The investigational review process

Unfortunately, as clinical research has expanded to become a commercial industry that conducts federally regulated research at a rate of $20 billion a year (1), the structure, operation, and guidance of IRBs have become problematic. Recently, concerns about IRBs have even led to formal FDA interventions. In 1999, for example, the FDA temporarily halted research because of inadequate IRB review and follow-up at two highly visible locations, Duke University (2) and the VA Medical Center of West Los Angeles (3). Without delving into the specifics of these two cases, it is fair to say that the current problems with IRBs are more a result of industry growth than of outdated regulatory provisions.

So how do IRBs operate, what issues have arisen as a result of research expansion, and how can the FDA address these various problems so that patient safety remains the overriding concern of clinical research?

The regulations
Formal requirements that establish IRBs, outline their functions, and provide directives for the scientific review of clinical trials are delineated in 21 CFR Part 56. These regulations are comprehensive and describe the composition and responsibilities of the boards. Specifically, in terms of membership, IRBs must contain at least five people of varying backgrounds, genders, and ethnicities who possess the “professional competence” to properly assess human research studies. Furthermore, the members must include one person who is concerned solely with the scientific aspects of research and one who is interested only in the nonscientific aspects. (As a result of this requirement, many IRBs include clerical members.) To eliminate conflicts of interest, at least one IRB participant must have no affiliation, professional or personal, with the institution or individual that will conduct the research. (Researchers who serve as members of IRBs must abstain from considering their own investigational projects.)

The regulations require that boards approve research plans for every site that will participate in a given trial. Even if a single protocol is to be conducted by several different investigators under review by the same IRB, that IRB must approve the protocol for each individual researcher. To grantsuch approval, board members must jointly agree, for a specific trial and its investigator, that

• patient risks are minimaland reasonable with respect to anticipated treatment benefits;
• enrollment criteria are appropriate and patient eligibility is determined without regard to race, gender, economic background, and other such factors (unless appropriate and necessary for the trial, such as recruiting only females to test birth control pills);
• informed consent documents meet all regulatory requirements;
• patient consent will be obtained properly;
• patient safety will be monitored continually;
• collected data will be reviewed for completeness and accuracy; and
• confidentiality will be maintained for all participants.

The requirements for the ongoing review of approved trials are somewhat vague. 21 CFR 56.109(f) states, “An IRB shall conduct [a] continuing review of research . . . at intervals appropriate to the degree of risk, but not less than once per year, and shall have authority to observe or have a third party observe the consent process and the research.” Typically, IRBs act on this regulation by requiring that investigative centers provide annual or semiannual written reviews of their projects. These reviews generally include listings of the number of research patients screened, enrolled, and terminated, reasons for screen-failure and study termination (if appropriate), listings of serious adverse events that occurred at the reporting site, and brief overviews of the trial’s progress. In addition to the reports, sites must notify their IRBs immediately of all “serious” adverse events and provide them with copies of all “serious” events reported by other investigative centers conducting the same trial.

The problems
The origins of research oversight date back to the Nuremberg Code of the 1940s, but IRBs were not formally created in the United States until almost 30 years later, after the disclosure of the Tuskeegee syphilis study (in which African Americans with syphilis were denied available penicillin treatments), and the Willowbrook hepatitis study (in which students were purposely infected with hepatitis A).

When IRBs were introduced, only a handful of physicians conducted clinical research, virtually all of them on government funded projects in academic settings. IRBs met fairly irregularly, spent a reasonable amount of time considering their approvals, and oversaw research conducted by only one investigative center.

Today, almost none of the conditions just described accurately reflects the state of IRBs. Research is an enormous industry in which hundreds of thousands of doctors in hospitals, medical schools, and private clinics perform research on thousands of treatments, much of which is paid for by pharmaceutical firms. Many local IRBs still exist, but several large, centralized boards routinely review protocols, granting near simultaneous approval to “member” physicians. Some IRBs meet weekly (centralized ones often meet daily), and all are pressed for time because they are inundated with new protocols to review.

The pressures from pharmaceutical companies, clinical research organizations, and investigators for immediate IRB review and unqualified approval are immense. As a result, approvals are often made without extensive deliberation and without the benefits of appropriate scientific insight. In many cases, board members have little research training and suffer from inescapable conflicts of interest with the organizations that seek their approval, although these problems are unusual for centralized IRBs.All in all, the current IRB system is becoming less effective at providing the protections for which it was created.

The solutions
In 1998, the Office of Inspector General for the U.S. Department of Health and Human Services (HHS) issued a report suggesting ways to improve the IRB system (4). The department’s suggestions included:

• IRBs should be made more flexible by reducing procedural requirements, but should also be subjected to enhanced, performance-based evaluations.
• Centralized IRBs should receive more feedback about developments in multisite trials, and a “data safety monitoring board” should review those trials.
• IRBs should be informed of FDA actions against investigators and notified of prior protocol reviews conducted on the same protocol by other IRBs. They should increase their awareness of research conduct at the sites they review.
• Educational programs should be instituted for IRB members, along with continuing research ethics programs for participating physicians.
• Membership representation by nonscientific, non-institution-related personnel should be increased, and corporate or academic owners of IRBs should be prohibited from participating in their reviews.
• New regulations should be created requiring that IRBs have access to scientific, medical, and other resources necessary to make informed decisions.
• Federal oversight of IRBs should be revamped and expanded to include a board registration process.

These are not the only proposals that could improve the functions of IRBs, and they may not fully address the problems inherent in the existing system. But they do go a long way toward improving the review process and, hence, the protection of human subjects. No substantive government action has been taken on the HHS’s proposals, which is unfortunate given the number of patients who annually participate in research and the potential adverse consequences of a hastily reviewed protocol or inappropriately approved investigator.

References

1. PhRMA Annual Survey; Pharmaceutical Research and Manufacturers of America: Washington, DC, 1999; p 106.
2. Weiss, R. U.S. Halts Research on Humans at Duke; University Can’t Ensure Safety, Probers Find; The Washington Post, May 12, 1999.
3. Monmaney, T. VA Not Alone in Ethical Shortcomings; The Los Angeles Times, March 29, 1999.
4. U.S. Department of Health and Human Services, Office of Inspector General. Institutional Review Boards: Time for Reform; Report OEI-01-97-00193; U.S. Department of Health and Human Services: Washington, DC, 1998.