THE NATIONAL ACADEMY OF ENGINEERING

Chapter 3: Methods for Ensuring Protection: Education, Certification, and Accreditation (Research Involving Human Participants V1)

Introduction

Education and Competence in Conducting Research Involving Human Participants

Mechanisms to Ensure Institutional, IRB, and Investigator Compliance

Exhibit 3.1: OPRR Compliance Oversight Investigations Resulting in Restrictions/Actions to MPAs, January 1990-June 2000

Conflicts of Interest

Summary

Footnotes

References

Introduction

Protecting the rights and welfare of research participants is the major ethical obligation of all parties in the oversight system, who, to meet this obligation, must demonstrate competence in research ethics by being able to conduct, review, or oversee research involving human participants in an ethically sound manner. Such competence entails not only being knowledgeable about relevant research ethics issues and federal policies but also ensuring that conflicting interests for institutions, investigators, or Institutional Review Boards (IRBs) are disclosed and managed so that there is no question that research participants are being protected. Finally, the oversight system must include a robust monitoring process so that lapses by institutions, IRBs, and investigators can be remedied.

This chapter addresses measures needed to ensure that all parties involved in the research enterprise are competent with regard to protecting research participants and are reliably carrying out their responsibilities in this area. They include measures:

• to educate all parties involved in research with human participants about the substantive and procedural requirements of the system of protections (which includes demonstrated competence in research ethics);
• to provide ongoing mechanisms to ensure that institutions, IRBs, and investigators are in compliance with regulations, guidance, and procedures pertaining to the protection of research participants; and
• to manage conflicts of interest of investigators, IRBs, and institutions.

Back to Top

Education and Competence in Conducting Research Involving Human Participants

Education is an essential ingredient for developing competence in the ethical conduct of research with human participants; it can assist not only in the identification of ethical problems, but also in their resolution. Through well-designed ongoing educational programs, one can learn, for example, the most practical and effective steps for protecting confidentiality or improving the quality of the informed consent process and its documentation. Tailored education programs can help prevent routine ethical issues from becoming needless impediments to research and can provide basic skills to assist investigators and IRB members in dealing with new or complex ethical issues.

The need for education was a major focus of the National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research (National Commission), the President's Commission for the Study of Ethical Problems in Medicine and Biomedical and Behavioral Research (President's Commission), and the Advisory Committee on Human Radiation Experiments (ACHRE). It also has been central in National Bioethics Advisory Commission (NBAC) reports.

The National Commission focused on education of IRB members and proposed that the federal government and individual institutions play a role in that effort (National Commission 1978, 10, 15). The President's Commission recommended a broad educational program targeted to investigators, IRB members, and research administrators, including site visits to institutions by experienced IRB members and administrators (President's Commission 1983, 135 - 136). ACHRE highlighted the importance of education by linking the protection of the rights and interests of research participants to the ability of investigators to "appreciate sufficiently the moral aspects of human 'research and the value of institutional oversight' (ACHRE 1995, 817). In two previous reports, NBAC recommended that professional associations develop topic-specific educational materials (NBAC 1998; NBAC 1999).

Despite this enduring recognition of the important role of education, the educational function of the oversight system has been only minimally implemented through federal programs. The former Office for Protection from Research Risks (OPRR) devoted most of its limited resources to negotiating assurances and overseeing compliance. Aside from the educational activities associated with obtaining an assurance, regional workshops sponsored in conjunction with the Food and Drug Administration (FDA) were OPRR's only other sustained educational program. The lack of focus on education at the federal level was repeated at the local level, with institutions often failing to provide educational programs to their investigators, research staff, and IRB members. It should not be surprising, therefore, that the lack of appropriate educational programs and the resulting lack of knowledge about the ethical principles and procedures relevant to protecting research participants are among the deficiencies identified in audits and suspensions of research programs by OPRR, now the Office for Human Research Protections (OHRP) (OHRP 2000). A National Institutes of Health (NIH) survey of IRB chairs, members, administrators, institutional officials, and investigators at institutions holding Multiple Project Assurances (MPAs) found strong support for additional, improved education (Bell et al. 1998).

Some efforts are currently under way to increase educational opportunities. In 1997, two NIH-sponsored educational programs were initiated as part of the government's apology to the victims and families of the Tuskegee Syphilis Study. In one program, investigators receive grants to participate in ethics training programs tailored to meet their research needs; in another, program grants are awarded to individuals to develop courses in research ethics. 1 In addition, in 1998 OPRR hired its first director of education to develop computer-based education programs for IRB administrators and institutional officials. Outside government, organizations, such as Public Responsibility in Medicine and Research (PRIM&R) and the Association of American Medical Colleges, have traditionally provided education to their constituents. Recently, PRIM&R developed a new "IRB 101" course, which is offered before its annual meeting and at the request of institutions throughout the year. 2 In addition, some academic institutions have developed their own courses (Dunn and Chadwick 1999; Sugarman 2000).

Recently, NIH implemented a funding eligibility requirement that all individuals submitting grant applications provide evidence that they have been trained in research ethics. 3 This requirement has increased the number of education programs academic institutions offer to faculty and staff with the goal of raising the level of knowledge about and awareness of research ethics among investigators. Although NIH provides a model education program on its website, no requirements or guidance are provided regarding what constitutes appropriate training. Investigators and institutions have considerable discretion in fulfilling the NIH requirement. This activity has generated a beneficial exchange of ideas on how best to teach research ethics to different audiences.

Educational programs should focus on ethical principles, their relationship to ethical standards, and the procedures necessary to implement them. It is critical to emphasize the links among these three components. Because each research study is different and some raise new ethical issues, investigators and IRBs must be able to refer to and apply general ethical principles. Yet, it is equally true that research ethics cannot be taught or learned entirely in the abstract. As one expert cautioned: "rote memorization of guidelines, principles, and regulations, absent the ability to apply such concepts in practical situations, will not ensure the protection of [research participants]." 4 Educational programs should include problem-solving components in which investigators and IRB members learn ways in which to identify, discuss, and resolve ethical issues pertinent to specific types of research studies.

For the behavioral and social sciences, IRB members and investigators should be educated in the ethical standards and associated procedures that are appropriate to human participant research in these disciplines. In addition, IRB members should be familiar with standards in professional societies' codes of ethics, particularly for fieldwork in anthropology or survey research in sociology or psychology. Such standards address issues of privacy and confidentiality, informed consent, and community protection. Complex questions concerning deception or stigmatization of individuals or communities may need particular attention, particularly in the context of research involving individuals with vulnerabilities.

NBAC supports the independence of institutions and other entities that would enable them to design their own education programs. However, it is important that all educational programs provide appropriate coverage of the ethical principles and standards and current federal regulations and guidance related to human participant research.

Although the primary focus of education is on those parties involved in the daily conduct of research (e.g., investigators and IRBs), other groups that carry important ethical responsibilities are also important targets for education. For example, sponsors, whether public or private, often are directly involved in providing advice and supervision to grantees and contractors. Federal agencies that have their own oversight offices (see Chapter 2) also should be competent in research ethics, regulations, and guidance.

If investigators are to conduct research ethically and research participants are to be able to make informed decisions about participation, education about the ethical conduct of research needs to be an integral part of science education in the United States (Pritchard 1999). Even for those students who do not pursue careers in science or research, education in research ethics could make them better informed in terms of future research participation or in terms of understanding the nature of the research enterprise. The American Association for the Advancement of Science (AAAS 1993) and the National Academy of Sciences (National Research Council 1995) have produced documents outlining content standards for science education, including standards for the ethical treatment of research participants.

While some curricula include courses or lectures on research ethics, such education generally is missing from most undergraduate and graduate school science curricula and is seldom required, including for medical and other health professions education (Coughlin et al. 1999; Mastroianni and Kahn 1998). This phenomenon is particularly disturbing because many students conduct research involving human participants as part of their graduate educational experience (Gunsalus 1997). Moreover, a large number of scientists and many health professionals expect to conduct human research. For example, in a survey of all graduating medical students conducted in 1996, 78.3 percent responded that they expect to be involved in research at least somewhat during their medical career, and 11.9 percent responded that they expected to be involved exclusively or significantly involved during their medical career. 5 Given the role of physician-investigators within academic medical centers and the growing number of community physicians conducting research or referring patients to research, education in research ethics and human research protection should be a meaningful component of medical education. Professional societies should assist institutions in developing educational programs and curricula.

