The BMJ changes its mind (kind of) (Edit: not really)

So last week, our friends at the BMJ published yet another ‘news’ item in one of the esteemed journal’s very-much-not-peer-reviewed sections. It carried a rather dramatic headline:

Covid-19: Boys are more at risk of myocarditis after vaccination than of hospital admission for covid

In a world swirling with anti-vaccination conspiracy theories, you could say this ‘finding’ was pretty topical, to say the very least. Unfortunately, however, it was also pretty rubbish. The purported finding — that boys are endangered more by Covid vaccines than by Covid itself — was simply fallacious.

The study that supposedly justified this claim had already been widely debunked by academics and scientists across many public fora for weeks.

It may have provided a basis for some kind of BMJ thought-piece about how dodgy research can so easily find an audience on the internet, but it was far from a newsworthy update on the current state of vaccination science.

All of which raised the question as to how the BMJ’s news reporter came to produce such a naively credulous piece in the first place — and, of course, more questions around how the BMJ’s editorial processes manage to approve such pieces for publication.

The reaction on social media was swift:


In short, the study was full of holes and errors, and rested on an unreliably biased methodology. One high-profile review (dare I say, higher-profile even than the BMJ’s news pages) had seen fit to dismiss the paper as little more than a “dumpster dive.”

And, in any event, it wasn’t a real paper; it was a preprint. As such, it has never been formally published or peer-reviewed. It occupies the same status as any other random Word document stored on an academic’s hard drive, somewhere in some university.

It was yet another example of a problem we have seen all too frequently during the pandemic: namely, the problem of science by press-release.

* * *

So, what did the BMJ do to rectify the situation? Well, one thing it did not do is retract the original ‘news’ story. Instead they doctored the online version to add a couple of paragraphs acknowledging the criticisms. These online edits had the effect of fundamentally altering the tone of the piece. They even changed the headline:

Covid-19: Study that claimed boys are at increased risk of myocarditis after vaccination is deeply flawed, say critics

The problem I have with this is as follows: these edits amount to face-saving, not fact-checking.

The BMJ should have clearly explained that the study they were writing about was known to be problematic before their original ‘news’ story appeared. Had the BMJ reporter (or editor) been aware of the problems in advance, I dare say that they would never have covered the study as ‘news’ at all.

Instead, this Frankenstein-monster of an article remains now on the BMJ ‘news’ website with the exact same publication details (e.g., the same doi number) as the original. It has the same web url (i.e., website) and official citation information (date and BMJ volume number).

This approach implies that there was no editorial slippage, and occludes the fact that the corrections that were required were substantial. It is makes it seem as though the edits were little more than typos. This wasn’t a published correction, it was an effort to absorb corrections into an existing article without drawing attention to the fact that they were required. The old piece — the embarrassing version — no longer exists.

Viewing the new after-the-fact rehash, a reader will not be able to appreciate the entire story of what took place here, or the BMJ’s role in it. But then I guess that is the whole idea.

Call me old-fashioned, but I thought the point of journalism was to inform people — not to leave them with a false and incomplete impression of the events they are reading about.

Comments are closed.