For decades, institutional review boards (IRBs) have been responsible for ensuring that research conducted on people is ethical, and that trial participants are adequately protected. But independent IRBs, sometimes called commercial IRBs, have been accused of placing profits ahead of their mission. Since the pharmaceutical firms pay independent IRBs for their reviews and oversight of trials, so the thinking goes, there’s a conflict of interest that casts doubt on the integrity of the IRB’s decisions.
Advocates of university- and hospital-based IRBs have been making these arguments since the Food and Drug Administration first allowed independent IRB review in 1981. Our company, Quorum Review IRB, has been doing IRB reviews for 25 years. Over that span, the arguments have not changed much.
What has changed is clinical research. Local IRBs, sometimes called institutional IRBs, have traditionally had jurisdiction over one research project in one location. Today, clinical trials tend to have multiple investigators in multiple locations, all following the same protocol. In such instances, independent IRBs, which routinely oversee dozens or hundreds of sites for a particular trial, are better equipped to oversee such studies.
After decades of debate over conflict of interest versus efficiency, we believe it is time to discard the blanket criticisms based on where an IRB is located or how it is structured. The focus should instead be on finding meaningful ways to gauge the quality of all IRB reviews.
STAT reporter Sheila Kaplan wrote that an inability to assess the quality of IRB review is part of the problem. We disagree. It is the problem. The clinical research community has not developed criteria to measure whether an IRB’s reviews are “good” or “bad.” We have tried looking at how quickly IRBs review protocols; at whether they have earned accreditation; at whether audits find they comply with regulations; and at what kinds of conflicts of interest may or may not exist within one IRB structure, compared to another. None of these measures are adequate proxies for judging whether an IRB’s decisions consistently follow ethical standards for protecting human subjects.
Admittedly, the criteria for IRB approval can seem subjective: risks should be “minimized;” they must be “reasonable” in relation to the anticipated benefits of the research; and informed consent must be “understandable” to the subjects. In making these determinations, individual IRBs should not rely solely on the values and experience of their particular members but also make decisions based on broadly held societal norms.
The only way to do that is to create a shared record of ongoing decisions and their justifications. Continued use will refine such a cumulative record into an evolving consensus. This approach could move the discussion about IRBs away from protocol details or regulatory compliance and toward ethical principles.
In clinical medicine, choosing a course of action is a matter of craft, experience, and values. When caring for an individual patient, much of medical quality comes down to a doctor’s experience. So-called evidence-based medicine and the learning health care system are reactions to this reality and seek to move medical decisions beyond the expertise of an individual doctor.
In a similar way, we believe that individual IRBs and the IRB enterprise as a whole should become a learning system. We should build a body of justified ethical decisions and gauge performance on the consistency of IRB decisions over time, as well as how departures from past decisions are justified.
What could an IRB learning system look like in practical terms? Overall, it would involve sharing decisions between IRBs, as well as more open communication between IRB leadership. For individual IRBs it would mean gauging how consistent each committee’s decisions were over time and across studies. The current generation of IRB management software is designed to facilitate processes and make review materials accessible, but it is technically feasible to build systems that recognize the similarities between protocols and populations. That would let IRB members easily take past decisions into account as a foundation on which to build an ongoing ethical edifice.
Establishing a generally recognized measure of quality in IRB review is more important now than ever. The National Institutes of Health has announced that by May 2017 it will not allow multiple, redundant IRB reviews of the same protocol, but will require that a single IRB oversee all investigators at multiple sites. The same mandate may extend to other federally funded studies if proposed changes to the Common Rule come into effect.
Those policies may or may not specify whether an IRB must be located in a university, be part of a government agency, or be run by a privately owned company. But they will definitely require that the IRB be equipped to oversee trials with multiple investigators. The NIH has already said it will need assurances that the IRB chosen for a study is appropriate for the research and up to the task.
An assessment system that draws on the quality of decisions — not on speed, not on ownership, not solely on compliance to the letter of the regulations — can ensure that an IRB can be trusted to protect the rights and welfare of research participants in this new and complex environment.
Stephen Rosenfeld, MD, is the chair of Quorum Review IRB. Jim Gearhart is senior manager of publications and online content and a co-owner of Quorum.