Thursday, February 5, 2015

Third year med student brilliantly teaches Integrity 101 to professors at Duke . . . and to all of us

I wish my field had something like the Cancer Letter.

Not sure how much this story has been chronicled outside of coverage in the Cancer Letter. Maybe the story has been widely disseminated and I just missed it before.

It is a truly impressive example of integrity. In a nutshell:

1. Duke third year medical student opts to do required third year research project in the lab of a rising star at Duke, Anil Potti.  Potti was collaborating with and working under the wing of a another, very prominent Duke faculty member running a large genomics program, Joseph Nevins.

2. Clinical trials were being designed and run--enrolling terminally ill cancer patients--on the basis of Potti's work.

3. The medical student becomes concerned about the research--based on reviewers' comments on a submitted manuscript--and acquires a sophisticated knowledge of the complex topic (analyzing large datasets from genetic studies and identifying markers of tumor susceptibility to specific therapies). His knowledge of the research rapidly advances to the point where he feels comfortable challenging the way the research is being conducted.

4. The student has enough confidence--and courage--to take his concerns up first with Potti and then, up to the next level, with Nevins, and then with the medical school leadership at Duke.

5. Remember, this is a third year medical student doing a required one year research project. He is planning to apply for a competitive residency program and not to go into research.  Potti is pressuring the student to get the manuscript out; if not to the journal that is asking questions, then to another one.  He has nothing to gain here by making waves and the paper would greatly bolster his chances of landing his desired residency.

6. The student next kicks his concerns up to Duke's associate dean for medical education and its director of student affairs. The former suggests that he reiterate his concerns to Nevins. The student then (March, 2008) writes a nuanced, clear memo of his research concerns. The Cancer Letter states, "According to top-tier biostatisticians who were asked to review these documents, his understanding of biostatistics was extraordinary for a med student—or even for someone with specialized training."  Impressive, no?

7.  Nevins "implored" the student not to take his concerns any further, according to the Cancer Letter, "because he did not want Duke or any other entity looking closely at the data underlying the clinical trials [that were ongoing and based on Potti's results]."

8. The Cancer Letter says:
Dr. Nevins validated many of Brad Perez’s [the student's] concerns by referring to the serious issues raised regarding Dr. Potti’s lab as “being somewhere along the spectrum between sloppy research and a difference of opinion to research fraud,” and that Dr. Nevins confirmed that he would “go back through each and every dataset that we have posted in relation to various publications to ensure that there are no errors.”
9. The student, having taken his concerns to Potti's sponsor (Nevins) and high up the chain at Duke then moves on.

And get this,
  • 10. He declines to resubmit his first author paper despite co-authorship with and reassurances from prominent Duke faculty members that all was above board and any errors would be fixed
  • 11. The student asks to have his name taken off every other paper from the laboratory--there are a number--on which he is listed as a coauthor.
  • 12. The student returns a merit award he received based on the research in question and turns down an opportunity to present at a major meeting
  • 13. He decides to repeat his third year of medical school rather than present or publish any results based on the flawed work in the Potti lab 
  • (and all this despite an upcoming application to a competitive residency program)

Epilogue:   The student wrote his letter of concern in April, 2008. According to the Cancer Letter, Duke administrators--aware of the letter--met in October 2008 to discuss Potti. Meanwhile, outside experts began to raise concerns. Nevins continued to defend the work and the ongoing clinical trials. Duke allowed the trials to continue.

In 2010 there was a slew of retractions of Potti's work--Lancet Oncology, Nature Medicine, NEJM, etc.--Potti left Duke, and the clinical trials were, late that year, finally halted. In March, 2011, Nevins admits "nonrandom data corruption."  The Cancer Letter has discovered that Duke--where Nevins still works--misrepresented its knowledge of the medical student's concerns and put out a false timeline regarding when they became aware of potential problems. Duke is currently being sued by participants in the cancelled trials.




Tuesday, January 20, 2015

Yet more from the NEJM

Just finished writing this post when I see yet another Perspective article in the New England Journal regarding the ACA.

The Perspective article details the enormous problems for Obamacare that  a win for the plaintiffs in the Supreme Court case King v. Burwell would entail. It is clearly important to proponents of the Affordable Care Act that the plaintiff's case be weak and that the defense prevails. Indeed, the strength of the plaintiff's case is of such great consequence for Obamacare that you would think that the editors would want to see that it was covered fairly.

