Testimony
Before the Committee on Government Reform Subcommittee
on Criminal Justice, Drug Policy and Human Resources
United States House of Representatives
Institutional Review Boards:
A System of Protections Still in Need of Reform
Statement of
Mark R. Yessian, Ph.D.
Regional Inspector General for
Evaluation and Inspections
December 9, 1999
Office of Inspector General
Department of Health and Human Services
Good morning. I am Dr. Mark R. Yessian, Regional Inspector General for Evaluation
and Inspections, in the Office of Inspector General's Boston office, U.S. Department
of Health and Human Services (HHS). I am pleased to testify at today's hearing
on the Federal role in protecting human subjects of medical research. My testimony
will focus on Institutional Review Boards (IRBs), the bodies that for more than
20 years have been charged under Federal law with ensuring human-subject protections.
With me is Laura C. McBride who was a key architect of these studies.
Mr. Chairman, the current system of protecting human subjects in medical research
is in need of major reform. This is a conclusion we emphasized in a series of
reports in June 1998. It is one we reiterate today.
BACKGROUND
In the late 1970s, Federal regulations were established requiring institutional review boards to review research protocols. The review is required to ensure human subjects are protected and are adequately informed of the risks and benefits of participating in research. Within the Department of Health and Human Services (HHS), both the Food and Drug Administration (FDA) and the National Institutes of Health (NIH) have responsibilities for human-subject protections and oversight of IRBs. NIH's Office for Protection from Research Risks, however, will soon be relocated to the Office of the Secretary for HHS where responsibility will transfer to the Assistant Secretary for Health. (For further background on IRBs and their functions, see the primer at the end.)
Background on Our Inquiry
In mid-1997, we initiated a broad, systemic review in response to concerns
raised in a prior Office of Inspector General study. In that study, we examined
clinical trials involving four investigational medical devices, and, in each
case, discovered limitations related to IRB review. These concerned serious
matters such as the implantation of a device in three times the number of human
subjects specified in the IRB-approved research protocol, the initiation of
a research effort without the changes called for in the informed consent document,
and the continuation of a research project for six weeks beyond when the IRB
had suspended it.
We were also aware of concerns about the IRB system raised by others--for instance,
the Advisory Committee on Human Radiation Experiments and the General Accounting
Office. Both groups raised questions in their respective reports about the adequacy
of the IRB review process.
In our most recent inquiry, we conducted a broad-based analysis of the IRB system
in order to gain an in-depth understanding of (1) the challenges facing IRBs
and (2) how the IRBs and the Federal government were meeting these challenges.
Toward that end, we developed a multi-faceted methodology drawing on many sources.
These included analyses of Federal records; an extensive literature review;
site visits to IRBs in 6 academic health centers; additional site visits accompanying
FDA inspectors; a survey on the electronic e-mail forum for those associated
with IRBs; and the systematic gathering of data from representatives of about
75 IRBs. This inquiry resulted in four reports issued in June 1998.
FINDINGS
Our overriding finding in these reports was that the system of protections,
which has been so carefully developed over the years, is in need of reform.
We based this conclusion on six main findings. We reiterate them below.
1. IRBs Face Major Changes in the Research Environment.
The research environment has undergone significant change in the 20 years since
the Federal IRB regulations were first established. In the late 1970s, research
was most often conducted by a single investigator at an academic medical center
under government funding. Today, almost the opposite is true. Many research
protocols are now multi-center trials involving thousands of subjects, numerous
investigators and institutions spread out across the country or even the world.
Each institution has little knowledge of what is occurring at other sites, if
problems have arisen, or even if other IRBs have called for changes in the protocol.
A greater proportion of research is funded by commercial sponsors. Research
is conducted increasingly at sites outside of academia where institutions can
more easily accommodate commercial sponsors desire for quick turnaround and
efficient research conduct. IRBs feel pressure to accommodate these sponsors
for whom time is money.
2. IRBs Review Too Much, Too Quickly, with Too Little Expertise.
IRBs, composed mainly of volunteer members, are working under increasing strains.
More and more research is being conducted and, as a result, IRBs across the
country are inundated with protocols to review. Some of the sites we visited
are now reviewing more than 2,000 protocols annually. These IRBs are also being
flooded with adverse-event reports from the multi-center trials they oversee.
In an effort to cope with the increased workload, many IRBs are forced to rely
on a pre-assigned reviewer to examine and summarize research plans. In some
IRBs, unless one of the assigned reviewers raises a question or concern about
the research, the board engages in little or no discussion at its meeting. Some
IRBs have been able to increase the length of their meetings, but many others
are forced to squeeze more reviews into a fixed block of time.
