University of Washington IRB Demanded Dangerous Consent Form

The recent Nature story on ethics consultancies includes an example of counterproductive interference by an intransigent IRB.

[Dolgin, Elie. “Human-Subjects Research: The Ethics Squad.” Nature 514, no. 7523 (October 21, 2014): 418–20. doi:10.1038/514418a.]

Amy Hagopian, a global-health researcher at the University of Washington in Seattle, found herself turning to an ethics consultant for help with a study in Iraq to find out how many people had died as a result of the US-led conflict that began there in 2003. Her team needed to obtain informed consent from participants, but the researchers on the ground in Iraq were concerned that including the University of Washington's name on the consent forms — a requirement for IRB approval — would make it difficult to get the data they needed. “They feared that being associated with American institutions would get them killed”, says Hagopian. “They dug in their heels and refused” to carry the form.

Hagopian wanted to strip the university's name from the consent document, but the IRB insisted that it was an important part of informed consent, which is meant to protect participants, not the investigators. The impasse brought Hagopian and her team to [ethics consultant Benjamin] Wilfond. He concluded that it would be ethical to remove mention of the institution, for three main reasons: first, research subjects would also be placed at risk by signing a document linking them to the University of Washington; second, apart from the link to the United States, the research involved minimal risk to the participants; and third, the study would not happen unless the name of the institution was removed.

The IRB eventually agreed with Wilfond. The researchers went ahead with the study and found that nearly half a million people had died from causes attributable to the Iraq war between 2003 and 2011 — a figure much greater than most previous estimates. “We couldn't have done this without him,” Hagopian says of Wilfond.

Of course, PRIM&R claims that "IRBs neither interact with subjects nor write consent forms."