Tuesday, March 15, 2011

BMJ Opens the Door to Conspiracy Theories

I don't like conspiracy theories.  I don't like wasting my time thinking that a group of people are secretly conspiring to pull something over on the rest of us.  It has been my experience that, most of the time, the conspiracy is nothing more than a figment of someone's imagination.

But, I do know that people are not always completely forthcoming with the truth.  I know that backroom deals happen all of the time for the benefit of those involved and the details of these deals are rarely made public.

I am also acutely aware that money corrupts and that if one group gives a good amount of money to a second group that first group has significant power over the second.  Money buys access, favors, and above all, leverage over the recipient.

When it comes to published science, money is very problematic.  I don't think it is any secret that when research is sponsored by a company, it tends to be more favorable to that company than independent research.  And then there are the (hopefully rare) cases when a company goes completely over the line and then we get ghost-written studies and entire journals that are fake.

But this is the stuff that conspiracy theories are made of.  What happens is that you have a company that appears to have bought some research to improve its bottom line.  The parties involved try and hide the details of the transaction but eventually they get caught and the affair is made public.

Hopefully at that point the parties are at least honest and admit to the affair (although I know that doesn't always happen) but the damage is done - our trust in the parties involved has been eroded.  The conspiracy has been started and we start having to constantly ask ourselves what else "they" aren't telling us.  And the next time that something gets published that involves these parties we have to ask ourselves if they are being completely honest with us.

Now, there are a few ways to deal with this problem.  The best way is for researchers and scientific publications to not accept any money from companies who could have an interest in the results.  But we all know that isn't going to happen.

A second way is to always be completely forthright about any financial arrangements between the researchers and publication and companies.  This doesn't get rid of the conflict but it at least gets it out into the open and defuses the conspiracy aspect.  But for this process the work, the disclosure has to be there from the start.  It doesn't work if you only admit to it after the fact.

Having said all of that, I think that the major journals are well aware of this problem and go to great lengths to even avoid the appearance of impropriety.  So imagine my shock when the BMJ was called out recently for failing to disclose a financial relationship with Merck in their recent pieces calling the ex-doctor Wakefield a fraud.

The problem is this.  A group of, ahem, interested parties combed through the records of the BMJ and found that they had a substantial financial relationship with Merck.  This group publicly called the BMJ on the lack of disclosure in the Wakefield articles, the journal responded with a resounding "whoops" -
Although Vera's claims may seem far fetched on this occasion, she is right that we should have declared the BMJ Group's income from Merck as a competing interest to the editorial (and the two editor's choice articles) that accompanied Brian Deer's series on the Secrets of the MMR scare.[2] [3] [4] We should also, as you say, have declared the group's income from GSK as a competing interest in relation to these articles. We will publish clarifications.
We didn't declare these competing interests because it didn't occur to us to do so. We saw this series not as pro-MMR vaccine or pro- vaccination in general, but as against fraud and corruption in medical research.
Go read the entire response from the BMJ, it is quite interesting.  The fact  that an article about "corruption in medical research" failed to disclose a significant potential conflict of interest is ironic, to say the least.

There are at least two ways that you could look at this situation.  The first is that the journal made an honest mistake and came clean when called on it.  And if this were another situation, that would be exactly the way that I would look at this.

But in the case, the journal was already on shaky ground.  Remember, this series of articles was commissioned and paid for by the journal.  The journal went out and hired a journalist - not a scientist or researcher - and paid him to write these articles.  And it wasn't just any journalist either - they hired Brian Deer who has made it his mission in life to hound Wakefield.  More importantly, these articles weren't the typical peer-reviewed articles that a journal would be expected to publish.  These articles were written with the explicit purpose of attacking a researcher and calling them a fraud.

Given all of this, I would have expected that the journal would have gone well out of their way to avoid even the appearance of impropriety.  After all, if you are going to call out someone else's misbehavior, you had best not be guilty of the same.

But they didn't.

So then we are left with the facts that the journal commissioned a paid attack on a researcher and that the journal had undisclosed financial ties to companies that have a vested interest in seeing this line of inquiry shut down.  And when you add in the idea that children are the victims of this potential collusion, what you are left with is a very powerful narrative.

I don't know whether the corporate sponsorship that the BMJ received in any way influenced the decision to commission and publish these articles on Wakefield.  I want to think that it didn't,  but the sad truth is that there is no way to know for certain if it did or didn't.  The potential for bias is there and the journal dropped the ball and made it look like they were trying to pull a fast one.

A new conspiracy is born.

5 comments:

  1. It looks messy, no doubt. But facts remain facts, regardless of funding or vested interests. Either the claims made by Brian Deer about the fraudulent practices of Wakefield, are true or the aren't true. I'm not particularly interested in the claims about Wakefield having a financial interest in law suits and alternative vaccines. In the same way that I don't find the BMJ's potential financial interests overly relevant either. I'm more interested in the claims about false and altered data.

    It's still an interesting post, that raises important questions. Financial interests do give us reason to be extra careful about findings; to double-check facts and methods. But that still has to be the focus - discovering the facts.

    ReplyDelete
  2. The word "facts" is a somewhat loaded concept and not one that science is normally concerned with. Rather, science is more concerned with evidence and whether all of the available evidence, taken together, supports the conclusion. The important part here is that you have to present all of the evidence - not just the particular bits that fit with your theory.