The need for education in research ethics in graduate curricula extends beyond the health fields. Graduate programs in social sciences and the humanities also should incorporate curricula relevant to the specific discipline, which could take advantage of the fact that professional societies have developed codes of ethics that deal specifically with conducting research involving human beings. Education also can be extended to other groups likely to be involved in research. Patient advocacy organizations can, and often do, provide education to their members. Informing groups about rights of research participants, protections afforded, and points to consider in deciding whether to enroll in research can empower a large number of prospective research participants.

Recommendation 3.1:

All institutions and sponsors engaged in research involving human participants should provide educational programs in research ethics to appropriate institutional officials, investigators, Institutional Review Board members, and Institutional Review Board staff. Among other issues, these programs should emphasize the obligations of institutions, sponsors, Institutional Review Boards, and investigators to protect the rights and welfare of participants. Colleges and universities should include research ethics in curricula related to research methods, and professional societies should include research ethics in their continuing education programs.

Recommendation 3.2:

The federal government, in partnership with academic and professional societies, should enhance research ethics education related to protecting human research participants and stimulate the development of innovative educational programs. Professional societies should be consulted so that educational programs are designed to meet the needs of all who conduct and review research.

The federal government should use a variety of incentives to encourage the enhancement of research ethics education.

Demonstrating Competence

Certification of Individuals

Educating all parties in research ethics and human research protections is effective only when it results in competence to design and conduct ethically sound research, including analyzing, interpreting, and disseminating results in an ethically sound manner. Such competence, however, cannot be assumed to follow from exposure to an educational course or program. As the complexion of research continues to change, new and challenging ethical dilemmas emerge with advances in technology, and as more people become involved in research as investigators or in roles particularly related to oversight, it becomes increasingly important for all parties to demonstrate their competence in the ethics of research involving human participants.

Currently, there is no regulatory requirement to demonstrate competence, even when the level of risk is high or the participants are vulnerable. In recent years, there has been much discussion about the need to implement regulations consistently, set standards for IRBs and measure performance in meeting those standards (OIG 1998a), and measure the competency of investigators (Cohen 2000).

Many IRBs and investigators have suggested that investigators should be required to demonstrate a certain level of competence in order to conduct human research. 6 NBAC supports this view. All individuals directly involved in the protection of human research participants should be certified, including investigators and research staff, IRB members and staff, and sponsors. Individuals employed by sponsors also should be certified when they are directly involved in the conduct or oversight of research, for example, grants or contracting officers who might make decisions relevant to human participant protections. Further details regarding just which individuals require certification will, of course, be context dependent.

Because roles vary substantially and among investigations the diversity in research design and subject matter is great, several certification options should be available. For example, certification of investigators may need to be specialized for particular disciplines or in relation to certain subject matter in order to increase its effectiveness. Different mechanisms for certification should be tested and evaluated. Currently, under the NIH requirement, certification of investigators is left entirely up to the institution. Certification of IRB staff and members is not required; however, there is a voluntary certification program for IRB administrators. The Council for Certification of IRB Professionals (CCIP) was formed to promote IRB administration practice and to advance the quality of research participant protection programs through the certification of qualified IRB professionals. CCIP administered the first certification examination in October 2000 and announced recently that 107 professionals were certified based on the results of these exams. 7 Certification efforts can surely be improved. Overall, the goal of such programs should be to ensure competence and consistency that is relevant to the performance of the particular individual parties.

Recommendation 3.3:

All investigators, Institutional Review Board members, and Institutional Review Board staff should be certified prior to conducting or reviewing research involving human participants. Certification requirements should be appropriate to their roles and to the area of research. The federal government should encourage organizations, sponsors, and institutions to develop certification programs and mechanisms to evaluate their effectiveness. Federal policy should set standards for determining whether institutions and sponsors have an effective process of certification in place.

Accreditation of Institutions and Independent IRBs

Related to measuring competence in individual investigators and IRB members and staff is measuring a core set of competencies in independent IRBs or institutions. In this arena, accreditation programs might be the most effective mechanisms for measuring compliance of such entities to a set of standards (Hamm 1997). In this report, the term accreditation refers to programs directed toward institutions, and the term certification refers to programs directed toward individuals.

Although accreditation and certification do not guarantee desired outcomes, they can be helpful in improving performance. These programs generally involve experts and peers developing a set of standards that represents a consensus of the best practices in the profession. Therefore, the choice of standards and the criteria for evaluating whether an institution has met them are critically important. The emphasis of these programs should be on education, on assuring that appropriate protections are in place, and on avoiding excessively bureaucratic procedures. Accrediting programs may lose their value if they are seen as merely another administrative burden.

Despite concerns that accreditation might add another layer of bureaucracy to an already regulated environment, several advantages of accreditation programs have been noted. They are generally voluntary and represent a profession's desire to self-regulate. Many accrediting programs strive to meet higher standards than are required by law, so that having the credential implies a higher level of competence than what may be minimally required. Properly governed and organized, accrediting bodies can improve competence (and thus performance) within a profession, which helps to achieve the goals of the profession and build public trust (Hamm 1997). In addition, accreditation serves an educative role.

Institutions seeking accreditation, for example, usually go through a periodic review process involving a self-study and a site visit from a team of experts in the profession. The self-evaluation of the institution in preparation for a site visit helps it identify strengths and weaknesses in its programs and make improvements. The site visit can be an especially useful learning experience that encourages the institution to sustain best practices.

Accrediting programs are widely used in the fields of health and education. These programs are seen as having a major and generally positive influence (Hamm 1997). One highly regarded accrediting program in research is the Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC). This private, nonprofit organization "promotes the humane treatment of animals in science through voluntary accreditation and evaluation programs." AAALAC accreditation is regarded as a validation of having a high-quality operation, and research programs are willing to spend resources to obtain such accreditation. AAALAC accreditation demonstrates that an institution is going beyond the minimum required by law, achieving excellence in animal care and use (Bayne 1998).

Recently, the Association for the Accreditation of Human Research Protection Programs (AAHRPP) was incorporated to 'provide a process of voluntary peer review and education among organizations concerned with research involving humans.' The overall goal of this accrediting organization is to improve protection of human research participants by developing "best practice" performance standards and by recognizing institutions that meet those standards. AAHRPP expects to begin accrediting institutions in the near future. 8 The Department of Veterans Affairs (VA) requires all institutions involved in conducting research sponsored by the VA to be accredited (VA 2000), but it is the only federal agency that requires institutions or IRBs to become accredited. OHRP is also moving in the direction of accreditation. Commissioned by OHRP to conduct a study on accreditation and other human participant research issues, the Institute of Medicine issued a report on the use of accreditation programs and offered recommendation regarding standards (IOM 2001).

Moving toward accreditation and certification is a positive step in ensuring improved oversight of human research. As the proposed oversight system evolves to include accreditation and certification programs, flexibility in testing different models and methods of accreditation and certification should be allowed to ensure that when these programs become fully functioning they are sensitive to the needs of the system and responsive to the various entities seeking accreditation certification. For example, institutions that do not have IRBs but that are conducting research involving human participants should be eligible for accreditation because they have education, monitoring, and accountability responsibilities; and independent IRBs that might not perform all the functions of an institution conducting research also should be eligible for accreditation to carry out their more limited responsibilities.

Given the diversity among institutions that might seek accreditation and the fact that there are likely to be several accrediting organizations, the proposed federal oversight office should set criteria for government approval of accrediting bodies. In addition, the federal oversight office should develop a mechanism for monitoring accrediting bodies, and accrediting bodies, in turn, should be monitoring the institutions they accredit for continued competency. In the end, the office should be able to evaluate the effectiveness of accreditation.

Although the establishment of accreditation programs is likely to improve protections to human participants, such programs will be costly. In addition to the cost of accreditation itself, all institutions seeking accreditation will need to devote additional resources and staff to support new functions or strengthen existing functions. For example, some institutions might need to add educational programs or add monitoring functions. However, the potential improvements in the system justify these additional costs.

