Research on ourselves: Do IRBs make it difficult?

What IRBs do to facilitate research on their own operations — or not

It's a well-accepted position in human subjects protection circles that in order to improve IRBs and their relationships with investigators, it's important to actually study IRBs — how they work, how long it takes to do things, the knowledge and attitudes that members carry into discussions of various types of research.

And some researchers who have attempted to do so report encountering institutional obstacles — IRBs that don't respond to queries, that say they are "too busy" to participate or that actively refuse to answer questions.

Two researchers who recently have focused on studying IRBs provide a more nuanced view of IRB cooperation. They say it can take time and effort, but that it is possible to get useful responses from IRBs.

And they point to other obstacles that may impact the study of IRBs, including lack of funding and lack of standardization of data necessary to compare institutions to one another.

When it was suggested to Charles Lidz, PhD, a professor of psychiatry at the University of Massachusetts Medical School in Worcester, that he observe IRB practices at multiple institutions, he was skeptical.

"I was pretty reluctant when one of my colleagues first proposed this — 'Are you kidding? You think we can do this?'" Lidz says.

"I actually would say that we have had fewer problems than I would have thought," he says. "I'm trying to do observational studies, so we do things like audio-record meetings and interview reviewers and staff and so forth. We've looked at 10 major IRBs in our current study. And I would say that maybe we had to go to 13 or 14 to get that total of 10."

Todd Wagner, PhD, a health economist with the Veterans Administration Palo Alto Health Care System in Menlo Park, CA, has been studying IRBs for 10 years, looking at costs and efficiency of IRB review in various settings. He's sent surveys to hundreds of IRBs, and followed up, often multiple times, in order to boost response rates.

"Generally they've been terrific, they've been really supportive," Wagner says. "I have yet to meet a combative IRB director. In general, they want to do the right thing. And I think most of them realize that the data can really help them do the right thing.

"It just takes a lot of time to get the right metrics and the right data," he says. "And the more detail you want, the harder it is for them to respond. If you want exact dates, if you want very specific data or a specific protocol, it just takes more time."

Outside reasons

Lidz says that when IRBs have turned him down, it's generally for an outside reason. "Two of the responses were: 'We're in the middle of revising our structures' or 'We just fired our director,'" he says. "It wasn't that they were saying no, they were just saying that it was a difficult time."

And two sites, he says, had legal objections. Because of their agreements with pharmaceutical companies, they felt that they could not allow him to see certain protocols, even after Lidz promised to remove all specific information about them, including the names of drugs.

"Almost everybody else thought that was fine, but these places said no, that's not enough," Lidz says. "They were willing to have us do it as long as we wouldn't look at pharmaceutical company protocols, but we said we couldn't do that."

One strategy he employed to help build cooperation was to seek out collaborators at every university he approached.

"A couple of my colleagues knew ethicists at every major university in the country," he says. "I can't think of a single one where we couldn't find somebody pretty easily and they certainly helped get us access."

Wagner says he generally approaches a board through the main IRB contact number. From there, depending on the institution and the information that he wants, the process can vary significantly.

"Some say, 'We'll pass along your survey to the appropriate person, thank you very much,'" he says. "Some might say, 'The person who coordinates that panel is Maria, and here's her phone number.'"

When an IRB is slow to respond, it can take several reminders to prompt someone to get back to him. "The response rate really does improve when you go up to three reminders."

Wagner tends to collect a large amount of specific data — how long protocols take to work their way through the IRB office, the amount of hours spent on a particular protocol, etc.

One of the biggest challenges to his research is finding a standardized way of collecting this information, which may be tracked differently at different institutions.

"If you're saying, 'How long did it take to approve this protocol?' and the protocol was returned to the investigator for some deficiency, then resubmitted, some IRBs will consider that the submission date, vs. others that use the first submission date. It's apples and oranges at that point.

"We've had to work really hard to standardize that in the survey questions that we use," Wagner says.

He would like to see efforts to standardize the way that IRBs routinely collect and store that type of information.

"So that when people are using these protocol-tracking systems for their IRBs, those metrics are built in and they're standardized so that we could then figure out ways of extracting data," Wagner says. He notes that more standard measurements would help streamline other multi-IRB data collection endeavors, such as accreditation and efforts by the Clinical and Translational Science Awards (CTSA) groups.

Finding funding

Wagner himself has proposed a project to help create such metrics, but was unable to get funding. Lidz has encountered funding issues as well. Both men have received NIH funding, but Lidz says there are few other sources interested in this type of research.

Lidz is currently seeking funding for another IRB project. He wants to create an experimental IRB at one of the universities he studied in his previous research.

"They agreed to do a little subset of their reviews, according to different procedures, all of which of course have to meet the Common Rule," Lidz says. "Basically, we negotiated a set of rules with them.

"The director was extremely helpful with thoughts and ideas," he says. "He was very thoughtful about these things and very much on board with what we were proposing to do."

All that's left, he says, is to see if the funding comes through.