Who Are Medical IRBs Really Protecting?

File this one under “games medical researchers play.”

In a recent essay for the BMJ (pointed out on Twitter by 1boringyoungman), Stuart Buck criticizes the Institute of Medicine for not being bold enough with regard to review of medical research data. The Institute of Medicine (IOM)—the medical branch of the National Academies—is calling for a new system in which “all NIH-funded or FDA-regulated clinical trials [must] share patient level data with independent researchers.”

Buck understands, as I do, that if this happens, it will be a very big deal. It will mean that the data from a very large number of medical research trials will become available to researchers unaffiliated with those studies. Independent researchers might very well have different perspectives, different skillsets, and different conflicts of interest than the original researchers, and this could mean that they could find all sorts of things in the data that would not otherwise be found. More transparency! More data analysis!

So what is Buck’s beef? This: The IOM doesn’t go far enough, because it is only calling for the policy to apply to futuretrials, not those already conducted. The IOM’s concern is that the subjects (i.e., people experimented on) who were in those older trials might not have signed consent forms that would allow for the sharing of their data with independent researchers.

True, they may not have consented to that specifically, but Buck asks us to “[r]emember, no one is talking about re-doing a trial and exposing new patients to new drugs. . . . For anyone worried about privacy, there is no more reason to be worried about the independent analyst than about whoever did the original trial, and it is trivially easy to protect privacy in the same way that the original trialists did—sometimes even more so, now that we have software that permits someone to analyze data without ever downloading or seeing an individual patient’s records. Requiring patients to consent a second time is therefore unnecessary and irrelevant.”

I’m with Buck, and find myself suspicious that what’s really going on here is, once again, medical institutions protecting researchers, not research subjects. Here’s one reason I’m skeptical:

All the time, Institutional Review Boards (i.e., ethics boards) at medical institutions approve what are called “retrospective chart review” studies. These are studies in which the medical records of patients who never consented to be research subjects are used to gather data that their doctor then publishes.

In Galileo’s Middle Finger, I mention a particularly egregious case of this—Cornell pediatric urologist Dix Poppas getting his Institutional Review Board’s (IRB’s) permission to do “retrospective chart review” to publish data on how well little girls could feel him touching their genitals years after he surgically “reduced” their clitorises for “psychosocial” reasons. (Read more here and here, if you can stand it.)

In cases like this, permission is given after the fact to share data collected from patients. In the case with Poppas, no one asked the girls or the parents in advance if it would be okay to do research-data collection on the girls’ genital sensations—none of these families agreed to be research subjects for this “sensory testing.” Poppas just did it, calling it “normal” medical care, collected the data in their charts, and then got permission from his ethics board to publish this data set he just “happened” to have. (By the way, I can’t find a single other pediatric urologist who does this kind of “clitoral sensory testing” on little girls as part of “normal” medical care.)

The way “retrospective chart review” works out in practice, patients get used for research they were never asked to be part of, and perhaps more importantly, no one is ever asked to make sure they are notified that their medical records have been used for a published study. They never know they became subjects of research.

Make no mistake: there are good reasons to allow retrospective chart review! If a doctor genuinely discovers something important in the course of her or his ordinary medical practice, it is critical that she or he be given permission to share that data in a peer-reviewed publication. But in practice, “retrospective chart review” appears to often be used to skirt around the system—because, as the saying goes, “it’s easier to ask forgiveness after than permission before.” (What IRB would have permitted Poppas’s study if he had applied to do this prospectively? None, I like to think.)

It seems ridiculous to me allow medical institutions to give passes on retrospective chart reviews—with no requirementeven to advise patients they unexpectedly became research subjects!—while saying it would be inappropriate to now open up to independent researchers old data sets from subjects who genuinely consented to the research.

Who, really, is this protecting?

A postscript: Buck’s bio led me to look up his organization, the Laura and John Arnold Foundation, which “strives to produce substantial, widespread and lasting changes to society that will maximize opportunity and minimize injustice.” Way cool!

Note: My book will be out March 10, but you can pre-order it now using the links at the top of the Kirkus advance review.