1 Boring Old Man

Fraud in the Scientific Literature
New York Times
Editorial
October 5, 2012

A surprising upsurge in the number of scientific papers that have had to be retracted because they were wrong or even fraudulent has journal editors and ethicists wringing their hands. The retracted papers are a small fraction of the vast flood of research published each year, but they offer a revealing glimpse of the pressures driving many scientists to improper conduct. Last year, Nature, a leading scientific journal, calculated that published retractions had increased tenfold over the past decade — to more than 300 a year — even though the number of papers published rose only 44 percent. It attributed half of the retractions to embarrassing mistakes and half to “scientific misconduct” such as plagiarism, faked data and altered images.

Now a new study, published in the Proceedings of the National Academy of Sciences, has concluded that the degree of misconduct was even worse than previously thought. The authors analyzed more than 2,000 retracted papers in the biomedical and life sciences and found that misconduct was the reason for three-quarters of the retractions for which they could determine the cause. The problem is global. Retracted papers were written in more than 50 countries, with most of the fraud or suspected fraud occurring in the United States, Germany, Japan and China. The problem may even be greater than the new estimates suggest, the authors say, because many journals don’t explain why an article was retracted — a failure that calls out for uniform guidelines.

There are many theories for why retractions and fraud have increased. A benign view suggests that because journals are now published online and more accessible to a wider audience, it’s easier for experts to spot erroneous or fraudulent papers. A darker view suggests that publish-or-perish pressures in the race to be first with a finding and to place it in a prestigious journal has driven scientists to make sloppy mistakes or even falsify data. The solutions are not obvious, but clearly greater vigilance by reviewers and editors is needed.

From the PNAS study [see what he said…]:

^{[reformatted for clarity]}

In certain areas, we hold people to a higher standard: elected and legal officials, airline pilots, doctors, scientists, even our lawyers. When accused, we come closer to a standard of "guilty until declared innocent" than the opposite from our criminal system – a standard of "above suspicion."

The point here is that the predictable reaction to that editorial or the graphs from the PNAS article is to be horrified that the scientists are sinners, and that scientific sin is on the rise. We don’t expect that. From my perspective as someone who has spent a lot of time looking at the scientific literature produced over the last twenty years in psychiatry, my reaction is different. I see that graph as the tip of an iceberg, and the result of allowing scientists the latitude afforded criminals instead of applying a time honored higher standard. In the case of the article often discussed here [Paxil Study 329], no one questions that it is not just wrong, but it is wrong on purpose. It was shepherded into the literature by an unethical pharmaceutical company using a talented ghost-writer and endorsed by scientists who were hardly moral giants. They didn’t change the data – instead they mis-used the tools of science to twist it beyond recognition to reach a conclusion that was the opposed to its true meaning [on purpose]. And we know that’s true because we have all the details, including a $3 B confession by the pharmaceutical cartel that masterminded the dirty deed.

So why hasn’t Paxil Study 329 been retracted? A medley of le·gal·is·tic [lg-lstk] arguments [see The Rules of Retraction]:

• they didn’t fabricate or falsify the data
• "no evidence of scientific errors"
• "nor any justification for retraction according to current editorial standards and scientific publication guidelines"
• "It generated all sorts of useful discussion which is the purpose of a scholarly journal"
• "There are very many papers where, if you looked at the data, you could argue that the conclusions are not justified"
• "Many highly expert child and adolescent psychiatrists were participants in that study, and others that were similar, and others equally expert contributed to the review process"

Le·gal·is·tic [lg-lstk] arguments rely on precedents [and sometimes on hypotheticals] and in this case, there really aren’t any solid precedents because these things don’t directly end up in courtrooms, at least not yet. There are just some guidelines scattered around with nothing like "case law." But the ramifications do end up in court – like this recent settlement with GSK or the numerous Civil suits won in part on the evidence of the misbehavior in this particular article. And Study 329 is not alone in the psychiatric literature – only a well-known example of the conscious, willful, and deceitful use of scientific methodology by a commercial enterprise in concert with academic physicians to defraud help-seeking patients by giving their physicians false information without fabricating or falsifying the data.

So we need a precedent. We need for someone to step up to the plate and do the obviously right thing, retracting the article, ignoring the le·gal·is·tic arguments as inappropriate for a respected scientific community. The only reason that Medicine is respected is that it is believed to operate on a higher ethic than the courts of crimes and misdemeanors. And who should move for retraction? The American Academy of Child and Adolescent Psychiatry. Paxil Study 329 was published by their official journal, and the vulnerable population they treat and represent were the inured party. It’s their precedent to set.