- Email: [email protected]
The Controversy Surrounding Central Institutional Review Boards
Ethical/Legal Feature
The Controversy Surrounding Central Institutional Review Boards Feature Editor Introduction: Kristi L. Kirschner, MD For this column, we are tackling an issue that many in rehabilitation may not be familiar with—namely, central institutional review boards (CIRBs). As rehabilitation research continues to grow, particularly in collaboration with pharmaceutical and device manufacturers, CIRBs will probably play an increasingly prominent role. Given the likely unfamiliarity of the issues posed by CIRBs, what follows is a brief commentary (by me, and my colleagues Teresa A. Savage, PhD, RN, and Debjani Mukherjee, PhD, of the Rehabilitation Institute of Chicago Donnelley Family Disability Ethics Program and Northwestern University Feinberg School of Medicine Medical Humanities and Bioethics Program) that lays out the issues, followed by expert commentaries and responses by Dr. John Whyte and Dr. Norman Harden. As always, we invite your comments and responses to this column!
INTRODUCTION A debate has ensued about the use of central or commercial institutional review boards (CIRBs) for research. At the heart of the matter is research integrity, public trust, and, of course, protection of human subjects. The 2 quotes below exemplify the debate: At a time when commercial interests threaten the safety of research participants and the integrity of medical research, it is remarkable that North American regulatory agencies have not seen any problem with entrusting the rights and well-being of human research participants to a lightly regulated commercial enterprise. . ..The protection of research participants is a critically important public mandate, and it merits a truly independent regulatory structure [1]. Making a distinction about the quality and merits of IRBs based on tax status is an antiquated piece of ideology reminiscent of Orwell’s Animal Farm—“for-profit bad, not-for-profit good.” The distinction obscures what we should care about and directly measure—the quality of IRB review of protocols and the monitoring of the safety of research participants [1]. In this column, we discuss the use of central or commercial (CIRBs) and summarize the historical context and debate. Throughout this feature, we welcome the reader to consider how the role of CIRBs may impact science and clinical care in the United States.
Contributors: Teresa A. Savage, PhD, RN Rehabilitation Institute of Chicago and Donnelley Family Disability Ethics Program, Chicago, IL Disclosure: nothing to disclose Debjani Mukherjee, PhD Rehabilitation Institute of Chicago, Donnelley Family Disability Ethics Program and Northwestern University Feinberg School of Medicine, Chicago, IL Disclosure: nothing to disclose
Feature Editor: K.L.K. PM&R, and Medical Humanities and Bioethics, Northwestern University Feinberg School of Medicine, Chicago, IL. Address correspondence to: K.L.K.; e-mail: [email protected] Disclosure: nothing to disclose
BACKGROUND Investigative journalists have largely led the way in elucidating research projects gone awry. When we open a newspaper or read an online account, questions are raised about how the projects were approved, who was monitoring the safety and effects, and how conflicts of interest may have impacted the process. The IRB is a committee developed to oversee research that involves human subjects, with the charge of protecting the rights and welfare of research participants. The need for IRBs was recognized in the wake of egregious examples of abuse in research (eg, Nazi physicians during World War II [2], the Tuskegee Syphilis Study [3], and the Willowbrook hepatitis vaccine study [4]), and mandated by Title 45 CFR 46 (Code of Federal Regulations) Part 46. The requirements for IRBs are set forth by law and regulated by the Office of Human Research Protections within the U.S. Department of Health and Human Services. PM&R 1934-1482/10/$36.00 Printed in U.S.A.
Guest Discussants: J.W. Moss Rehabilitation Research Institute, Albert Einstein Healthcare Network, Elkins Park, PA Disclosure: nothing to disclose R.N.H. Rehabilitation Institute of Chicago and Department of Physical Medicine and Rehabilitation, Northwestern University, Evanston, IL Disclosure: 2A, 7B, 8B Disclosure Key can be found on the Table of Contents and at www.pmrjournal.org
© 2010 by the American Academy of Physical Medicine and Rehabilitation Vol. 2, 57-63, January 2010 DOI: 10.1016/j.pmrj.2009.12.013
57
58
Kirschner et al
Historically, IRBs were set up by universities and medical institutions involved in human-subjects research. During the last 40 years, there has been the addition of for-profit or commercial IRBs, which are also subject to the same federal laws and regulations. [Although most CIRBs are commercial, there are a few exceptions, such as the National Cancer Institute IRB (http://www.ncicirb.org). For the purposes of this column, we are concentrating on for-profit, commercial (or central) IRBs and will refer to them as CIRBs.] CIRBs are being used increasingly for pharmaceutical and medical device clinical trials. From 2000 to 2006, clinical trials in the United States increased nearly 50% to 59,000 drug trials [5]. Research moved from academic medical centers to clinical research organizations or private physician offices. One of the oldest and most respected CIRBs, the Western IRB (WIRB) “reviews more than half of all trials of new drugs submitted for FDA approval. . .” and some academic medical centers are outsourcing their IRB work to CIRBs [5]. Cancer researchers have been the most vocal advocates for the use of CIRBs because they focus on the complexity of the trials, the duplication of reviews, and the ease of coordinating multisite, multi-investigator research. Although a growing number of academic and governmental institutions use CIRBs, some core ethical issues have been raised, including protection of human subjects, actual and perceived conflicts of interest, and balancing the specialized knowledge and speed of CIRBs, with the local knowledge and due process of local IRBs. Regardless of the type of IRB used, harm to research subjects has been documented, and regulations or mandates have been developed. For example, on January 15, 1994, then-President Clinton formed the Advisory Committee on Human Radiation Experiments. These experiments took place for 40⫹ years after World War II and the Cold War. The Advisory Committee on Human Radiation Experiments issued a detailed report of the multiple abuses that took place, including infringement of informed consent and the abuse of vulnerable populations [6]. In the more recent past, there are a number of questions raised about erosion of protections with current practices involving U.S. researchers who engage in clinical trials in developing countries [7]. Additionally, several high-profile deaths of research subjects have raised issues to consider. The deaths include Jesse Gelsinger at the University of Pennsylvania in 1999 [8], the death of a healthy volunteer Ellen Roche in an asthma study at Johns Hopkins in 2001 [9], and the death of Jolee Mohr, a 35-year-old woman in 2007 at the University of Chicago involved in a gene therapy study for rheumatoid arthritis [5]. Detailed reviews of these deaths revealed concerns about safety with the protocols, adherence to protocol, and the informed-consent processes. Some evidence also exists that drugs and devices that are approved by the Food and Drug Administration (FDA) have a high incidence of postmarketing safety concerns. Examples of drugs that have been withdrawn from the market after
