The Researchers Who Forgot Their IRBs

What’s behind paper retractions? (4)
by Adam Marcus and Ivan Oransky, Labtimes 05/2011

Photo: zettberlin / Source: PHOTOCASE

What should happen to data obtained unethically?

In 1966, Henry Knowles Beecher, a prominent Boston anaesthesiologist, wrote a groundbreaking article in the New England Journal of Medicine about the sorry state of research ethics.

Beecher cited nearly two dozen instances, in which subjects and patients were subjected to questionable methods in clinical trials, and his article helped usher in a series of critical changes. One of those, a provision of the National Research Act of 1974 in the U.S., was the development of the Institutional Review Board (IRB), a panel of impartial members of the community, tasked with making sure that research protocols do not violate the rights of volunteers.

Obtaining IRB approval is an essential step for any clinical investigator who hopes to evaluate a new drug or test a theory of medicine on human beings. Even something as seemingly banal as administering a questionnaire on, say, a person’s foot-fungus history, requires the okay of an IRB. No IRB approval, no study.

IRB approval is also absolutely necessary for publishing. Any­one who thinks it’s not should try to get a manuscript into press in a reputable journal.

Thank goodness, then, that we’ve come such a long way since Beecher’s landmark paper. But have we really? Over the past year, Retraction Watch has covered two cases, one small, one enormous, involving researchers who misled journals about having received IRB approval for their studies. Both cases illustrate how horrifyingly easy it is for unscrupulous scientists to violate the public trust.

First, the small fry. In August, a journal called BMC Musculoskeletal Disorders retracted two articles from a group of Australian exercise specialists who’d looked at aspects of lower back pain in rugby and football players. The notice read:

“The journal has been informed by the authors’ institution that, contrary to the statement in this article, the Macquarie University Human Ethics Committee did not receive an application for ethics approval for this study. As the study was conducted without institutional ethics committee oversight, this article has been retracted.”

Note that the editors didn’t say that IRB approval hadn’t been obtained. Rather, it said the researchers not only didn’t have it but they lied about having it, too.

The second case is even more alarming. It involves Joachim Boldt, who, until last October, was considered one of Europe’s leading anaesthesiologists and an internationally recognised figure in cardiovascular anaesthesia and critical care in particular. But that month, the American journal Anesthesia & Analgesia retracted a December 2009 paper by Boldt that had two troubling issues: it appeared to contain falsified data, and Boldt couldn’t provide proof that he’d received IRB approval for the trial.

What followed was one of science’s more spectacular falls from grace. Boldt, it seemed, lacked IRB approval not only for that study but for dozens of others. He lost his job as head anaesthesiologist at the Klinikum Ludwigshafen, and faces massive fines and possible jail time in addition to the loss of nearly 90 articles from 18 specialty journals, due to retractions.

So, are these two cases likely to be the only situations where IRB approval has not been sought for a while? We would like to think so, but there is doubt. Speaking of the IRB situation in the U.S., Anesthesia & Analgesia editor in chief, Steven Shafer, notes:

“We have a system of private IRBs that are paid for by investigators, and those IRBs are hired to review protocols and either approve them or not. But for a hospital that assumes that investigators are using a private IRB – because they don’t have any IRBs themselves – there would be no way of checking up on this.”

Indeed, in a 2009 sting by the U.S. Government Accountability Office, one private IRB approved a study of a surgical adhesive gel, submitted by a company that didn’t exist.

In the end, scientific publishing relies on good will – and good word – of the people (emphasis there) who submit manuscripts. As we have described here, the loopholes (from forgery of authors’ names to manipulation of material) are too large to plug up, despite the best efforts of editors and publishing houses.

Certainly, violation of research ethics is a profound problem that journals must take seriously. But some observers raised a provocative and interesting point: What if a study lacks IRB approval but produces impressive (and reproducible) results? Is retraction really the only option? Should other researchers be denied the chance to learn from the tainted finding?

Do we need to remind readers of a certain group of researchers who conducted horrible, often fatal, experiments during the 1940s? Some scientists would like to use this data, too. They say, it should just always include some sort of disclaimer.

We respectfully disagree. Protecting vulnerable subjects in clinical trials is too important.

(The authors run the blog Retraction Watch:

Last Changed: 03.05.2012

Information 4

Information 5