Recommendation 3.4:

Sponsors, institutions, and independent Institutional Review Boards should be accredited in order to conduct or review research involving human participants. Accreditation should be premised upon demonstrated competency in core areas through accreditation programs that are approved by the federal government

Back to Top

Mechanisms to Ensure Institutional, IRB, and Investigator Compliance

Assessing institutional, IRB, and investigator compliance can help to ensure that standards are being followed consistently. Current mechanisms for assessment include assurances of compliance issued by the Department of Health and Human Services (DHHS) and several other federal departments, site inspections of IRBs conducted by FDA, other types of site inspections, and institutional audits. In addition, some institutions have established ongoing mechanisms for assessing investigator compliance with regulations. 9 However, institutions vary considerably in their efforts and abilities to monitor investigator compliance, from those having no monitoring programs to those conducting random audits. 10

Institutional/IRB Compliance

Assurances of Compliance

Assurances of compliance (referred to as assurances) were developed in the late 1960s when the NIH Division of Research Grants was assigned responsibility for implementing the federal requirements for protecting human research participants. (See Appendix C.) Assurances were negotiated with each institutional grantee, with the negotiations allowing each institution to create its own policies and procedures for protection as long as they were fully consistent with federal regulations. The negotiation process also allowed federal officials to educate institutions about requirements and procedures for participant protection. Because the assurance indicated what an institution intended to do to protect research participants, it was essentially a pledge or commitment on behalf of the institution to comply with all appropriate regulations and guidance.

The National Commission believed that there should be uniform implementation of the federal regulations and recommended that each institution engaged in regulated research provide assurance to a single office that all research would be conducted in accordance with federal regulations (National Commission 1978). This perspective was reinforced by the President's Commission, which suggested further steps to ensure that federal monitoring would be coordinated to minimize the bureaucratic burden imposed on institutions.

The President's Commission argued that both assurances and on-site evaluations were needed to monitor implementation of the federal regulations and that a single office should coordinate such procedures for all regulated research (President's Commission 1983, 133). A single office was never created. Instead, each federal department and agency may issue its own assurance, although many rely on DHHS assurances. (See Table 3.1.) The assurance process provided a means of identifying certain problems and correcting them before research was conducted. However, it provided only a limited indication of how an IRB would actually work. The value of the recommended site visits was the opportunity to learn how IRBs were actually functioning and for institutions to receive feedback on their IRBs' performance. However, to rely only on after-the-fact, "for-cause" site visits to determine whether an IRB met basic regulatory requirements is not the most effective strategy for ensuring that research protocols are being reviewed by a properly constituted and procedurally sound IRB (President's Commission 1983).

Table 3.1: Federal Agency Practices Regarding Assurances of Compliance

Agency	Issues Assurances of Compliance	Relies on Other Agency Assuranecs
Central Intelligance Agency	No	No
Department of Commerce
National Telecommunications and Information Administration	No	No
National Institutes of Standards and Technology	No	Department of Health and Human Services (DHHS)
Department of Defense	Yes	DHHS
Department of Education	Yes	DHHS
Department of Energy	Yes	DHHS
Department of Health and Human Services	Yes	No
Department of Energy	Yes	DHHS
Department of Health and Human Services	Yes	No
Department of Housing and Urban Development	No	Not reported
Department of Justice	Some components	DHHS
Department of Transportation	No	DHHS
Department of Veterans Affairs	Yes	DHHS
National Aeronautics and Space Administration	Yes	DHHS
National Science Foundation	Yes	Yes
Social Security Administration	No	Yes
Agency for International Development	Yes	DHHS, Department of Defense
Consumer Product Safety Commission	Yes	DHHS
Environmental Protection Agency	Yes	DHHS

Although DHHS and some other departments that adopted the Common Rule continue to issue assurances, site visits as envisioned by the President's Commission were never implemented. In addition, for the most part, the negotiation process for assurances developed into a routinized and somewhat tedious procedure. Standardized documents that mirror the federal regulations substitute for the independently negotiated assurances specific to the institution's culture, policy, and procedures. The educational function of assurances was lost. 11 Currently, few seem to view assurances as meaningful commitments or pledges by institutions to sustain a broad-based program of review and monitoring. 12 Instead, they are viewed as unnecessary, additional bureaucratic paperwork with limited, if any, value in protecting research participants. 13 Additional problems reported to NBAC were long delays in processing assurances (between 6 and 12 months in some cases), when the expectation is that OHRP should be able to process assurances in a much shorter period, confusion about which institutions required assurances, and OHRP's review of consent forms submitted by institutions seeking an assurance when the same consent form was reviewed and approved by another collaborating institution with an MPA. Problems with the assurance process are not unique to domestic research. In the context of international research, the assurance process was most often criticized because it requires foreign institutions to follow rigidly U.S. requirements and procedures. The need for greater flexibility by the United States in conducting international research would allow for procedural variations that are responsive to local culture and practice while still ensuring compliance with substantive ethical principles (NBAC 2001).

OHRP revised the assurance process effective December 2000 and is currently testing the new procedures. 14 The revised process entails the use of one Federalwide Assurance (FWA) document for domestic institutions and another assurance document for foreign institutions. Each legally separate institution must obtain its own FWA, and assurances approved under this process will cover all the institution's federally supported research involving human participants. The domestic assurance replaces MPAs, Single Project Assurances, and Cooperative Project Assurances. Individual federal agencies can still choose whether to use the new assurance or issue their own. The assurance document is similar to previous assurance documents in content. However, institutions must meet additional requirements, such as that certain institutional staff complete OHRP's computerized educational training. The new assurance process is intended to reduce burden on institutions by allowing all institutions to qualify for the one FWA that may be renewed every three years. The revised process also includes registration of IRBs, regardless of whether they review research sponsored or regulated by a federal agency that follows the Common Rule. 15 Modification to the revised system will be made based upon feedback and experience during the first few months of implementation.

FDA Site Inspections of IRBs

FDA conducts site inspections under its own regulations. In 1976, FDA developed its Bioresearch Monitoring Program in response to Congress' instruction to expand its monitoring activity. The program encompasses monitoring of investigators, sponsors, program monitors (individuals selected by the sponsor to oversee the clinical investigation), and IRBs. The purpose of the Bioresearch Monitoring Program is to ensure the quality and integrity of data submitted to FDA to 'demonstrate the safety and efficacy of regulated products, and to determine that human rights and the welfare of human and animal research subjects are adequately protected' (FDA 1998a). FDA conducts inspections of sponsors to determine how they ensure the validity of data submitted to them by investigators and to determine compliance of sponsors, contract research organizations, and monitors with applicable regulations (FDA 1998b). The purpose of FDA inspections of investigators is to help ensure compliance of investigators with the regulations (including the regulations to protect research participants) and to assess through audit procedures whether records substantiate data submitted to FDA (FDA 1998a). These inspections also include assessment of whether the investigator obtained appropriate IRB approval of research studies and whether participants have signed consent forms. However, FDA's own findings for fiscal year 1998 show that for studies involving medical devices, over 50 percent of sponsors may not be monitoring trials adequately (OIG 2000). Most FDA inspections of investigators are conducted after the trial is complete. Thus, any detected violations of regulations to protect research participants are found after the point when participants in the particular trial could have received adequate protections. However, the inspections are helpful in improving compliance of investigators and, therefore, protection of participants in future research. More recently, FDA's inspection process has focused more on real-time complaints. 16 FDA conducts surveillance (routine) and directed (when information "calls into question" regulated practices) inspections of IRBs. Usually IRB inspections are scheduled every five years, although if there are major problems, inspections can occur more frequently (FDA 1994). During an inspection, an FDA field investigator (inspector) chooses a few studies that received initial IRB review within the past three years and follows them through the IRB review process. Inspectors look at IRB policies and procedures; minutes; membership; and records of studies, including protocol, consent form, investigator's brochure, and correspondence between the IRB and investigator. IRBs that are found to be out of compliance may be subjected to sanctions ranging from a warning letter to rejection of the data from the trial to prosecution (FDA 1994). FDA conducts approximately 300 IRB inspections annually as part of its Bioresearch Monitoring 17 FDA does not require its sponsors to provide assurances of compliance; however the agency does require investigators to provide a written commitment that, before initiating an investigation subject to an institutional review requirement under 21 CFR 56, an IRB will review and approve the investigation in accordance with 21 CFR 56 (21 CFR 312.53(c)(1)(vi)(d); 312.53(c)(1)(vii); 21 CFR 812.43(c)(4)(i)). The sponsor makes similar commitments (21 CFR 312.23(a)(1)(iv); 21 CFR 812.20(b)(6)).

Other Site Inspections

Some other federal departments also conduct site visits. For example, the Department of Energy (DOE) routinely conducts educational site visits at its contract facilities. A team of experienced IRB professionals reviews policies and procedures at individual facilities and makes recommendations for improving protections tailored to the operation at the local facility. 18 The DOE human subjects program manager makes determinations of noncompliance.

