Central vs. local IRB for multisite research

University study compares efficiency of both

The research oversight system, says a University of North Carolina at Chapel Hill researcher, has not evolved to keep up with the volume and complexity of the research it oversees. In the 1970s and 1980s, the research community consisted of a handful of single-site, single-researcher studies. Now, phase 3 studies can consist of thousands of subjects at hundreds of sites.

“Multicenter research is essentially a take-it-or-leave-it proposition for individual sites by the time we get our hands on it,” says Daniel Nelson, MS, CIP, director of the Office of Human Research Ethics and professor of social medicine and pediatrics at UNC-Chapel Hill. “It is predicated on local review of single sites.”

The premise is that the protocol will be performed identically at all research sites. If one site finds major issues with the protocol, that site will likely be excluded altogether, Nelson says. There are few effective means to modify the underlying protocol, he says, and local IRBs “are left tinkering around the edges with the things they can control.”

“The net result is ineffective oversight of study-wide issues by a patchwork quilt of independent sites,” Nelson says. “It’s broadly accepted that redundant IRB review by dozens of local IRBs is not commensurate value added.”

UNC-Chapel Hill is considering a policy change to address some of these issues. Nelson and colleagues conducted a randomized, controlled study to compare the use and efficiency of a central IRB versus a local IRB for approving multisite research. The aims of the UNC-Chapel Hill pilot project were to test a model that allows reliance on any central IRB involved with a multicenter clinical trial, Nelson says. The sample involved 43 consecutive industry-sponsored multisite trials, and UNC investigators and IRB members and staff were all blinded to the status of the individual studies. The studies were reviewed by the UNC Biomedical IRB, then randomized. Experimental investigators were given permission to register with the central IRB for that study.

To be eligible for the study, central or independent IRBs had to be accredited by the Association for the Accreditation of Human Research Protection Programs (AAHRPP), be in good standing with the Office of Human Research Protections (OHRP) and the Food and Drug Administration (FDA), and be willing to establish a master service agreement with UNC.

During the study, UNC-Chapel Hill collected data on the effect on the local IRB workload; administrative issues with dealing with multiple central IRBs; researcher satisfaction and feedback; turnaround time; and contingencies identified by the local IRB on studies already reviewed or approved by central IRBs. In all, 20 central/independent IRBs were eligible for the pilot study. UNC executed agreements with 13 IRBs, and in the end eight IRBs were used.

Over a six-month period, 43 studies that had 32 sponsors and eight central IRBs were reviewed. Researchers found a potential time savings of around 20 days per trial, provided standing agreements were already in place. They also found that of the 260-300 full trial reviews at IRB meetings, one-third of initial reviews conducted by the Biomedical IRB are potentially eligible for outsourcing. Eighty-five percent of study researchers felt that reliance on a central IRB would ultimately speed up the process, and any problems encountered would resolve as the model as more experience with the model is gained.

“Our desire to explore an alternative to the traditional ‘sole provider’ arrangement was well founded,” Nelson concludes.