Red Flags: Citation Kiting, etc.

Citation Kiting: The industry’s “Authoritative” prevalence has no external, contemporary source.

Dr. Kay Dickersin[1] was the expert witness in the Pfizer/Parke-Davis affair. She detailed work on an organized “publication strategy” and the resulting exploitation of biases in published scientific papers. “Citation bias” was one of the abuses exposed in her report:  selecting only favorable citations and neglecting the unfavorable, previously published works. What we have with the FH publication strategy however is over-the-top. I call it “Citation Kiting.”

Bankers are on the alert for a crime called, “Check Kiting.” It takes advantage of the lag between the writing of a check and the final transfer of funds from the bank. So if a real or fake check for $10,000 is deposited into Bank A, another check for $10,000 can instantly be written against Bank A and deposited into Bank B.  During the time it takes for both checks to clear, the total of printed balances would be $20,000. Now, our fraudster can show a third party “proof” of $20,000 in assets. The scheme can involve many banks or other financial institutions, until eventually he comes across a bank with lax policies or personnel who break with policy. Or perhaps he can find someone who will lend him cash against his inflated bank statement.  He can even use the increasing balance of a target bank as “proof” of increasing revenue. Fortunately, there is a clearance mechanism in place for bank checks, forcing the fraudster to accelerate his pace as his balances swell, leaving it only a matter of time and a little research before this financial epidemic ends. 

But what if there were no clearing house for bank checks? How absurd would that be? And yet there is no clearance mechanism for the debt one science report owes to another. A similar “kiting” scam can happen within the hallowed corridors of science. Here is an excerpt from Novelion’s latest 10-K. (Emphasis mine.)

“Medical literature has historically reported the prevalence rate of HoFH as one person in a million, based on an estimated prevalence rate for HeFH of one person in 500. Analysis of HoFH prevalence have been evolving in recent years cumulating in published medical literature that suggests that the actual prevalence of both HeFH and HoFH may be significantly higher than the historical estimate of one person in a million. For example, in 2014, the European Atherosclerosis Society (EAS) Consensus Panel on Familial Hypercholesterolaemia (FH) published an article citing research that would result in an estimate of the prevalence of HoFH in the range of between one person in 300,000 and one person in 160,000 or 3.33 persons per million to 6.25 persons per million, which is consistent with estimates that can be derived from other publications from the last few years. The FDA cited this estimate in its review of PCSK9 inhibitor products in June 2015.” ~ (emphasis mine) Novelion (acquired Aegerion) 2016 10-K[2]

What concerns us here is the citation trail. The EAS paper cited by Novelion in the above did no prevalence study of its own. It cites another “Authoritative” report,[3] which also did no prevalence study of its own. This Authoritative report did not even use the results of the study it cited.  Consequently, there was no external, contemporary source in this Authoritative report for the new result and new criteria. The only source to the corrected prevalence rate is a self-citation within this very, selfsame, Authoritative report, found in a caption to an illustration, referring to a “personal communication” with the lead author: Borge Nordestgaard. Did someone actually talk to him? … or did he simply talk to himself?

Chronology of European publications with citation mapping

Links of Communication Dependence: Citation Kiting

There was no external, contemporary source in this Authoritative report for the new result and new criteria.

Chain of information dependence

There is no “citation clearing house” – even when human health and billions of dollars are at stake. The 1st report has vanished from concern … the source.  Below, we can see the first step in the removal of the prevalence conclusion from its own source.

The new doubled FH prevalance has no external contemporary source

From here on the Authoritative report is cited in the industry as if it were the source material itself. As of November 2016, Google Scholar lists 583 citations:[4]

The Nordestgaard report is widely cited

This is a multibillion-dollar industry. And the new prevalence and new criteria rests on a personal communication, with the selfsame lead author?  There is no external, contemporary source for the correct prevalence result.

Here is a circumstantial but scary thought: Dr. Peter Gotzsche of the Nordic Cochrane Centre wrote a report on ghost authorship in Danish scientific research, having found that 75% of randomized trials over a 2 year period involved ghost writers.[5]  Who consulted with Dr. Nordestgaard? Again, this is only a circumstantial juxtaposition of facts, so caution is warranted.  But here is Dr. Peter Gotzsche again,

It is still commonly accepted that department chairs claim authorship of all papers emanating from the department, and newspaper articles celebrating a professor’s 60th birthday may note that he or she has written more than 500 papers. The professor may have contributed, but almost certainly wrote only a minority of them.[6]

It is a perfect game. If all goes well, the professor simply says nothing and absorbs all the prestige. If it goes badly, he simply says, “It wasn’t really me.”  Now, I can’t say that Dr. Nordestgaard simply hired out his name and that ghost writers have done much of his work. I can say that the University of Copenhagen credits him with 438 scientific publications.[7] For the last couple of years he has been averaging about 1 per week.