The Department of Education authorizes its extramural research director to investigate allegations of noncompliance. At the National Science Foundation, allegations of noncompliance are referred to the Office of Inspector General for investigation. The Department of Veterans Affairs uses several new mechanisms to monitor compliance. For example, the Multi-Assessment Program is a systematic, prospective approach to assist VA medical center research programs and investigators fulfill their responsibilities to conduct research so that there is adequate protection of human participants. The VA Office of Research Compliance and Assurance also uses Special Inquiry Force Teams to investigate allegations of research noncompliance, and if warranted, may initiate a Comprehensive Research Integrity Program review.

OHRP/OPRR also conducts both educational and 'for-cause' site visits. In 1999, OPRR reported conducting 'between zero and one not-for-cause site visits a year' to monitor IRB and institutional compliance with the federal regulations (Foubister 1999). In 1998 - 1999, OPRR's number of 'for-cause' site visits increased, along with suspensions of assurances for several prominent research institutions (Exhibit 3.1).

Institutional Audits

Institutions and IRBs should regularly review policies and procedures and monitor their implementation. Moreover, they should take steps to open their activities to greater public review and accountability. In a previous report, NBAC suggested that IRBs should make descriptions of their policies and procedures public and provide annual summary statistics regarding the nature and scope of the research approved. NBAC also suggested that institutions adopt internal audit procedures to assure themselves that their IRBs are following regulations (NBAC 1998). Institutions can also assume greater responsibility in ensuring that investigators comply with regulations and carry out protocols as approved by the IRB, as discussed below.

Assurances of compliance, site inspections, and institutional audits are all potentially useful mechanisms for improving institutional and IRB compliance. No one mechanism used in isolation of the others will be sufficient to monitor compliance. As certification and accreditation programs become more widespread, the utility of assurances and site inspections should be examined in order to avoid duplication of effort.

Back to Top

Exhibit 3.1: OPRR Compliance Oversight Investigations Resulting in Restrictions/Actions to MPAs, January 1990 - June 2000

Year Institution and Action by OPRR

1990

1991

Massachusetts Eye and Ear Infirmary.

Certain research was excluded under the existing MPA pending negotiation of a renewed MPA. New participant accruals were suspended pending re-review by appropriately convened IRB and notification of former participants about inadequate informed consent.

NIH.

Certain collaborative research was excluded under the existing MPA. Designation of a signatory official with authority to ensure institutional human participant protections, modification of institutional policies and procedures to ensure IRB review of all human participant research, and development of an education program were required.

University of California - San Francisco.

New participant accruals in all pediatric oncology research were suspended pending modification of informed consent documents and revised consent for all current pediatric oncology participants, and mechanisms to ensure regulatory compliance in such research were required.

State University of New York College of Optometry.

Use of expedited IRB review procedures was suspended. Suspension of human participant involvement in certain research, re-review of research by an appropriately convened IRB, modification of IRB review and record keeping procedures, and designation of a signatory official with authority to ensure institutional human participant protections were required.

Florida State University.

Use of expedited IRB review procedures was suspended. Re-review of research by an appropriately convened IRB, modification of IRB review, and record keeping procedures were required.

1992

1993

NIH.

Progress reports on implementation of education program were required.

1994

University of California - Los Angeles.

Schizophrenia research was excluded under the existing MPA. Modification of informed consent documents, information materials, and record keeping for schizophrenia research, the addition of participant representatives to the IRB, and the establishment of a Data and Safety Monitoring Board for psychiatric disorders research were required.

West Virginia University.

Special monitoring of urology department research and modification of IRB review and record keeping procedures were required.

Memorial-Sloan Kettering Cancer Center.

Special monitoring of melanoma immunization research, modification of IRB review and record keeping procedures, verification of completeness of informed consent documents, development of handbook of guidelines for investigators, and designation of a signatory official with authority to ensure institutional human participant protections were required.

Medical University of South Carolina.

Review of research by appropriately convened IRB, modification of institutional mechanisms to ensure IRB review of human participant research, and development of an education program were required.

University of California - San Diego.

Human participant involvement in certain research was suspended, and re-review of research by an appropriately convened IRB, modification of IRB review and record keeping procedures, and designation of a signatory official with authority to ensure institutional human participant protections were required.

University of Florida.

Use of expedited IRB review procedures was suspended. Re-review of research by an appropriately convened IRB, modification of IRB review and record keeping procedures, and revision of IRB guidelines for investigators were required.

University of Minnesota.

Research conducted by the Department of Surgery was excluded under the existing MPA, and human participant involvement in certain research was suspended. Re-review of research by an appropriately convened IRB, modification of IRB review, and record keeping procedures were required.

1995

University of Virginia.

Behavioral and educational research were excluded under the existing MPA, and human participant involvement in behavioral and educational research and certain medical sciences research was suspended. Re-review of research by appropriately convened IRBs, modification of IRB review and record keeping procedures, provision of an enhanced educational program and materials, review of staff support and resources for IRBs, designation of a signatory official with authority to ensure institutional human participant protections, and comprehensive review of IRB policies and procedures were required.

Centers for Disease Control and Prevention.

Provision of updated information to measles research participants, development of conflict of interest guidelines for IRB members, review of staff support and resources for IRBs, development of an educational program, a proposal for a mechanism to ensure performance site assurances for international research, and a comprehensive review of IRB policies and procedures were required.

University of California - Los Angeles.

Development of an educational program, revised IRB guidelines for investigators and the procedures manual, review of staff support and resources for IRBs, provision of a locked filing system and computerized tracking system to ensure confidentiality of IRB records, and development of mechanisms to ensure that all human participant research receives IRB review were required.

Center for Molecular Medicine and Immunology.

Modification of IRB review and record keeping procedures were required.

Veterans Affairs Medical Center, West Los Angeles.

Modification of IRB review and record keeping procedures were required.

Cook County Hospital Hektoen Institute for Medical Research.

Participant enrollment in research not receiving valid initial or continuing IRB review and use of expedited review procedures were suspended. Revision of IRB review procedures was required.

Cornell University Medical Center.

Enrollment in research not receiving adequate initial or continuing IRB review was suspended. Modification of IRB review and record keeping procedures, a plan for increased staffing, a plan for education of IRB members and investigators, and a mechanism for investigation of non-compliance were required.

University of Rochester.

Enrollment in research not receiving appropriate initial or continuing IRB was suspended. Review and development of written operating procedures and an investigator handbook, a finalized organization structure of institutional human participant protections, and development of an education program for IRB members and investigators were required.

Wayne State University.

Enrollment in research not receiving appropriate initial or continuing IRB review was suspended. Increased staffing and resources for the IRB, development of a mechanism for prompt review of adverse event reports, development of an educational program for IRB members and investigators, and review and revision of IRB policies and procedures were required.

City University of New York.

Development of an educational program for IRB members and investigators, review and revision of IRB policies and procedures, revision of sample informed consent documents to comply with DHHS regulations, and revision of IRB record keeping procedures were required.

University of Maryland - Baltimore.

Enrollment of participants in research not receiving appropriate initial or continuing IRB review was suspended. Review and revision of informed consent documents for psychiatric research to ensure compliance with DHHS regulations, development of an educational program for IRB members and investigators, and revision of IRB review procedures, including procedures for research involving vulnerable participant populations, were required.

University of California - Irvine.

Modification of initial review process to ensure compliance with regulations, revision of continuing review and oversight procedures, increased documentation of IRB actions in accordance with regulations, enhanced education for investigators, IRB members, and staff, and increased support for social and behavioral sciences IRB were required.

Western Carolina Center.

Due to limited DHHS support to the Western Carolina Center, Single Project Assurances (SPAs) were required for DHHS-supported human participant research, and the MPA was deactivated on October 30, 1998.

1996

1997

1998

Rush Presbyterian St. Lukes Medical Center.

OPRR suspended the MPA for five days pending development of adequate corrective action plans. The MPA was subsequently reinstated with restrictions. OPRR required correction of 17 identified deficiencies in systemic human participant protections, restructuring of systemic human participant protections including changes in leadership, enhanced institutional commitment, and expanded IRB membership, development of an education program for IRB members, IRB staff, and research investigators, suspension of new enrollments in all federally supported research pending review of all such protocols by the reconstituted, convened IRB, and suspension of IRB review of federally supported research by expedited review procedures.

Scripps Clinic and Research Foundation:The Scripps Research Institute.

