Actually, I would like to speak in defence of Andrew Wakefield,” said Guardian Bad Science columnist Dr Ben Goldacre in a BMJ video, long after I first skewered the man’s research but before Wakefield was struck off by the GMC. “I’m not sure it was necessarily a bad piece of research.”

We actually discussed this very point. First off, this quote was from a period before the GMC found Wakefield guilty of serious professional misconduct and ethical breaches. I disagreed that Wakefield’s work was an acceptable case series, and told him so. Indeed, I still can’t believe that The Lancet published Wakefield’s report. It was, in my estimation, a mediocre little case series that, when you come right down to it, didn’t show much of anything, really, even taking it at face value and even if its data hadn’t been falsified and the research itself tainted by all of Wakefield’s undisclosed conflicts of interest. Usually papers with such weak data don’t end up in The Lancet. That’s why I’m still puzzled as to why this particular case series did end up in The Lancet, given that, even knowing that none of the editors or peer reviewers was aware at the time that Wakefield’s manuscript was fraudulent, it quite simply just wasn’t a very good paper. Sure, if it hadn’t been fraudulent it would have been perfectly acceptable to publish it, but it belonged in a lower tier journal. I also agree that the behavior of The Lancet’s editors has been–shall we say?–less than admirable, probably because their decision to publish such a crappy little case series so spectacularly in their faces. If Deer is planning on including that aspect of the Wakefield case in his next article, I’ll be in agreement.

Another aspect of Deer’s criticism might be that Dr. Goldacre has taken a strong stance that the British press was largely to blame in this debacle, having aided and abetted Wakefield in promoting the MMR scare. He also has strongly criticized the press for focusing on the man rather than the science. Just as Dr. Offit has argued that science should matter more than the man, so has Dr. Goldacre. Both are correct, but Brian Deer is also correct in that, once the man has become an issue, his misdeeds matter, and science might not prevail without help. I agree with that as well. However, in an ideal world, science would reign supreme when it comes to settling matters such as whether the MMR causes a syndrome of regressive autism and enterocolitis, and the press would “get it” and not focus on the human conflict more than the science. In such a world, findings of fraud would be “irrelevant,” at least with respect to the conclusions of science. Unfortunately, Deer appears to see this view, not for what it is, but rather as contempt for him and his methods:

So, what’s my point? I think these comments reveal a striking pattern: doctors default to defending other doctors. In fact, until recently there was a GMC regulation that banned them from bad-mouthing colleagues.

But in the specifics of their stance there seemed the idea that scholarly debate, epidemiology and suchlike, should arbitrate. Truth would emerge from the “scientific method”, not from “we can reveal” media muck-raking.

Such faith in science was also the apparent view of Wakefield’s medical school, when in 1998 he launched the MMR scare in a five-page paper in The Lancet.

I’m afraid I’m about to risk getting in Deer’s bad side right here. I don’t want to do it; but I have to say this: Count me in 110% as one of those physicians who believe “that scholarly debate, epidemiology, and suchlike should arbitrate issues of science.” That is the way science should work. That is how scientific questions like whether the MMR causes regressive autism associated with enterocolitis should be settled. That is what should persuade journalists and politicians, who then persuade the public.

Sadly, though, I also recognize that such is not always how science is done, how scientific questions are resolved, how the press, government, and population are persuaded. (The last case, in particular, is probably rarely how the world works.) I recognize that fraud happens and that, because science tends to assume honesty as the default, sometimes the scientist does matter. Science is a human activity that can’t be divorced from the humans that do it. That’s why it is sometimes necessary to focus on the scientist rather than the science. I don’t like it. I don’t like it at all. However, I understand that this is not an ideal world, and, because this is not an ideal world, we can’t have that scientific utopia that I wish we could.

Sometimes we need pit bulls like Brian Deer to latch onto the dishonest and drag them into the light, the better to reveal their perfidy to the world.

At the same time, I must question Deer’s choice of Drs. Goldacre and Offit as examples of the medical community closing ranks to protect one of its own. Indeed, even though I found Dr. Michael Fitzpatrick’s characterization of the GMC hearings against Wakefield as a “witch hunt” to be clueless in the extreme, I must also question Deer’s choice of him as well, because Dr. Fitzpatrick’s defense of Andrew Wakefield appears to derive far more from his politics and AAPS-like distrust of authority–medical authorities included–than from a knee-jerk reaction to circle the wagons and defend a fellow physician. (Dr. Fitzpatrick has said similar things about HIV/AIDS denialists and global warming “skeptics, for example, although, if anyone comes close to deserving Deer’s criticism in his article, it’s Dr. Fitzpatrick.) Worse examples of such behavior than Drs. Goldacre, Offit, and Fitzpatrick I would have a hard time imagining, as all three of them have been very clear and consistent in criticizing Wakefield’s “science.” There is no conflict between believing that scientific questions should be settled by scientific studies, debate, and evidence whenever possible and accepting that science often doesn’t deal with liars and fraud very well. Nor does expressing a desire that scientific questions be settled scientifically, as Dr. Offit does, for instance, preclude recognizing that we do not live in an ideal world and sometimes someone skilled at detecting fraud becomes necessary to overturn a false result. These are complementary, not opposite, techniques.

None of my disagreement with him alters my admiration for Brian Deer’s amazing skill and dedication or the huge service he has done for the U.K. and the world by exposing Wakefield’s research misdeeds at great personal cost. I will, of course, read (and very likely blog about) Deer’s final installation, anticipating that it will certainly document dodgy behavior on the part of some physicians in the U.K. closing ranks behind Wakefield, in particular that of The Lancet‘s editors defending their decision to have published his case series in the first place. I have little doubt that the charges will, as always, be well documented and persuasive. There’s plenty of blame to go around when it comes to the question of how Wakefield’s fraud could have spread so far and so wide, to such horrific effect. I just wish that Brian Deer hadn’t, in the lead up to the publication next week of his article describing how British physicians closed ranks around Wakefield, chosen such inappropriate examples to use to illustrate his point. Given that I know that Deer sometimes reads this blog, at least whenever vaccine stories are going around, I also hope that my words will lead him down a road that will show him that he’s made a mistake.

ADDENDUM: Steve Novella has weighed in as well.