James Watson (not the racist dude who, in 1998, said that a cancer cure was coming in 2 years) writes:
About a month ago, when the infamous Lancet hydroxychloroquine/chloroquine paper was still “real science” (i.e. in the official scientific record), we decided to put extra pressure on the authors by writing an open letter to the New England Journal of Medicine. Prof Mehra and his Surgisphere colleague Sapan Desai had published a paper derived from the same “database” in the New England Journal of Medicine a few weeks before their Lancet publication. This was an important but rather boring paper (ACE inhibitors not dangerous in COVID). I guess because it did not have immediate ramifications, none (to our knowledge) had critiqued it openly. But on close examination it didn’t take long to see glaring inconsistencies. Numbers from Turkey and the UK (they made the mistake of giving country-level numbers!) were basically impossible, and the relationship between age and mortality didn’t agree with later reports. The open letter can be found here. A few hours after we posted this letter online NEJM posted an expression of concern re Mehra et al, and hours later Lancet did the same (I’m not claiming full causality but it could be argued!). A few days later both papers were retracted.
On the 3rd of July we got the (expected) reply from NEJM:
Dear Dr. Watson,
I am sorry to inform you that your submission, “Expression of concern regarding data integrity and results,” has not been accepted for publication in the Journal. It was evaluated by members of our editorial staff. After considering its focus, content, and interest, we made the editorial decision not to consider your submission further. We are informing you of this promptly so that you can submit it elsewhere.
Thank you for the opportunity to consider your submission.
Eric Rubin, MD, PhD
I’m not sure that one month counts as prompt given the importance of the concerns. And given that 174 researchers/clinicians signed the letter, from around the world. I can understand people not liking these massive group signatory letters, but this aspect is important in my opinion. Every statement was carefully reviewed by each signatory – many changes were proposed and made. Most of the signatories are individuals actively working in COVID-19 and the fact that none of them had heard or knew of Surgisphere activities gave considerable confidence that it was all fraudulent.
It’s a shame that NEJM didn’t publish our letter. I don’t understand why they can’t openly accept that a mistake was made. Publishing the letter formally records the worrying patterns that point to pure data fabrication and fraud. And it highlights that Prof Mehra and co-authors made a serious error appending their names to a publication with no knowledge of data provenance. In The Lancet it is stated that the corresponding author (Mehra) and coauthor ANP (Patel) had full access to all the data in the study and had final responsibility for the decision to submit for publication. Mehra and Patel must have signed this – yet when retracting the paper they claimed they did not have access to the data! (and asked a third party to audit the data).
As discussed many times on your blog, this shows a major problem in academic publishing: there are few (any?) incentives for post-publication review . And journals don’t take post-publication review seriously. Journals like NEJM and Lancet appear to believe that 2-5 anonymous reviewers will be better at spotting problems than hundreds/thousands of individuals who read it after publication. This is obviously silly. So why not encourage post-publication review instead of stamping it out?
I recently came across this somewhat remarkable example of how post-publication might happen with a responsible and modern journal, not surprisingly it’s from PLoS.
A very nice example of transparency.
Further background here.
It’s that never back down “culture of poverty,” attitude. But . . . the editors of the New England Journal of Medicine are not poor people. They have no excuse!