OPRR required correction of 20 identified deficiencies in systemic human participant protections, re-review of all DHHS-supported human participant protocols to include review of complete grant application, development of an educational program for IRB members, IRB staff, and research investigators, and quarterly progress reports.

Duke University.

OPRR restricted the MPA and required several corrective actions.

Duke University Medical Center.

OPRR restricted the Duke University Medical Center MPA and required several corrective actions.

1999

Friends Research Institute.

OPRR removed from coverage under the existing MPA any performance sites outside the Maryland area, removed recognition of one IRB under the MPA, and withdrew approval of all inter-institutional and cooperative amendments to the MPA.

Mt. Sinai School Of Medicine.

OPRR restricted the MPA and required satisfactory implementation of a series of corrective action plans.

Veterans Affairs, Greater Los Angeles Health Care System (Formerly Veterans Affairs Medical Center,West Los Angeles).

OPRR deactivated the MPA on March 22, 1999. Enrollment of new participants in all federally supported research was suspended, and involvement of previously enrolled participants was allowed only when it was in the best interest of the participant.

Fordham University.

Due to the limited number of DHHS-supported projects at Fordham University and persistent concerns about compliance with human participant protections requirements, SPAs were required for all DHHS-supported human participant research, and the MPA was deactivated on May 31, 1999. Duke University Medical Center. OPRR suspended the MPA for five days. OPRR reinstated the Duke University Medical Center MPA with restrictions and required several corrective actions, including re-review of DHHS-supported research by the IRB, implementation of a second IRB, and implementation of appropriate education programs for IRB members, IRB staff, and all investigators.

University of Illinois at Chicago.

OPRR restricted the MPA and required corrective actions in response to 29 identified deficiencies, including revision of the MPA to indicate enhanced institutional commitment to the protection of human participants and development of an educational program for IRB members, IRB staff, and research investigators. New enrollments in all federally supported research pending re-review of all such protocols were suspended.

St. Jude Children's Research Hospital.

OPRR restricted the MPA and required submission of a progress report describing implementation of all corrective actions three months after the site visit.

Virginia Commonwealth University.

OPRR restricted the MPA and required corrective actions, including development of an educational program for IRB members, IRB staff, and research investigators and submission of a list of all active IRB-approved protocols. New participant enrollments in all federally supported research pending re-review of all such protocols were suspended.

Virginia Commonwealth University.

OPRR suspended the MPA and required revised corrective action plans, including designation of a new signatory official and new IRB chairpersons.

University of Alabama at Birmingham.

OPRR restricted the MPA and required suspension of new participant enrollments in all federally supported research that had not had appropriate initial or continuing review. Submission of corrective action plans to address all deficiencies and concerns related to systemic human participant protections and submission of an educational plan for all IRB members, all IRB staff, and all research investigators were required.

2000

Adapted from Compliance Oversight Investigations Resulting in Restrictions/Actions to Multiple Project Assurances, 1/90 - 11/99 (OPRR 2000).

Monitoring Investigator Compliance with IRB-Approved Protocols: Verification

Institutional monitoring to verify that investigators are following IRB-approved protocols is an important means of determining whether investigators are actively meeting their ethical obligations to protect research participants while a study is in progress. Such verification should include activities such as observing research, especially the enrollment of participants, auditing research records, and establishing clear procedures for receiving complaints or concerns from participants or staff.

Some argue that as collegial review bodies, IRBs should not question the information provided by investigators.19 On the other hand, others have argued that IRBs should develop mechanisms for continuing review of research that include verification of the investigator's conduct of the study (Weijer et al. 1995). There are cases in which investigators' deviations from IRB-approved protocols have affected participant protections (OIG 1995). In one instance, a case of financial fraud by two clinical investigators may also have affected the welfare of research participants (Teegardin and Whitt 1997). Failure to follow IRB-approved protocols may not be limited to serious abuses. One recent survey of clinical research professionals found that as many as 40 percent had encountered cases of questionable consent, and many had encountered cases of inappropriate alteration of study eligibility criteria, alteration of care provided to participants, and breach of confidentiality (Shefrin and Harper 2000)

IRB review alone does not address these types of problems. Although IRBs are not and should not be the research police, protecting participants will sometimes require IRBs to verify at least on a sample basis the conduct of approved research studies. IRBs are not fully meeting their responsibilities when they make detailed changes to the language in a consent form, for example, but have no way to verify how the consent form is discussed with participants.

Under current regulations, IRBs have the authority "to observe or have a third party observe the consent process and the research" (45 CFR 46.109(e); 21 CFR 56.109(f) [FDA regulations differ in reference to regulation]), and institutions are required to have procedures for reporting on "serious or continuing noncompliance with this policy or the requirements or determinations of the IRB" (45 CFR 46.103(b)(5); 21 CFR 56.108(b)(2) [FDA regulations differ in reference to regulation]). Few IRBs carry out this type of monitoring, because they do not have sufficient resources.20 The desire for such verification is not new. In its report on IRBs, the National Commission intended for IRBs to observe investigators and hear directly from research participants: "At the discretion of the IRB, the consent process or the research itself may be observed on a sample or routine basis, subjects may be interviewed about their experience in research, and research records (including consent forms) may be reviewed. Also at the discretion of the IRB, investigators may be required to provide subjects with a form on which they can report to the IRB their experiences in research" (National Commission 1978, 17). Moreover, noting the difficulty of such observation, the National Commission made it clear that these processes were not supererogatory on the part of IRBs; for at least some research studies, observing the conduct of research to ensure the protection of the research participants was an 'obligation' (National Commission 1978, 18).

More recently, the DHHS Office of Inspector General (OIG) criticized IRBs' 'minimal field presence...[for] as a result, IRBs typically have little basis to know for themselves how research teams approach the informed consent process, how well human subjects understand the implications of their participation in research, and how fully teams remain true to the research design set forth in their approved protocols' (OIG 1998b, 8). Nevertheless, there are cases in which IRBs have developed systems for monitoring compliance. For example, one IRB hired an intermediary, a person who observes the consent process in psychiatric research studies. Another IRB randomly selects studies from a list of approved high-risk research and audits research records, concentrating, however, primarily on consent forms.21 The President's Commission also noted the need for verification and the difficulty it can present for IRBs. It suggested that institutions, rather than IRBs, develop mechanisms for receiving complaints and conducting investigations. As articulated by the President's Commission, 'the duty ought to be that of the institution itself but need not necessarily be lodged with the IRB' (President's Commission 1981, 80). Institutions could choose to delegate these functions to an IRB or to an entity other than the IRB. The results of all investigations, however, should be reported to the IRB.

Another advantage to having someone other than the IRB monitor investigator compliance is that investigators may be reluctant to ask informal questions of the IRB if there is a chance that a question may trigger closer monitoring of a research study. Likewise, IRBs also hesitate to seek advice from OHRP/OPRR on difficult cases, because such an inquiry might trigger an investigation.22 Assessing compliance of investigators is an important part of protecting research participants and should be taken seriously as a responsibility of each institution. Although a number of practical issues arise in monitoring investigators while they conduct their research studies, investigators, IRBs, and institutions should discuss these issues and provide input into the regulatory process.

Recommendation 3.5:

The process for assuring compliance with federal policy should be modified to reduce any unnecessary burden on institutions conducting research and to register institutions and Institutional Review Boards with the federal government. The assurance process should not be duplicative of accreditation programs for institutions (see Recommendation 3.4).

Recommendation 3.6:

Institutions should develop internal mechanisms to ensure Institutional Review Board compliance and investigator compliance with regulations, guidance, and institutional procedures. Mechanisms should be put in place for reporting noncompliance to all relevant parties.

Back to Top

Conflicts of Interest

Conflicts of Interest Involving Investigators

A research setting that involves human participants necessarily creates a conflict of interest for investigators who seek to develop or revise knowledge by using individuals to obtain that knowledge. In the words of the National Commission:

The Commission's deliberations begin with the premise that investigators should not have sole responsibility for determining whether research involving human subjects fulfills ethical standards. Others, who are independent of the research, must share this responsibility, because investigators are always in positions of potential conflict by virtue of their concern with the pursuit of knowledge as well as the welfare of the human subjects of their research (National Commission 1978, 1).

Overzealous pursuit of scientific results could lead to harm if, for example, investigators design research studies that pose unacceptable risks to participants, enroll participants who should not be enrolled, or continue studies even when results suggest they should have been modified or stopped. In addition, decisions about study design, inclusion criteria, and adverse events require judgment and discretion on the part of the investigators, who can be influenced, consciously or unconsciously, by conflicting interests. Thus, it is important to address prospectively the potentially harmful effects on participants that conflicts of interest might cause (Emanuel and Steiner 1995; Lo 2000).