Science is becoming increasingly complex and many IRBs find that they lack sufficient
scientific expertise on their boards or staffs to adequately assess protocols.
This is particularly evident for protocols involving advanced biomedical techniques--such
as genetic testing--that raise scientific issues as well as moral and ethical
questions that may not be apparent to the untrained eye.
3. IRBs Conduct Minimal Continuing Review of Approved Research.
The IRBs' ongoing review of research can serve as an important safety net for
human subjects. The review involves a re-review of the protocol and an analysis
of the adverse event data to ensure the risk to subjects is minimal. However,
continuing review has become a low priority and is also often limited to a paper-based
review at many IRBs. Board members and officials we spoke with reported that
they seldom left the board room to visit the research site. Although many IRBs
would like to, few oversee the consent process.
4. Neither IRBs nor HHS Devote Much Emphasis to Evaluating IRB Effectiveness.
IRBs have little basis for knowing how well they are accomplishing their mission
of protecting human subjects. Seldom do IRBs seek out feedback from human subjects
or their families. Federal oversight does not compensate for these deficiencies
as it, too, is not geared to evaluating effectiveness. OPRR's oversight is limited
almost entirely to an up-front assurance process. The assurance is a document
stating an institution's commitment to adhere to Federal requirements. Most
IRB staff we spoke with told us the assurance has little impact on IRB functioning.
The OPRR generally goes on-site only in instances of alleged breakdowns in IRB
protections.
FDA oversight involves a more frequent on-site presence. However, their visits
focus almost entirely on IRB compliance with the procedural requirements set
forth in Federal regulations--such as attendance at review meetings, completeness
of minutes, and a review of the informed consent document. Such matters can
be important indicators of performance, but they give FDA little direct feedback
on the actual effectiveness of IRBs.
5. IRBs Face Conflicts that Threaten Their Independence.
Research monies, particularly from commercial sponsors, is an important source
of revenue and/or prestige for most institutions. For example, at one of the
academic medical sites we visited, about 25 percent of the operating budget
(nearly $200 million) derives from research activities. This revenue source
is increasingly important as medical institutions are squeezed by managed-care
cost pressures. An important counterbalance to these sorts of pressures is the
perspectives of certain IRB members whose concerns are primarily in nonscientific
areas or who are not otherwise affiliated with the institution. However, Federal
regulations require only one of each. We found few such "outside"
members on the boards.
6. IRBs and Their Institutions Provide Little Training for Investigators
and Board Members.
The review process can involve complicated ethical issues and scientific questions. Because of this, the education of board members is important. An understanding of the issues is also essential for research investigators who, themselves, initiate the informed consent process and interact directly with research subjects. However, IRBs face significant obstacles which include not only insufficient resources, but the reluctance of many investigators to participate in training sessions. For new IRB members, their orientation to the role is seldom much more than a stack of materials to read and on-the-job learning.
RECOMMENDATIONS
We found the stresses on the IRB system to be significant enough for us to make a number of strong recommendations to NIH/OPRR and FDA. The thrust of our recommendations was for a more streamlined approach to providing human-subject protections, both at the local and Federal levels. At the same time, we called for a greater emphasis on accountability, performance, and results. Our recommendations included a number of actions, many of which, in the near-term, could help to address the vulnerabilities in the system. Some of our key recommendations were as follows.
We called for eliminating or loosening a number of the procedural requirements that Federal regulations currently impose on IRBs, to allow IRBs to meet the significant challenges facing them. Greater flexibility could allow IRBs, for example, to concentrate their limited time and resources on protocols involving substantial risks to subjects. A quid pro quo for allowing IRBs greater flexibility is an increased emphasis on accountability. Therefore, we recommended regular, Federal performance-focused evaluations, whose results should be made public, and more extensive representation on IRBs of nonscientific and noninstitutional members.
We suggested reorienting the Federal oversight system to focus on IRB performance. In particular, NIH/OPRR should rework its assurance process so that it rests essentially on an institutional attestation to conform to the Federal IRB requirements in order to free scarce OPRR resources for on-site reviews and education sessions. FDA should search for ways of revamping its inspections, so that they focus less on narrow compliance matters and more on performance issues. FDA and OPRR should combine efforts to require a registration process whereby IRBs would submit basic descriptive information annually to HHS. This simple information would allow Federal bodies to more effectively target their oversight and facilitate dissemination of education materials.
The IRB system has long relied on an ethic of trust in the research community.
However, trust alone does not provide sufficient continuing protection. IRBs
need to be more aware of what is actually happening at the research site. For
example, they spend a significant amount of time reviewing and editing the language
of an informed consent document-- documents that can run up to 20 pages in length.
