Soon, all IRBs will need to be ready to contract with central IRBs on cooperative research studies. While many IRBs already participate in reliance agreements, they say it’s becoming more complex and will take time and practice to prepare organizations for the big change.

The revised Common Rule, published in January 2017, requires U.S.-based institutions to use a single IRB for NIH-funded multisite studies. Unless President Donald Trump’s administration changes or suspends the rule, IRBs and research institutions have until Jan. 19, 2020, to implement the change.1

Though the revised Common Rule states that relying on a single IRB will reduce institutions’ burden over time, some say this can be complicated and time-consuming — at least in the short term.1

The Emory University IRB in Atlanta has been relying on a single, independent IRB for several years. “We found we were not adding as much value as an institutional IRB, and we wanted to focus more on investigator-initiated studies and federally funded studies,” says Rebecca Rousselle, CIP, IRB director at Emory University.

Most of the institution’s single-study reliance agreements were for data analysis. But as organizations prepare for the Common Rule’s 2020 deadline, this is changing.

“For clinical trials involving Emory patients, we’ve always chosen to review these ourselves until now, and they just require a lot more attention even when we’re relying on a single IRB,” Rousselle explains.

When the University of Wisconsin–Madison became involved with research with the Inner City Asthma Consortium, the IRB learned that relying on a central IRB requires substantial administrative support and continuous analyses of efficiency and effectiveness, as well as ongoing education and infrastructure support.1

The university was asked to be the IRB of record. “We said, ‘No, we’re not ready yet to handle that. There are many studies with many sites,’” recalls Brandy Stoffel, JD, LLM, IRB facilitator for both the University of Wisconsin–Madison and the Inner City Asthma Consortium.

Relying on a single external IRB is challenging, she notes.

“It’s hard for people to agree. They’re scared and uncertain,” Stoffel says. “Because our university already had a relationship with Western IRB, we thought that might be the best place to go.”

For all new grants, the consortium relies on an independent IRB of record. Studies that are older and ongoing stay with their local IRBs, she adds.

The same kind of decisions could be made with any independent or central IRB as multisite studies increasingly look for a single IRB of record. The key issue now is how local IRBs can structure their workflow and processes to handle these changes.

“We had a robust process for studies that were reviewed by WIRB that we worked out a long time ago,” Rousselle says. “We had certain people in our office who would handle the administrative and local Emory part of study review, and then WIRB would handle the ethical review with our input from the local context.”

From that experience, the Emory IRB developed a special process with the National Cancer Institute IRB, which was different from its process with the independent IRB. For example, the National Cancer Institute does not perform a privacy board review, so Emory had to work out a way to conduct that review for those studies, she adds.

Those experiences were positive and worked out well within the Emory IRB’s workflow. But beginning last year, things began to get complicated. From working with just a few central IRBs, the institution now works with about a dozen central IRBs. These have included Partners, UCLA, Utah, One IRB, UNC in Chapel Hill, and Children’s Hospital of Philadelphia.

“We’ve gotten a growing number of requests to rely on other single IRBs, and it’s taken us a long time to get the agreements reviewed, negotiated, and to process the kind of local context review they request because each of their mechanisms retaining local context is different,” Rousselle explains.

The IRB had to develop a system to use its own electronic admission system along with the submission system used by external IRBs.

“We realized we definitely needed a reliance specialist,” Rousselle says. “Junior staff can handle the rest of the work.”

But a specialist is needed to negotiate agreements and interpret different local contexts. It’s challenging to keep track of reporting requirements and standard operating procedures for each of the central IRBs so Emory researchers won’t run afoul of any rules, she adds.

“We need to restructure our staff to allow for a reliance specialist to do that work,” she says.

Regardless of the revised Common Rule and reliance agreements, local IRBs ultimately will be responsible for any incidents at their institution, Stoffel says.

“I created a chart, saying, ‘What does your institution require for reportable event reporting?’” Stoffel says.

The chart shows study teams what to report locally and what to report to the central IRB.

So far, the reliance has worked well, Stoffel notes. “What’s instrumental in the success of this format is having one person — or at least a central location — for communication,” she explains. “Anytime any of our coordinators have a question, they can call me.”

The central IRB has an online submission system, and it’s able to send all approval documents to a distribution list. New sites automatically get all of the approved documents, Stoffel says.

As IRBs create more reliance agreements, they’ll find that their culture will change, she suggests.

“Over the last two years, I’ve seen culture changes by leaps and bounds,” Stoffel says.

IRBs must know which requirements are state laws, versus which are institutional and cultural rules. “Do we have a written policy of how the written assent is obtained, or do we rely on the central IRB to make sure all sites are doing it the same way?” she says. “I went with the most conservative site’s consent policy, and we all did that one — everybody agreed to it. It was also difficult to drill it down to one HIPAA template. I wanted it to be consistent across all sites.”

Another challenge is that each site has its own approved informed consent language. “This required language was not actual policy, and it wasn’t just culture,” Stoffel says.

Stoffel put all required language that is based on policy in a checklist each human research protection program uses.

“We took those and meshed them with NIH language, and we got them all approved,” she adds.

As more local IRBs enter into reliance agreements with external IRBs, the logistical details might become less difficult. For instance, Rousselle has noticed that the agreements seem to be merging toward a common set of terms.

“The more we rely on each other, the more we see these terms and agreements that look good, and the terms seem a little more homogenous, which is good,” she says. “We developed a checklist of things we need in a reliance agreement to make sure everything is in there, and it’s not as time-consuming now.”


  1. Federal policy for the protection of human subjects. Fed Reg. 2017;82(12):7149-7273.