New England Journal of Medicine editor Jeffrey Drazen has chosen an ominous course for his journal. JunkScience.com ended the career of ex-Journal of the American Medical Association editor George Lundberg for dabbling in politicized junk science. We could use another scalp.
Last week, the NEJM ran letters related to its June study/editorial on PM2.5. Yesterday, I wrote to Drazen asking him to retract the response from the study authors because it was just a total lie. My letter is below, followed by Drazen’s response and my response to his.
October 17, 2017
Dr. Jeffrey M. Drazen
The New England Journal of Medicine
10 Shattuck Street
Boston, MA 02115-6094
Re: Request for Retraction of letter by Di, Dominici and Schwartz and renewed request for retraction of related article and editorial
Dear Dr. Drazen,
I am writing to request that you retract of the letter by Qian Di, Francesca Dominici and Joel Schwartz appearing in the October 12, 2017 issue of the New England Journal of Medicine on the basis of its flagrant dishonesty and continued scientific misconduct. This dishonest letter further highlights the dishonest nature of their related study and your NEJM editorial. Accordingly, I am renewing my earlier request (included below) that the study and editorial be retracted.
The relevant part of the Di-Dominici-Schwartz (DDS) letter is as follows:
First, DDS is dishonest in claiming they ignored contradictory studies because of the existence of “hundreds” of studies they “might have cited” in support of their study.
As you know and to quote Albert Einstein, “No amount of experimentation can ever prove me right; a single experiment can prove me wrong.”
It only takes a single well-conducted study to show that an entire body of literature or thought is wrong. That is, in fact, how science works. Science is not a competition to which which viewpoint has more publications, citations or awards. If science was decided by the number of publications, etc., science would never advance.
Additionally, some of the ignored null studies are recent studies (i.e., published in 2017). It is certainly possible for a study published in 2017 to invalidate hundreds of studies published before 2017. And if you didn’t know this already, the reason there are “hundreds” of studies allegedly supporting the view of DDS is because the U.S. EPA has spent hundreds of millions of dollars purchasing dubious “science” (including the DDS study) to advance its regulatory agenda. By the way, the NEJM ignores this blatant conflict of interest when it comes to government-funded studies yet applies that same standard to pharma industry research. You pretend that government-sponsored research has no agenda that might affect reported research outcomes. But let’s back to the DDS letter.
DDS next misleads readers by stating, “Enstrom points to some null studies we didn’t cite” [Emphasis added] But, of course, DDS didn’t cite ANY null studies. DDS did not allude to the existence of null studies, or even the possibility that null studies exist. This is entirely dishonest.
Additionally, you have personal knowledge that contradictory studies exist, yet you allowed DDS to get away with giving readers this false impression.
Third, the DDS comment that they ignored contradictory evidence because their study was “not a meta-analysis” is a non sequitur that can only be considered a dishonest effort to distract readers.
No one suggested that a meta-analysis be conducted. Enstrom mostly pointed out that contradictory studies were entirely and inexplicably ignored by DDS. Although you should (and likely do) know better, you allowed this dishonest strawman argument by DDS.
Fourth, the DDS comment that their study was a “nationwide… assessment of exposure with high spatial resolution” is another non sequitur and strawman argument. How exactly does the guess-timation of PM2.5 exposures negate the obligation to address contradictory evidence? Answer: It doesn’t.
Fifth, DDS claims they didn’t control for smoking because a “sensitivity analysis” showed it would be “unlikely” to confound the association. This claim is misleading because the area where smoking does have an effect is in exposure to PM2.5.
The Medicare population studied is a generation of Americans that was known to smoke heavily and be exposed to quite a bit of ubiquitous secondhand smoke. A single cigarette can expose a smoker in five minutes to 100 days worth of the PM2.5 contained in average US air. So the Medicare population’s exposure to PM2.5 in outdoor is dwarfed by its smoking/secondhand smoking experiences, not to mention other occupational and lifestyle exposures to PM2.5. The notion that smoking history doesn’t matter is simply false.
Sixth, DDS claims that other PM2.5 meta-analyses support their results. This claim is of course irrelevant as a response for why they ignored to even mention the existence of contradictory evidence, much of which is recent and post-dates the meta-analyses. Also, just because a meta-analysis or any study is published, does not mean it is valid.
Seventh — and possibly most dishonest, if that is even possible in this case — is the DDS claim that they “reported strong not weak associations.” Since when is a hazard ratio on the order of 1.08 “strong.” As pointed out in my earlier request for retraction, actual epidemiologists disregard hazard ratios under 2.0 because they are unreliable. Given all the uncertainties in the DDS study data (exposure and health-related), a hazard ratio of 1.08 is itself a null result. That the DDS hazard ratio has been calculated in the context of a large population is irrelevant, especially since there is no physical or medical evidence that inhalation of PM2.5 causes death — another bit of contradictory evidence omitted by DDS.
The whole sorry incident reflects poorly on you, your colleagues, your journal, and all at the Massachusetts Medical Society who allow you to operate in such a shoddy, non-scientific manner.
Referencing the ongoing saga of Harvey Weinstein, I suggest you take a good look at what happens when the conspiracy of silence comes to an abrupt halt.
Save yourself and your journal — retract the letter, study and editorial before the junk science bubble bursts all over you.
Committee on Publications of the Massachusetts Medical Society
NEJM Editorial Board
Here is Drazen’s same-afternoon response:
To which I immediately responded as follows:
Your response is dishonest.
This dispute is not about the validity of the contradictory studies. This dispute is about whether they should have been mentioned in the original study and editorial.
If the study authors had mentioned, discussed and dismissed the contradictory studies, then the point might be disagreement over interpretation as you suggest.
But the study authors pretended that the contradictory studies and other evidence didn’t exist. They do. Those omissions are dishonest. Their letter is an effort to cover up their dishonesty.
You have, of course, aided them in all this. Possibly you should recuse yourself. You are driving your journal into a wall. Think Harvey Weinstein.
The battle lines are drawn.