But here is another circumstantial fact which may or may not have place in this jigsaw puzzle: Dr. Adriane Fugh-Berman, who was expert witness in the Wyeth trial, helped expose the widespread practice of ghost writing in medical literature.  Here is the first summary point of her paper, “The Haunting of Medical Journals: How Ghostwriting Sold ‘HRT’”:

“Some 1500 documents revealed in litigation provide unprecedented insights into how pharmaceutical companies promote drugs, including the use of vendors to produce ghostwritten manuscripts and place them into medical journals.”[8]

The new prevalence count in the Authoritative report has no external, contemporary source. And yet now the medical community is reporting that there is a higher prevalence than previously thought and it is this Authoritative report, itself, that is widely cited as if the actual source of the “discovery.”  But that would mean that the industry accepted an obscure comment in the caption to an illustration which doubled the prevalence of a serious disease, doubling also the addressable market for a multibillion-dollar industry … without blinking.

Other facts can be blinked out of the source material. The studies in Denmark were of whites of Danish descent.  At the end of successive research papers this restricted population was eventually declared to be a study of the “general population.” Here’s how the same study of “whites of Danish descent” becomes a study of the “general population.”

Evolution from specific to general Nordestgaard et al

Only Benn, et al, performed a prevalence study. The following two reports simply cited Benn.  Cuchel, by citing Nordestgaard and not the original, can take Nordestgaard’s statement of “general population” and omit mention of Benn’s original disclosure of “all whites of Danish descent.” Now as if representative of the USA, we move from a very specific slice of demographics, to something which represents all of the USA, with one and the same source population: whites of Danish descent. These studies were funded by Big Pharma players and were promoted in the USA. Nordestgaard’s and Cuchel’s papers have been highly influential, and have been cited by the FDA. And then in the next step, the FDA is cited by Novelion (formerly, Aegerion) to add credibility. A full circle is completed.  Pharma punches in the very information and credibility it needs to pull back out. (The emphasis is mine.)

“… in 2014, the European Atherosclerosis Society (EAS) Consensus Panel on Familial Hypercholesterolaemia (FH) published an article citing research that would result in an estimate of the prevalence of HoFH in the range of between one person in 300,000 and one person in 160,000 or 3.33 persons per million to 6.25 persons per million, which is consistent with estimates that can be derived from other publications from the last few years. The FDA cited this estimate in its review of PCSK9 inhibitor products in June 2015.” ~ Novelion Therapeutics Inc., Dec. 2016 10-K[9]

Now for the industry, and even the FDA, the original source material has virtually disappeared. Real money has been lent, borrowed and invested based on these reports and subsequent citations of them.  But there’s more. I will demonstrate mathematically in the second half of my analysis that the actual source for the above swapped out genuine mutation carriers and swapped in false positives. These false positives have been prescribed risky, new drugs. This is citation kiting: the industry draws credibility from an FDA citation, which drew credibility from the Authoritative report and yet the Authoritative report is an empty account drawn against an original empty account. They may even be ghosts which are “writing checks,” as it were, against this original phantom balance. Who can know? 

  • Question: Why does the industry cite the Authoritative report and not the Corrigendum to the 1st Report?
  • Answer: The 1st Report was clearly manipulated. See the pages: here, here, here, here, and here.

It is a simple mechanism, like a rudder to a ship: citation kiting steers the phantom “source” cited by the industry away from the genuine source.


[1] Expert witness in Prizer trial: “Reporting and other biases in studies of Neurontin for migraine, psychiatric/bipolar disorders, nociceptive pain, and neuropathic pain,” Dr. Kay Dickersin, MA, PhD,August 10, 2008

[2] https://www.sec.gov/Archives/edgar/data/827809/000082780917000012/nvln-12312016x10k.htm#sD4A2BC7F37462DC63C36481CE7B73D5C

[3] Click here for a list of key reports, including the “Authoritative” report.

[4] https://scholar.google.com/scholar?as_sdt=1,48&hl=en&sciodt=0,48&cites=5902718471118150349&scipsc=

[5] Ghost Authorship in Industry-Initiated Randomised Trials” http://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.0040019

[6] “What Should Be Done To Tackle Ghostwriting in the Medical Literature?”, Peter C Gøtzsche, Et al, http://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1000023  

[7] http://research.ku.dk/search/?pure=en%2Fpersons%2Fboerge-nordestgaard(733103b5-ddc2-416f-a686-48e694154a82)%2Fpublications.html

[8] http://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1000335

[9] https://www.sec.gov/Archives/edgar/data/827809/000082780917000012/nvln-12312016x10k.htm