IRB review of research studies is one method for identifying and dealing with conflicts of interest that might face investigators. By having IRBs review research studies prospectively and following an IRB-approved protocol, investigators and IRBs together can manage conflict between the investigators' desire to advance scientific knowledge and the need to protect the rights and welfare of research participants. IRBs should, for example, determine whether investigators are being paid for enrolling participants in a research study or whether they or their family members have a proprietary interest in the outcome of the research. Steps to strengthen the IRB oversight system should reduce the probability that investigators' conflicts of interest will cause harm to research participants.

It is worth exploring the problems that result when successful research creates benefits, such as prestige, power, and profit, for those who conduct the research studies. Although conflicting interests associated with obtaining promotion and tenure have always been present for academic investigators, recent changes in the research enterprise increasingly have created opportunities for investigators and institutions to profit financially from research (Steiner 1996). Investigators' financial conflicts of interest include capitated payments or bonuses for enrolling participants, indirect payments through consul-tantships or honoraria, and equity holdings in companies or royalties from patents whose value may be affected by the research. Institutions' conflicts of interest also include equity holdings in companies and the economic benefits of patents they hold.

As a result, concern is growing that financial conflicts of interest may compromise scientific objectivity and the welfare of research participants, ultimately leading to decreased public trust in the research enterprise (Bodenheimer 2000; Shalala 2000). The factors contributing to these financial conflicts (e.g., erosion of the academic medical center's revenue base, encouragement by government to transfer technology into the marketplace, and industry's increasing financial support of research) are unlikely to wane. In fact, these pressures on investigators and institutions are likely to increase.

On the surface, financial conflicts are tangible and extraneous and therefore might be more easily managed or eliminated than some other conflicts (Angell 2000a; Angell 2000b). However, simply prohibiting financial conflicts may not be feasible in an environment in which private funding of research and new strategies for transferring technology to the marketplace are being encouraged (Emanuel and Steiner 1995; Korn 2000). Thus, managing financial conflicts of interest will remain an important part of preserving the public's trust in research.

IRB review alone, however, is not sufficient to manage financial conflicts, because the options available to IRBs to eliminate such conflicts are limited. IRBs can disapprove a research study in which the financial conflicts of interest of the investigator are perceived to be too high, or IRBs can require that study design or methods be altered, but they cannot dictate directly the conduct of investigators concerning their financial interests. Moreover, although financial conflicts of interest raise concerns about bias and the reliability of research results (Bodenheimer 2000), attention to these concerns should not rest solely on IRB review. Thus, the responsibility for the review of investigators' conflicts should also lie with sponsors and institutions. National Bioethics Advisory Commission

Some have proposed disclosure of conflicts of interest to institutions (AAMC 1990; AMA 2000), IRBs (DeRenzo 2000), and participants (Cohen 2000) and subsequent management of these disclosed interests as a mechanism for dealing with financial conflicts (AAMC 1990). Although disclosure might encourage investigators to think carefully before agreeing to arrangements that pose conflicts or might provide others, such as institutional officials, an opportunity to assess the risks and potential benefits of financial arrangements (Blumenthal 1992), it is not a complete solution (Lo 2000; Thompson 1993).

Organizations, particularly academic institutions, should become more actively involved in managing investigators' financial conflicts and should increase their self-regulation efforts in this area. Several recently published articles indicate that policies governing conflicts of interest at major academic research institutions vary widely and often lack basic safeguards, such as providing clear guidance regarding the types of relationships that are prohibited (Cho et al. 2000; Lo et al. 2000; McCrary et al. 2000). Federal regulations already require recipients of certain federal funds to establish policies for disclosure, reduction, management, and elimination of financial conflicts of interest (42 CFR 50.601 - 50.607). As a first step, the regulations and the institutional policies responding to them could be improved (Cho et al. 2000; Cohen 2000). FDA also requires financial disclosure by investigators as part of its review of marketing applications (21 CFR 54).

Some financial conflicts may be inappropriate regardless of the amount of oversight provided either by the institution or the IRB. For example, certain per-capita bonuses are designed solely to provide an incentive to recruit participants; when such payments are excessive they serve only to provide an inappropriately strong incentive to investigators or referring physicians (Lind 1990; Wolf and Lo 2000). Or, in a clinical trial, it might be unacceptable for investigators to own stock or options in the company manufacturing the product being tested, because their remuneration might be greatly influenced by whether the trial results are favorable or unfavorable (Healy et al. 1989; Lo et al. 2000; Topol et al. 1992). The concern would be greatest in a start-up company that has not yet made a public offering. However, the lack of guidance and policy makes it difficult for individual institutions to set strict standards, especially ones that include the prohibition of certain types of financial relationships. Institutions should work with professional societies and sponsors to clarify types of conflicts and develop mechanisms for managing such conflicts through education and oversight.

IRBs should be aware of investigators' financial arrangements with respect to research under review (DeRenzo 2000). Most academic medical centers, however, require investigators to disclose financial conflicts of interest to a university official or to a committee, but not to the IRB (Lo et al. 2000; McCrary et al. 2000). Payments to the investigator should not, for example, affect an IRB's evaluation of the risks and potential benefits of a research study or the equitable selection of participants. Rather, the presence of financial conflicts of interest might affect an IRB's assessment of the protocol in its entirety and whether the research should be approved, or it might affect its assessment with respect to the amount or type of monitoring needed (AAMC 1990; Gottlieb 2000).

One area in which an IRB must be involved is determining what information about financial conflicts of interest should be shared with research participants as part of the informed consent process. Recently, the Association of American Medical Colleges added its support to the policy of disclosing of investigators' financial conflicts to research participants (Cohen 2000). A number of concerns surround the issue of disclosure, primarily related to the privacy of investigators and the relevance and understandability of the information to participants. Participants clearly need to understand the nature of the research study in which they are participating, including who is likely to benefit from it. Such information may be relevant to an individual who is deciding whether to participate in a research project.23 However, disclosure to participants should not serve as a substitute for the institutional management of conflicts of interest. Disclosure, although necessary, is not sufficient for managing and resolving these issues.

Conflicts of Interest Involving IRBs

Independent review is the primary mechanism for managing conflicts of interest in research, and assuring the independence of IRBs has been a matter of concern since their inception. The federal regulations require two types of independence for IRBs. First, no IRB member should participate in the review of "any project in which the member has a conflicting interest" (45 CFR 46.107(e); 21 CFR 56.107(e)). Second, to prevent the IRB from representing solely an institutional viewpoint, at least one IRB member must be unaffiliated with the institution (45 CFR 46.107(d); 21 CFR 56.107(d)). These two types of independence specified in the federal regulations reflect concerns about the personal conflicts of individual IRB members and the conflicts an IRB may have as part of the institution.

Individual IRB Members

Conflicts of interest for individual IRB members are currently addressed in the federal regulations by a requirement that "no IRB may have a member participate in the IRB's initial or continuing review of any project in which the member has a conflicting interest, except to provide information requested by the IRB" (45 CFR 46.107(e); 21 CFR 56.107(e)). The most obvious case of conflict would occur when an IRB member reviews a protocol for which he or she is an investigator. However, the regulations are often interpreted to include conflicts that involve finances, collaborations, or even personal relations. NBAC supports this more inclusive interpretation.

Financial and other obvious conflicts of interest for IRB members, such as collaboration in a research study, are often less difficult to identify and manage than some of the more subtle and pervasive conflicts (Francis 1996). Subtle conflicts are often not detected. For example, IRB members may have intellectual or practical commitments related to a research study with which they are not directly involved. IRB members may also have professional or personal connections to the investigators whose research they review. Although familiarity with the research was an intended part of the IRB system, it can also cause IRBs to become enmeshed in institutional, interdepartment, or interpersonal interests. Thus, a trade-off might occur on IRBs between the level of expertise of its members and their independence. In some cases, IRB members may fear that an unfavorable review might result in personal repercussions. Many IRBs try to keep board discussions confidential in order to protect individual members, but some IRB members have expressed discomfort in reviewing the work of direct supervisors, for example.