    Which brings me to the Wakefield GMC hearings. The GMC clearly had more evidence available and yet did not find Wakefield guilty of outright fraud. If I am not mistaken, the GMC used the equivalent to "beyond a reasonable doubt" to make that judgement (although the typical standard that the use is less strict than that).

    The standard of evidence required for a journal is much higher than a hearing like the GMC. The standard for a journal is that all of the evidence, taken together, must support the assertion being made AND you have to show that result is unlikely to have occurred by random chance.

    So, if the GMC could not find that the evidence supported the assertion that Wakefield committed fraud, beyond a reasonable doubt, what does that say about the BMJ?

    I think the articles in the BMJ fail to reach a scientific level of certainty. I would also question whether the evidence would even reach the standard of beyond a reasonable doubt. None of this is to say that any of the "facts" in the story are wrong, I am just not certain that it is the complete picture.

    Which brings me to the issue of trust. I don't have access to all of the same information that Deer did when he wrote the articles so I can't judge the validity of the claims for myself. I have to trust that Deer and the BMJ are being completely honest in their presentation of the evidence and are presenting a balanced look at the evidence.

    But when I look at how the BMJ did these articles, I find it difficult to trust them.

    The journal went out and paid someone to write the articles. They didn't go to a neutral party but rather someone who has had a firm position that predates the GMC hearings (the evidence used in the articles). Having a preconceived notion of what the facts say before you look at them and only using the facts that fit your idea is a big no-no in science.

    The BMJ also hired a journalist rather than a researcher or scientist. Journalism is quite different than either law or science - it is picking certain facts and using them to tell a compelling story. The articles in the BMJ seem to be journalism more than anything else. Which I have to wonder what journalism is doing in a peer-reviewed journal.

    Now we find out that the journal had undisclosed financial ties to interested companies. And, while I have to give the journal credit for admitting that the ties are a conflict, I find their reason for not disclosing in the first place to be very lame, at best.

    But just to be clear, I am not trying to say that Wakefield is innocent of all charges but rather that the actions of the BMJ are really starting to smell bad.

    ReplyDelete
  3. One clarification to your comment there MJ - fraud is not a matter of science. Fraud is something that can be factual or false, and it was only the fraudulent data that I was referring to (not the scientific claims pulled from the data).

    You're right that courts of law also deal in matters of evidence - in as far as there is a burden of proof - but that doesn't mean that fraud cannot be talked of as a factual issue that can be definitively proven.

    I too am skeptical of (some, not all) journalists' stories, in as far as their purpose is to find a story. But if a journalist's claims are incorrect and defamatory, then the affected party has legal recourse to a defamation suit. Yet I understand that Wakefield has not taken a legal challenge against Deer. That doesn't mean Deer is right, but considering how serious the defamation is and how much Wakefield could get in significant damages, it is surprising that he is not taking the matter to court. (Yet..?)

    ReplyDelete
  4. Fraud is a matter of intent though and when you are talking about intent you leave the realm of hard facts. After all, the only person who knows the "fact" of whether Wakefield intended to commit fraud is Wakefield himself. But, we can hardly take his word for it. The rest of us can only look at the evidence that we can see and guess as to what he was trying to do.

    If the intent wasn't there than the worst we can say is that Wakefield was just horrible at his job and managed to get a piece of junk published. He wouldn't have been the first person and he certainly won't be the last to get bogus research published.

    As for filing a lawsuit, Wakefield tried that once before and lost. I don't know what the laws of the UK are like so I can't judge whether he would have a good chance of success or not.

    I know in the US that such a suit can he hard to win and really depends on how much time and resources you want to spend. For example, Dr Offit was quoted in wired magazine as saying that Barbara Fisher "lies" -

    "Fisher, who has long been the media’s go-to interview for what some in the autism arena call “parents rights,” makes him particularly nuts, as in “You just want to scream.” The reason? “She lies,” he says flatly."

    You would think that directly calling someone a liar in a major magazine would be defamation, but Fisher lost the case.

    If Wakefield chose to sue he would have to sue the BMJ and I would think that they have a slightly larger legal budget than he does. And remember, he has just gotten done with a multi-year fight with the GMC and has lost the ability to work in his professional field. I wonder if he has the funds to mount a challenge (or if he even has the will to fight it anymore).

    So I don't think we can conclude anything solely on the basis of Wakefield not suing.

    ReplyDelete
  5. It hardly matters now does it? - the whole point was so that the world could name Wakefield a fraud, and the (uneducated) villagers could "lynch" him. vaccinations could resume the - godlike never question status that the pharmaceuticals want - and with the current journalistic rules that vaccines can't be questioned everything can go back to business as usual - with the added bonus that all doctors are now very aware of what happens to those who speak out (just in case they weren't already)

    Autismandoughtisms-

    science doesn't deal with "facts" so much as "theories" that try to make sense of "observations"

    my observations are - gut problems and autism are linked. both my very bright kids have the "Autistic like symptoms" associated with coeliac. Moreover MOTHERS with coeliac are more likely to have allergies/more likely to have autistic kids. (remember that damage often happens during pregnancy) There's a clear picture and there are clear pathways that are in the scientific record - completely separate to Wakefield's research. Don't you think it's time we started listening to the science instead of $$$$ pushing an agenda onto the gullible.

    ReplyDelete