But, they know little about what actually happens in the interaction between
subject and investigator.
Another key mechanism of ensuring IRBs' meaningful reviews of approved research
is requiring certain outside sources to provide IRBs with valuable information.
Data Safety Monitoring Boards oversee many large-scale trials and continually
monitor the safety and efficacy of trials. Assessments of adverse events across
many disparate sites could help IRBs in their continuing reviews and reduce
their workload. Also, the FDA should provide IRBs with feedback on actions it
takes against investigators that are engaged in research at the IRBs' institution.
Investigators are the ones who actually interact with and consent subjects. Investigator education can help convey to investigators that along with the considerable independence that they have in the research process, there exists a significant responsibility to ensure human-subject protections. Investigators, who are adequately sensitized to human-subject protections, can also serve to minimize the need for regulatory intervention by the Federal government or by IRBs themselves. Board members must also be trained on Federal requirements. There are also many hidden ethical issues, particularly with the newer types of research such as genetics, to which IRBs must be sensitive.
AN UPDATE
Our reports were released a year and a half ago. Since then there have been
some encouraging developments. But, the system of protections is still in jeopardy
and the need for strong federal action is still compelling. I'll close by elaborating
on each of these points.
Encouraging Developments
Federal Initiatives. Progress has been made at the Federal
level, mainly in the activities of OPRR. In the year before our report was issued,
OPRR made only one on-site visit to a research institution. Since that time,
it has conducted a number of such visits, including one that resulted in the
temporary termination of all federally sponsored research at the Duke University
Medical Center. Some of their reviews represent the most probing and results-focused
inquires we have found of IRB performance, resulting in strong recommendations
to the IRBs. They are also taking a more active role in issuing further educational
outreach programs. NIH has issued guidance on the transmittal of information
to IRBs from data-safety boards used for trials it funds.
Local Efforts. These Federal efforts, in particular the recent
shutdown at Duke University, in tandem with our reports and recent media attention
have served as a catalyst for change at IRBs and their institutions. Many IRB
representatives told us that they have been able to garner more support and
resources from their institutions. For example, an IRB Chair from a small community
research institution called to ask for 20 copies of our report in order to hold
their first-ever IRB training program.
Professional Responses. There is a newly developed momentum
to certify IRB administrators and to develop a system for the private accreditation
of IRBs. Both of these efforts represent potentially important ways of raising
the bar in current IRB practice.
A System Still in Need of Reform
As encouraging as the above developments are, the system of protections provided
by IRBs remains in jeopardy. The same danger signs are widespread: expanded
workloads, quick reviews, threats to IRB independence, inadequate information,
and insufficient training.
Underlying pressures on IRBs continue to build and make it more difficult for
them to do their job:
Continued Need for Federal Action
A number of IRBs have enacted policies to help balance the pressures and minimize
conflicts of interest. However, most IRBs are hard-pressed to keep up with the
changing environment around them. Even if they are able to enact changes, they
may be at a competitive disadvantage to do so. Research sponsors often view
a burdensome and lengthy review process as a negative when deciding which sites
will conduct their research. Therefore, we need stronger Federal action. Federal
requirements can help safeguard adequate protections and establish a level-playing
field among research sites. Adequate protections would also help maintain the
integrity of the research enterprise. Just one unfortunate incident can damage
public confidence in research and, in turn, limit progress.
We continue to support all of the recommendations offered in our report. We
regard the following as particularly significant actions that Federal bodies
could take to shore up protections for human subjects.
1. Maintain adequate Federal on-site presence. Visit by Federal
bodies can help reinforce to institutions the importance of IRBs. These visits,
particularly when conducted randomly, can serve to identify cracks in the local
IRB system before problems occur and serve as technical assistance. When unfortunate
incidents or complaints do arise, quick Federal responses can provide an outside
perspective on necessary corrective actions.
2. Require investigator and IRB member education. The NIH should
require that institutions which receive funds for human-subject research under
the Public Health Service Act have a program to educate their investigators
about human-subject protections. Investigators, before participating in this
research, should be asked to sign a written attestation, similar to the one
required for FDA research, acknowledging their responsibility for human-subject
protections. Also, IRBs should be required to provide an orientation program
for new members and a continuing education program for all members before reviewing
research funded by NIH or under FDA purview.
3. Require more extensive representation of nonscientific and noninstitutional
members on IRBs. The current requirement of only one nonscientific
and one noninstitutional member on an IRB is inadequate. This regulation can
be met with the selection of just one person; boards often have 15-20 members.
