NCI's Central IRB saves time, money

Lower costs, faster views, and less variability of review time

Using the National Cancer Institute's Central IRB (CIRB) saved affiliated institutions money, staff time and time spent on review, according to a recent study of the costs and benefits associated with the CIRB.

The study found that affiliation with the CIRB was associated with more than $700 savings for each initial review conducted. Affiliated sites reported reviews were approved nearly a month faster than those at unaffiliated sites, and required an average of about six hours less research staff work.1

Todd Wagner, PhD, a health economist with the Veterans Administration Palo Alto Health Care System in Menlo Park, CA, says the results represent an important first insight into the workings of a publicly funded central IRB.

But he says the efficiencies demonstrated in this study might not necessarily extend to other central IRBs in the process of being formed or currently being discussed.

"The process that NCI does is a very different process than, for example, the one the VA is creating now," he says. "So despite our results, I'm not sure that they generalize."

Wagner's group used surveys to glean information from research staff and IRBs at institutions participating in trials run by the NCI's Clinical Trials Cooperative Group Program — both those who enrolled with the CIRB and unaffiliated institutions.

Unaffiliated sites conduct their own full-board local reviews of the trials, just as they would for any other study.

Those that enroll with the CIRB do a different type of review, called a 'facilitated review.' In a facilitated review, the local IRB chairperson or a small subcommittee of the IRB reviews the recommendations of the CIRB, which are posted on the CIRB Web site. They can choose to accept the CIRB as the IRB of record for that trial, and can get minor changes made to the informed consent to meet local requirements. If an affiliated institution rejects a particular CIRB review, the local IRB conducts its own full board review.

If the CIRB is chosen as the IRB of record for a study, it takes responsibility for all continuing reviews, amendments and serious adverse events reports (except that SAEs occurring at a local site must be reported to that local IRB).

Variability is key

Wagner's survey asked about the timing of reviews and the effort involved in conducting them. His group also calculated the costs of the reviews carried out at both types of sites.

The results:

For each initial review, a CIRB-affiliated site saves $717 — $321 of it in research staff savings and $396 in IRB staff savings. The average time that elapsed from initial paperwork to approval was 33.9 days less for CIRB-affiliated sites.

Wagner says CIRB affiliation also was associated with less variability in review times, a point that groups such as trial sponsors might actually find more compelling than cost

savings.

"It means it's more predictable," he says. "If you're a sponsor of trials, that might be really valuable information to determine how many trial sites you need. Variance is a very important result. In quality improvement, that's the whole goal, to remove the variance."

Jacquelyn Goldberg, JD, head of the NCI CIRB Initiative, says she was not surprised by the results.

"I expected the CIRB process to save both time and money," she says. "But also we were hearing it anecdotally. The time part we were hearing anecdotally from a lot of sites that were using it, what a time saver it is."

Wagner noted one interesting fact that surfaced in the study: Some CIRB-affiliated sites appear not to be using the CIRB as it was intended. Those sites reported spending time on continuing reviews and amendments for the studies, despite the fact that the CIRB is supposed to be responsible for those procedures. The study suggests that there could be further savings if this issue is addressed.

Goldberg says her organization had been aware of this problem anecdotally as well. She says it's not simply a matter of new affiliates getting comfortable with the process.

"Some places were in it for a couple of years and doing an extra level of review that they didn't need to do," she says.

Goldberg says the CIRB is trying to address the issue through education.

"We do outreach in different environments such as the cooperative group annual meetings, PRIM&R. We have question-and-answer sessions," she says. "In these sessions we bring it up and we also refer people to the Web site to make sure they have clarity on it."

Different models

How will this study affect the ongoing discussion over centralized IRB review?

Wagner says that his study doesn't necessarily make the case for centralized review in general, because different central IRBs can have very different operations models.

"If there's any take-a-way from this, it's not that this model is right, because I don't believe that any one model is right," he says. "What I hope people see from this is there might be benefits to being creative. How do we collect data to show if this is making a difference? If we're really intrigued about improving quality, how do we take the next step and actually do that?"

Lowell Schnipper, MD, professor of medicine at Harvard Medical School, is a strong proponent of centralized IRB review because of its potential to speed up cancer research.

"We have for a long time been very dismayed in the clinical oncology community about the long time lags it takes to activate clinical trials," Schnipper says. "In the world without the NCI CIRB, each and every institution has to use its own review board to review a trial. Some of them are efficient and some are extremely slow."

But he says that he's continued to encounter opposition to the idea of central IRBs from local institutions both large and small. Studies such as Wagner's will necessarily convince those institutions to put aside their reluctance, he says.

"The one major sticking point that vice presidents for research of various universities have brought forward has to do with liability," Schnipper says. "The issue that I have observed more often putting a real stop to this progress is the issue of who's responsible when something goes wrong."

More IRBs will come around as they see central IRBs that work well and safely, he says.

"The more people see that this is not resulting in harm to patients, I think it will be a process of gradual acceptance. But it takes a very long time."

Reference

  1. Wagner TH, Murray C, Goldberg J, et al. Costs and Benefits of the National Cancer Institute Central Institutional Review Board. J Clin Oncol. 2009 Oct 19 (Epub).