Here is the sentence regarding the argument at the heart of the case: "The case hinges on enigmatic statutory language that seems to link the amount of tax credits to a health plan purchased 'through an Exchange established by the State.' According to the plaintiffs in King, that language means that consumers who buy insurance through federally run exchanges don't qualify for subsidies"

The authors cannot get around the fact that the law states that only participants in exchanges established by state governments are entitled to subsidies (contrary to how the subsidies are currently being distributed).  When the authors say that the wording "seems" to link eligibility for tax credits to buying insurance through a state--rather than the federal--exchange, what they mean is that the law does make that link: they just can't bring themselves to state so plainly, hence the weak attempt to introduce uncertainty.

Of greater interest, the NEJM editors also let the word "enigmatic" slip by. Surely they could have at least provided some hint that the rationale here is not at all enigmatic. Perhaps the best way to understand the rationale is listen to Jonathon Gruber, a key architect of Obamacare, plainly lay the rationale out (lately proponents of the bill have--for obvious reasons--tried to make Gruber's role disappear). Anyone paying attention to the ACA litigation--including the NEJM editors--must be aware of the relevant Gruber videos: they received a lot of attention when they were uncovered. The linkage of subsidies to state exchanges is logical, not mysterious: the purpose is to strongly incentivize states to set up exchanges. Relevant discussions and links to the pertinent Gruber recordings can be readily found through Googling: here is one and another and a third.

Dear NEJM: please think about the balance between functioning as a journal of political advocacy and your general credibility.

Sunday, January 11, 2015

The Lancet: much, much worse

(Continued from prior post:  The politicalization of two major medical journals: bad and much, much worse)

In contrast to the NEJM, the Lancet has long evidenced obvious biases. This has done much to severely damage its credibility. The Journal is run for Elsevier by Jon Horton, who seems to be a man of the European left. His editorial judgments have landed the journal in trouble before. For example, he published two politically-motivated studies from a group at Hopkins purporting to look at deaths in Iraq attributable to US military actions. These studies were later found to be seriously flawed and biased by politically-motivated funding.

Worse, though, is Horton's turning what should be a prestigious journal of clinical medicine into a platform for obnoxious anti-Israel propaganda. Recently, as just one example, he discredited himself and his journal by publishing a letter in the middle of the most recent Gaza war authored by a group of viciously anti-Israel activists. Among others, the authors included two who have worked with the former Ku Klux Klan member and current neo-Nazi David Duke and also a maoist, pro-Hamas, strident anti-Israel activist. Later, as the antisemitism underlying the letter became more evident, Horton claimed to be sorry for publishing it. He has yet to remove it from the Lancet's website.

The Lancet reflects poorly on Elsevier and, arguably, on the British medical establishment. Elsevier needs to rethink its staffing of the journal.

The politicalization of two major medical journals: bad and much, much worse

Bad:  
The New England Journal of Medicine purports to take no position, pro or con, on Obamacare. The Editor-in-Chief and Executive Editor have come out in favor, yet, somehow, despite their biases--they claim--the Journal itself has not staked out a position. This is risible.  I subscribe to and very much like the NEJM, but ever since the Affordable Care Act (ACA) was first proposed, the Journal has been flogging Perspectives and other articles written by advocates as if it were a journal of politics and political policy: routinely publishing articles that could just as easily be found on the Op-Ed pages of newspapers or in any number of--primarily left-leaning--political/policy journals. Yes, they have published articles written by skeptics here and there (I can recall only one right now), but basically the Journal has been pushing Obamacare since the get-go: it's not subtle.

No forum for pro-ACA writings would be complete without an essay or two from the egregious Jonathon Gruber. The NEJM's conflict of interest policies has spared the Journal criticism for not disclosing Gruber's financial links to the ACA: if readers took the time and looked closely enough, they could find the NEJM's mandated disclosure (there has typically been no disclosure elsewhere he has been published or quoted). However, I doubt that even NEJM readers who found the disclosures knew at the time of the magnitude of Gruber's financial interests in the ACA. These amounted to millions of dollars: $400,000 from the feds alone.  Would the Journal let a consultant being paid millions of dollars by Amgen write a perspective advocating for more Epogen use?  If so, would a hard-to-notice financial disclosure statement be adequate, or would perhaps a disclosure need to be included in the text?

One final question: now that we know that Gruber has intentionally obfuscated important facts about Obamacare and has attempted to deceive the Congress and US electorate on key policy points, does the NEJM need to go back and look at his articles?  I bet they would have a second look at, say, a clinical trial that was published by an investigator later have found to have published fraudulent manuscripts.

I really like the Journal. It has been painful to see the Journal become so politicized during the ongoing debate surrounding Obamacare. The claim cited above that the Journal has not taken a position is belied by the articles that make it into its pages.

Continued in next post: Much, much worse