Plausible deniability: Avastin, Lucentis, AMD, ARVO, and a chink in the Ingelfinger armor
By now, if you follow health and science news, you will have probably heard about the results of a trial that isn’t scheduled to be announced until Sunday at an ophthalmology conference. But the top-line findings of the CATT study, comparing Avastin and Lucentis for macular degeneration, appeared in the New York Times yesterday thanks to a scoop by Andrew Pollack. The embargo on the full results of the study has just been lifted by the New England Journal of Medicine, which is why this post can go live without breaking any kind of embargo, although the study itself hasn’t quite appeared online yet. (If you’re reading this after it went online, the link on “the study itself” will work.)
The incident isn’t over as far as Embargo Watch is concerned, however. There are a few issues here:
- Did Pollack break an embargo?
- Why didn’t NEJM invoke the Ingelfinger Rule, refusing to publish because results had already appeared?
- Will anyone be sanctioned for this? As the Times noted: “Investigators in the National Eye Institute trial had a day-long meeting on Tuesday in Chicago to learn the results. But they were sworn to secrecy.” Two of those investigators spoke to Pollack anonymously.
I asked NEJM’s media relations manager Karen Buckley about the decision:
Why is NEJM still publishing the study, despite the Ingelfinger Rule?
Why we are still publishing the study, even though some of the results were leaked to a newspaper and reported prior to publication? It would not be fair to the authors- or the public- at this stage to not publish the article, and make the complete, peer-reviewed and edited study available where everyone can read it.
Why publish the study without sending it out to reporters under embargo first?
It’s difficult to place an embargo on information that has already been reported. Since some information is being circulated, it makes the most sense in this case not to further delay the public availability of the study.
Is there a way to hold someone accountable for this leak?
We hadn’t provided any information to the media under embargo when the New York Times story was published, so we can’t call this an embargo break, and therefore can’t hold a reporter accountable. To the best of our knowledge, the authors on the paper behaved appropriately in the way they communicated to the investigators (in advance of the media, with signed confidentiality agreements, without disseminating any information by hand), so we cannot hold them accountable. The investigators are accountable to the NEI, therefore it is up to the NEI to handle any breach of their agreement.
I also asked the Association for Research in Vision and Ophthalmology (ARVO) whether they would be moving the presentation forward from Sunday. At first, they too passed the buck to NEI:
ARVO is hosting this presentation as a special session, so it is outside our program. The embargo is being maintained by NEI. I’m not able to supply further information and suggest you contact the NEI press office.
Once it became clear that NEJM was going to release the study early, ARVO confirmed their plans — but still left things to the NEI:
At this point, the presentation at ARVO is still planned to go ahead on Sunday, which is the first day of the meeting. It’s important to NEI and the investigators involved that they present the findings in detail to their peers. The presentation includes time for discussion/Q&A. The ARVO Annual Meeting is the most viable venue for this.
NEI may or may not do a press briefing ahead of this, but the ARVO special session, while open to the press, is aimed primarily at ARVO attendees.
I’ve asked NEI whether they plan to investigate, and will update with anything I hear back. (See update at end.)
Regardless of the NEI’s plans, however, this episode highlights a few uneasy truths about the embargo system and the Ingelfinger Rule:
Plausible deniability: Despite the fact that they both stand to benefit in revenue and exposure, NEJM and ARVO are trying to claim that what happened here was someone else’s problem.
I’m not discounting the frenzy that this probably threw both of their press offices into. In fact, I think NEJM made all the right calls by releasing this sans embargo as soon as they could — I understand they really didn’t have anywhere near a final draft until the moment they published it — and deciding to publish it at all, despite being the ancestral home of the Ingelfinger Rule. (I have a minor quibble with lifting an embargo a half hour before actually publishing the study, which is just sort of strange. But that’s minor.)
But the fact is that journals that will gain impact factor and reprint revenue from major trials like this one — and societies who allow researchers to use their prestige as a mark of approval, then benefit from the exposure — have more of a responsibility than just throwing up their hands. Journals like to do this with retraction notices too, we’ve found at Retraction Watch, and we don’t think that’s justifiable either.
There’s an Ingelfinger loophole: Even if the NEI ends up investigating this leak, and finding out who did it, what will they actually be able to use as a sanction? That person wasn’t an author, so NEJM would be acting in an “extra-judicial” way if it punished him or her by taking away the right to publish there. (And, by the way, I’m not encouraging sanctions — I like leakers a lot and think they’re important for a fully-informed public.)
We’ve seen this before, just last year, in fact, with David Graham and a paper on Avandia that he eventually published in JAMA but that was leaked to Pharmalot ahead of time. Because JAMA didn’t think Graham had leaked it, they didn’t invoke Ingelfinger. And as far as I know, no one has ever faced sanctions for this.
In other words, in both of these cases: Although the original purpose of Ingelfinger was to prevent researchers from touting their work to the press before it was peer-reviewed — an understandable goal — it apparently doesn’t matter if someone else touts your work to the press before it’s peer-reviewed.
This allows journals to avoid invoking Ingelfinger. And that’s instructive. I can tell you, as can hundreds of science and health reporters around the world, that the rule strikes enough fear into researchers that many will refuse to talk to journalists, even if they’ve just spoken about their work in a public forum. And yet journals find ways not to use it — the same way societies and journals often find ways not to sanction embargo breakers, a trend I’ve been meaning to take a harder look at too.
Still, journals haven’t done much to dispel any fears about Ingelfinger. Maybe this is their version of Don’t Ask, Don’t Tell. Ingelfinger’s specter helps maintain their stranglehold on scientific data at least as much as any particular reality of it. That inconsistency, however, is another reason I think Ingelfinger must go, in the name of the free flow of scientific information.
There’s a word I thought about using to describe this, which begins like Ingelfinger does but ends in a way that wouldn’t be very family-friendly. Embargo Watch readers are clever. You’ll work it out.
Need a hint? Take a look at this T-shirt design apparently inspired by another Embargo Watch post on the subject.
Update, 2 p.m. Eastern, 4/29/11: NEI responded to my request for comment as follows:
We are still sorting out what happened and it would be premature to speculate on any next steps at this time.