"Independent" members can provide an important counterbalance to institutional
interests and potential conflicts. They also provide an important voice which
may be attuned to the needs and sensitivities of the community. There should
be enough of these members to ensure their voices are heard.
4. Require IRBs register with the Federal government. Current
estimates suggest the number of IRBs in the country is somewhere between 3,000
and 5,000. However, no one knows for certain how many IRBs there are, nor where
they are. IRBs should be required to register with the Federal government and
provide minimal descriptive information annually. Identifying information would
help target more effective oversight. Also, it would ensure that Federal bodies
be able to more efficiently disseminate important guidance and notices of upcoming
educational opportunities.
CONCLUSION
In closing, Mr. Chairman, I would like to underscore that the current system
of protections is supported by many conscientious researchers committed to protecting
human subjects and many competent, dedicated IRB staff. But the system they
function in has become brittle, with a few cracks. The effectiveness of the
system is in jeopardy.
It is vital to reform this system so that human subjects receive adequate protections,
so that we can confidently recruit the participants we need for clinical research,
and so that the research process itself remains on a productive, fast track--one
that will improve our understanding of disease mechanisms and introduce new
and effective products and treatments for addressing them. It is particularly
important for the Federal government to exert leadership in triggering reform
because in an increasingly commercialized, competitive research environment,
the Federal government is best suited to establishing the level playing field
on which contending parties participate.
Thank you for the opportunity to testify on this most important topic. At this
time, I would be happy to answer any questions which you or the other members
of the Subcommittee may have.
INSTITUTIONAL
REVIEW BOARDS: THE BASICS
What Do They Do? The responsibilities of IRBs fall into two main categories: initial
review and continuing review of research involving human subjects. Initial Review: IRBs review and approve a research plan before
the research is carried out. This review encompasses the research protocol,
the informed consent document to be signed by subjects, any advertisements
to be used in recruiting subjects, and other relevant documents. In
carrying out this review, the boards seek to ensure any risks subjects
may incur are warranted in relation to the anticipated benefits, that
informed consent documents clearly convey the risks and the true nature
of research, advertisements are not misleading, and the selection of
subjects is equitable and justified. IRBs focus much attention on the
informed consent document as it is the vehicle for providing information
to potential research subjects. Continuing Review: The continuing review process is multifaceted
and includes required reviews "at an interval appropriate to the
degree of risk but not less than once per year." In addition to
this continuing review, study amendments and reports of unexpected adverse
experiences by subjects are received periodically and reviewed to ensure
that the risk-benefit ratio of the research has not changed and remains
acceptable. Why Were They Established? As public awareness and concern about the treatment of human subjects
in research increased, the need for additional review mechanisms was
evident. These concerns grew from stories of the abuse of subjects during
the World War II trials at Nuremberg, the promotional distribution of
thalidomide resulting in numerous children born with birth defects,
the administration of cancer cells to chronically ill and senile patients
at a hospital in New York, and others. A 1966 article by Henry Beecher
brought prominent attention to human research abuses in medical schools
and hospitals citing 22 cases involving highly questionable ethics.
The formal requirements for the establishment of IRBs were outlined
in regulations stemming from the National Research Act of 1974 and in
FDA regulations issued in 1981. Where Are They Located? An estimated 3,000-5,000 IRBs can be found across the country. They
are most commonly associated with hospitals and academic centers. Boards
also exist in managed-care organizations, government agencies (such
as the National Institutes of Health, the Centers for Disease Control,
and State governments), or as for-profit entities that are independent
of the institutions in which the research takes place. How Are They Organized? Federal regulations require that boards have at least five members
with varying backgrounds. At least one member must have primarily scientific
interests, one must have primarily nonscientific interests, and one
must be otherwise unaffiliated with the institution in which the IRB
resides. A quorum, with at least one member whose interests are primarily
nonscientific present, is needed for voting. How Does the Department of Health and Human Services (HHS) Oversee
Them? Two agencies within HHS share responsibility for IRB oversight: the Office for Protection from Research Risks (OPRR) in NIH and the FDA. The OPRR’s main tool for oversight is the assurance document. Any institution that intends to conduct HHS-funded research must have an assurance on file with OPRR. The assurance is a written statement of an institution’s requirements and procedures for its IRB and human subject protections. Institutions conducting multiple HHS-supported studies can apply for a multiple project assurance (MPA) which can be renewed every five years. Institutions with smaller HHS-funded workloads, however, use a single project assurance (SPA) for each such project it conducts. The OPRR also conducts a small number of site-visits. The FDA’s main mechanism for IRB oversight is the inspection process. The FDA also inspects research sponsors and scientists (known as research investigators).
|