Resolving these tensions is not always easy. In cases of clear, strong conflict (e.g., an IRB member's participation in a research study as an investigator), the member with the conflict can recuse him or herself from the review. However, in cases of weaker conflict (e.g., close working relationship with the investigator; particular academic or other commitments, but no actual involvement in the study), disclosure of potential biases to the IRB should be required and should be sufficient. The IRB should be confident that each member voting on a specific protocol provides a knowledgeable, independent assessment of the research under review.

Guidance should be developed to assist IRBs in identifying various types of conflict. In addition, IRBs can reduce conflicts that employees of the institution may have by increasing the number of members who are not otherwise affiliated with the institution and the number of members who are nonscientists and are more likely to represent the views of the prospective participants rather than the investigators.

The IRB as an Institutional Body

IRBs are institutional entities (Annas 1991), a feature that was intended to reflect an institution's commitment to safe, ethically sound research. An institution designates an IRB and gives it authority to review research, although the institution retains the power to disapprove research that has been approved by an IRB (45 CFR 46.112; 21 CFR 56.112). This structure applies even when an institution relies on an independent IRB or an IRB at another institution. The IRB's authority still derives from the institution that has delegated the power of review. Because most institutions have an understandable desire to increase their research activities, institutional IRBs may, themselves, face conflicts of interest. And, even independent IRBs have a strong incentive to consider the interests of their institutional contractors. As Leslie Francis notes: 'Consistently unfriendly reviews might be thought to threaten ongoing relationships between IRBs and the institution [sic] with which they contract' (Francis 1996, 424).

Some have suggested that the primary function of IRBs has shifted from protecting the participant to protecting the institution (Annas 1991). In general, institutional interests in protecting themselves from liability, bad publicity, and regulatory sanctions are consistent with an interest in the protection of research participants. However, institutions can have other interests that conflict with the goal of participant protection, and such conflicts might increase when institutions hold equity in companies whose products are tested in research by their investigators or when they have patent licensing arrangements whose value may be affected by specific studies. Institutions therefore aim to maintain a research-friendly environment. Institutions can reduce such conflicts by divesting questionable holdings, by not conducting research that may affect those investments, or by conducting research that affects those investments only as part of a multi-site trial or with external monitoring (Emanuel and Steiner 1995). However, the primary strategy for limiting the influence of inappropriate institutional interests on IRBs is to impose requirements on IRB membership. For an IRB, being research friendly can mean not only being efficient in handling paperwork, but also being less stringent in its review than it should be. An IRB that disapproves protocols, that often requires significant modifications, that requires investigators to provide frequent and detailed information on research studies in progress, or that conducts frequent continuing review might be perceived as overly intrusive or bothersome.

The involvement of particular institutional staff on IRBs has also been questioned. Recently, OHRP suspended assurances at some institutions where institutional grants and contracts officials served as IRB members. OHRP argued that staff responsible for bringing research funds into an institution have a fundamental conflict of interest when they serve on the IRB (OHRP 2000). Such strategies challenge the balance between functioning independently and having members who are able to provide relevant information about investigators or their research. Institutional counsel is another example of a potential source of conflict. An institution's attorney is generally concerned with protecting the institution from liability and might require that certain language be included in the consent form to protect the institution, even though the language might not be understandable to prospective participants.

IRB membership is one mechanism for reducing institutional conflicts of interest. Current federal regulations require that each IRB have 'at least one member who is not otherwise affiliated with the institution and who is not part of the immediate family of a person who is affiliated with the institution" (45 CFR 46.107(d); 21 CFR 56.107(d)). Interpreting the definition of not otherwise affiliated literally can result in only a small degree of separation between the individual and the institution. For example, ex-employees are sometimes chosen to serve as the unaffiliated member. Moreover, unaffiliated is often interpreted as "representing the community or participants,' which is incorrect, because unaffiliated members can be scientists; other professionals, such as lawyers, business people, or clergy; or individuals who represent participants. The only characteristic that unaffiliated members share other than participating on the IRB is lack of ties to the institution.

Concerns have been raised regarding whether having only one unaffiliated member on an IRB is sufficient to avoid institutional influence, especially when IRBs have 15 to 21 members on average.24 One or two "outsiders" might hesitate to mention concerns or challenge the group (McNeill 1993), and unaffiliated members do not need to be present in order for the IRB to conduct review and approve research studies.25 Thus, IRBs can approve research with only institutional representation present as long as a nonscientist and a quorum are also present.

Others have proposed requiring the presence and participation of more unaffiliated IRB members to reduce the influence of institutional interests on IRB decision-making (OIG 1998a). However, finding appropriate unaffiliated members who are willing to serve on an IRB can be difficult. Paying unaffiliated members for their efforts, as originally proposed by the National Commission (National Commission 1978, 15) might improve the yield, but excessive compensation could call their independence into question. Currently, no rules or guidance are available describing criteria to meet the definition of unaffiliated, how long such members should serve, under what circumstances they may be removed, or what payment they should be provided. Institutions should be careful to select unaffiliated members who are truly separated from the institution, except for their role on the IRB. Procedures for the selection and removal of unaffiliated members should be established in a way that empowers the independent voices of those members, and providing reasonable payment to IRB members who are otherwise unaffiliated with the institution can be a valuable way to strengthen their role.

Another strategy for reducing institutional influence on an IRB is to use a noninstitutional IRB. These IRBs review research studies for a number of institutions and are, therefore, presumed to be less likely to be influenced by any one institution. It is claimed that one of the advantages of this system is the almost complete elimination of influence by individual institutional interests.26 However, others have argued that paying IRBs for review of research, as the United States does when noninstitu-tional (independent) IRBs are used, constitutes a conflict of interest.27 It appears that neither system of review, institutionally based IRBs or fee-for-service independent IRBs, is free from all real or perceived conflicts of interests. However, it does seem that institutionally based IRBs can reduce conflicts of interest by including more members who are not affiliated with the institution. Independent IRBs, like institutions, can also reduce conflicts by separating business functions from review functions and by using agreements that clearly describe what is covered by the fee (i.e., the fee is for review only and clearly does not guarantee approval).

Another way to limit the influence of an institution's research interests on an IRB is to ensure that the views of both scientists and nonscientists are considered. Current regulations require that each IRB include 'at least one member whose primary concerns are in scientific areas and at least one member whose primary concerns are in non-scientific areas.' A scientist is defined by OHRP as a person who has "substantive training or experience in a scientific discipline or in the scientific method.' 28 Thus, physicians, nurses, or individuals trained at the bachelor's, master's, or doctoral level in the basic sciences or social sciences are all considered scientists. Unlike the requirement for unaffiliated members, at least one nonscientist must be present for an IRB to review research (45 CFR 46.107(c); 21 CFR 56.107(c)).

As with unaffiliated members, it is sometimes assumed that nonscientist members represent the views of the participants. However, nonscientists might or might not be members from the community of participants. Professionals who are nonscientists might be no more able to represent the view of low-income or seriously ill participants, for example, than scientists.

From an ethical perspective, balancing the views of scientists and nonscientists is important, because although IRBs must be technically competent to identify and quantify the risks and potential benefits of a protocol, they must also be able to put those risks and benefits into a social context in order to judge how ordinary citizens would value them. This kind of assessment requires the input of nonscientists because "experts have a unique commitment to research that favours the interests of science" (McNeill 1993, 184). This is not to suggest that scientists and nonscientists have different roles on an IRB, although nonscientists often do tend to focus on the adequacy of consent documents (Porter 1987). It is important to remember that all IRB members should evaluate protocols according to established ethical principles and the federal regulations. It is also important to remember that they cannot avoid doing so from different perspectives and with different emphases.

The current IRB system requires that unaffiliated and nonscientist members serve on these groups. Although each brings valuable experience, knowledge, and insight to the IRB, neither may reflect the views of the research participants. For this reason, IRBs should include members who are specifically chosen because they represent participants' interests.

IRBs should strive to complement their membership with members who are clearly unaffiliated with the institutions, members who are nonscientists, and members who represent the perspectives of participants. Federal policy should specify membership and quorum requirements for each of these categories. Individuals who meet the membership criteria in more than one of these three categories should be able to fulfill the requirements for each. For example, a member of the clergy can fulfill both the nonscientist and unaffiliated membership requirements. NBAC recommends that an IRB have collectively at least 25 percent of its membership comprised of individuals representing these three categories.

Other IRB Composition Issues

Scientific or discipline expertise on an IRB is clearly needed to judge the risks and potential benefits of the research protocols it reviews. However, scientific expertise is to a large extent discipline specific, meaning that it is essential that at least some IRB members have expertise and experience that is directly related to the research being reviewed. Indeed, federal regulations require that IRBs possess "the professional competence necessary to review specific research activities" (45 CFR 46.107; 21 CFR 56.107). NBAC heard repeatedly from investigators conducting research in the social sciences and humanities that IRBs with primarily medical expertise sometimes impose unreasonable, and even irrational, requirements that are inappropriate for nonbiomedical research (AAUP 2000). Although ethical principles for review should be consistent for all research, factors such as standards for research design, participants' expectations of privacy and confidentiality, and the appropriate consent processes to be used can, and often should, vary by type of research. Appropriate expertise and experience are essential both for the protection of participants and the promotion of ethical research.

At the same time, however, IRBs are intended to be interdisciplinary and may not "consist entirely of members of one profession" (45 CFR 46.107(b); 21 CFR 56.107(b)). Thus, an appropriate balance must be found, which for large institutions might mean multiple IRBs with specific areas of expertise and which for smaller institutions might mean one IRB with a varied membership and a roster of consultants. In either case, institutions should have discretion in determining the most appropriate arrangement.

Recommendation 3.7:

Federal policy should define institutional, Institutional Review Board, and investigator conflicts of interest, and guidance should be issued to ensure that the rights and welfare of research participants are protected.

Recommendation 3.8:

Sponsors and institutions should develop policies and mechanisms to identify and manage all types of institutional, Institutional Review Board, and investigator conflicts of interest. In particular, all relevant conflicts of interest should be disclosed to participants. Policies also should describe specific types of prohibited relationships.

Recommendation 3.9:

Federal policy should establish standards and criteria for the selection of Institutional Review Board members. The distribution of Institutional Review Board members with relevant expertise and experience should be commensurate with the types of research reviewed by the Institutional Review Board (see Recommendation 3.10).

Recommendation 3.10:

Institutional Review Boards should include members who represent the perspectives of participants, members who are unaffiliated with the institution, and members whose primary concerns are in nonscientific areas. An individual can fulfill one, two, or all three of these categories. For the purposes of both overall membership and quorum determinations 1) these persons should collectively represent at least 25 percent of the Institutional Review Board membership and 2) members from all of these categories should be represented each time an Institutional Review Board meets (see Recommendation 3.9).

Back to Top

Summary

This chapter discussed the important role that education plays in protecting research participants. Individuals should be able to demonstrate that they understand their ethical obligations and how to carry them out. To help investigators and IRBs fulfill their responsibilities, the federal government should promote the development of education and certification programs that apply to all investigators as well as all IRB members and staff.

Institutions and independent IRBs must do a better job of meeting their ethical obligations to research participants. Accreditation programs are one way in which all organizations involved in conducting or reviewing research can develop highly efficient and effective protection programs.

A key protection is that of ensuring institutional, IRB, and investigator compliance. At all levels in the system, various mechanisms, including assurances of compliance, site inspections, and internal audits, should be used. Regular periodical monitoring of compliance is the most likely way to detect and remedy problems.

Institutions and IRBs increasingly are confronting conflicts of interest as more investigators and institutions enter into financial arrangements in which they stand to benefit from the results of their research. Conflicts of interest are not new, but they have changed and intensified as the research enterprise has evolved. Disclosure and management of investigator and institution conflicts still seem to be the best strategies. Conflicts affecting the IRB can be handled by increasing the percentage of members who do not have any direct interests in the institution or its research program (unaffiliated members). Increasing the percentage of nonscientists and members who represent the views of participants can also reduce conflicts. Ultimately, clear policies and guidance will provide the best way for IRBs to fulfill their responsibilities and meet their obligations in educating their members, monitoring for compliance, and avoiding conflicts of interest.

Back to Top

Footnotes

• 1 NIH Guide for Grants and Contracts, 'Short Term Courses in Research Ethics.' PA-99-051. January 22, 1999; NIH Guide for Grants and Contracts, 'Mentored Scientist Development Award in Research Ethics.' PAR-98-006. November 7, 1997. Both available at http://grants.nih.gov/grants/guide. Last accessed December 4, 2000.
• 2 PRIM&R, 'IRB 101 On the Road.' Available at http://www.primr.org/ 101road.html. Last accessed December 4, 2000.
• 3 NIH Guide for Grants and Contracts, 'Required Education in the Protection of Human Research Participants.' Notice OD-00-039. June 5, 2000. Available at http://grants.nih.gov/grants/guide. Last accessed December 4, 2000.
• 4 See Peckman, S., 'Local Institutional Review Boards.' This background paper was prepared for NBAC and is available in Volume II of this report.
• 5 Association of American Medical Colleges Medical School Graduation (GQ) Questionnaire, 1980 - 1996.
• 6 NBAC Town Meetings: February 9, 2000, Houston, Texas; April 5, 2000, Pittsburgh, Pennsylvania; May 3, 2000, Orlando, Florida; June 7, 2000, Chicago, Illinois; August 14, 2000, Portland, Oregon.
• 7 See http://www.ptcny.com/PDF/ccip2000.pdf. Last accessed May 1, 2001.
• 8 See http://www.primr.org/aahrpp.html. Last accessed May 1, 2001.
• 9 NBAC Town Meetings: February 9, 2000, Houston, Texas; April 5, 2000, Pittsburgh, Pennsylvania; May 3, 2000, Orlando, Florida; June 7, 2000, Chicago, Illinois; August 14, 2000, Portland, Oregon.
• 10 Ibid.
• 11 See McCarthy, C.R., 'Reflections on the Organizational Locus of the Office for Protection from Research Risks.' This background paper was prepared for NBAC and is available in Volume II of this report.
• 12 NBAC Town Meetings: February 9, 2000, Houston, Texas; April 5, 2000, Pittsburgh, Pennsylvania; May 3, 2000, Orlando, Florida; June 7, 2000, Chicago, Illinois; August 14, 2000, Portland, Oregon.
• 13 Ibid.
• 14 OHRP Guidance: 'Procedures for Registering Institutional Review Boards and Filing Federal Wide Assurances of Protection for Human Subjects.' December 3, 2000. Available at http://ohrp.osophs.dhhs.gov/irbasur.htm. Last accessed December 8, 2000.
• 15 See http://ohrp.osophs.dhhs.gov/humansubjects/assurance/faq.htm.
• 16 Letter from D.A. Lepay, Food and Drug Administration, to E.M. Meslin, NBAC. February 20, 2001. Providing comments on a draft of this report.
• 17 Lepay, D., 2000, FDA's Program for Human Subject Protection. 16 October. Presentation to ORCA Teleconference.
• 18 NBAC, 'Federal Agency Survey on Policies and Procedures for the Protection of Human Subjects in Research.' This staff analysis is available in Volume II of this report.
• 19 NBAC Town Meetings: February 9, 2000, Houston, Texas; April 5, 2000, Pittsburgh, Pennsylvania; May 3, 2000, Orlando, Florida; June 7, 2000, Chicago, Illinois; August 14, 2000, Portland, Oregon.
• 20 Ibid.
• 21 NBAC Town Meeting: April 5, 2000, Pittsburgh, Pennsylvania.
• 22 NBAC Town Meetings: February 9, 2000, Houston, Texas; April 5, 2000, Pittsburgh, Pennsylvania; May 3, 2000, Orlando, Florida; June 7, 2000, Chicago, Illinois; August 14, 2000, Portland, Oregon.
• 23 Moore v. The Regents of the University of California et al. 793 P.2d 479. (Cal. 1990).
• 24 NBAC Town Meetings: February 9, 2000, Houston, Texas; April 5, 2000, Pittsburgh, Pennsylvania; May 3, 2000, Orlando, Florida; June 7, 2000, Chicago, Illinois; August 14, 2000, Portland, Oregon.
• 25 39 Fed. Reg. 18915.
• 26 See Holm, S., 'The Danish Research Ethics Committee System, Overview and Critical Assessment'; Chalmers, D., 'Research Ethics in Australia'; and Heath, E., 'The History, Function, and Future of Independent Institutional Review Boards.' These background papers were prepared for NBAC and are available in Volume II of this report.
• 27 See Heath, E., 'The History, Function, and Future of Independent Institutional Review Boards.' This background paper was prepared for NBAC and is available in Volume II of this report.
• 28 'Instructions for Registering an IRB or IEC.' Available at http://OHRP.osophs.dhhs.gov. Last accessed March 7, 2001.

National Academy of Engineering

500 Fifth Street, NWWashington, DC 20001T. 202.334.3200F. 202.334.2290