THE CONTROVERSY SURROUNDING CENTRAL IRBS
FDA approval, sometimes years later, include rofecoxib, valdecoxib, trovafloxacin, mibefradil, terfenandine, troglitazone, cerviastatin, alosetron, and tegaserod. Other drugs have had black box warnings added after approval, such as selective serotonergic reuptake inhibitors (for suicide risk with adolescents) and rosiglitazone (for congestive heart failure and myocardial infarction risk). Some of the devices with postmarking safety concerns include drug-eluting stents, biventricular pacers, and implantable defibrillators. Since the passage of the Prescription Drug User Fee Act in 1992 (which allowed the pharmaceutical industry to pay a user fee to expedite the review of the drug), there has been marked increase in rates of postapproval safety problems in drugs that were approved in the 2 months before the approval deadline imposed by the Prescription Drug User Fee Act as compared with drugs approved in other months of the review cycle (odds ratio of 5.5 for drug withdrawals, blackbox warning odds ratio 4.4, and discontinuation of at least one dosage from 3.3—all P ⱕ .02) [10]. In the analysis of many scholars who study the problems of drug safety in the United States, there have been substantial concerns raised about the undue influence of money in the erosion of protections of human subjects, particularly when pharmaceutical companies and device makers are involved [8,11-14]. The erosions of protections can occur in a number of ways, including: (1) independence of the researcher (financial ties, ownership of data, writing manuscripts, reporting harms); and (2) independence of the review bodies (eg, the IRB, FDA).
COMPLIANCE ISSUES Federal regulations on the conduct of research apply to both local IRBs and CIRBs. In addition to protecting research participants, local IRBs also are motivated to uphold their high standards to protect the reputation of their institution. CIRBs also desire to protect their reputations, although any missteps or transgressions by the CIRB may not be public information, unlike for local IRBs. One needs only to search the Office of Human Research Protections Compliance Oversight webpage to find institutions that have been out of compliance. In the late 1990s and early 2000s, there were a number of university IRBs that had research activities suspended by the NIH Office for Protection from Research Risk (now Office for Human Research Protections). Duke University, Rush University, University of Rochester, and University of Illinois at Chicago are a few whose research was suspended and all active protocols had to be re-reviewed. There have been lawsuits against IRBs, such as in the melanoma vaccine case in Tulsa, Oklahoma, and the efalizumab (Raptiva) case, a psoriasis medication [15]. At the Medical College of Georgia, a psychiatrist stole more than $10 million of research money and allowed untrained employees to administer experimental drugs. The university IRB was unaware of the
PM&R
Vol. 2, Iss. 1, 2010
59
Table 1. Pros and cons of a CIRB and local IRB Pros CIRB
Specialized expertise in reviewers Large, well-trained IRB staff
Consistency in reviews Perception of quick approval thereby decreasing costs and speeding up scientific discoveries Local IRB
Naive community member more representative of human subject Can coordinate with other aspects of local human research protection program, eg radiation safety, conflicts of interest On-site monitoring geographically easier; investigation of adverse event reports geographically easier
Cons Lack of local context and expertise Creates more distances between principal investigator/research team and human subject protection processes and procedures Potential coercion by drug company to agree or be excluded There are no data supporting decreased costs
Lack of specialized expertise re certain conditions or drugs Fewer staff limit number of studies reviewed per meeting, thereby increasing interval between submission and approval Responsible for on-site research even when CIRB is IRB of record
study; the CIRB (WIRB) was criticized for not conducting any on-site audits, which may have uncovered the misconduct [15]. Dr. Angela Bowen, founder and former CEO of the WIRB, takes issue with the accuracy of the report by Evans et al, but she did not set the record straight in her response [16].
Local IRBs are criticized for being slow, inefficient, and too focused on unimportant minutiae; CIRBs are criticized for their “fundamental conflict of interest. They are in a clientprovider business relationship with the commercial entities whose studies they review. Because commercial IRBs gener-
Table 2. Identifying and addressing key challenges Key challenges Assurance of review quality ● Loss of safety net of redundant review ● Insufficient attention to local concerns ● Inappropriate consent forms
Suggested strategies ● ● ● ● ● ● ● ● ● ●
Sensitivity to local context Failing to understand culture, concerns, ethnic groups
●
●
●
Liability (institutional and individual)
● ● ● ● ● ●
Identify independent indicators of quality Provide performance data Consider local IRB representation on the central IRB Provide evidence of benefits Ensure that all policies and procedures are transparent Conduct reciprocal visits Consult references Consider a demonstration project Create detailed agreements Take steps to build public trust Develop means of accessing and using local information Develop strong communication Show that the institution is fulfilling its responsibilities Develop solid policies and procedures Establish a crisp agreement between the local facility and the alternative model Ensure “due diligence” in the selection of the alternative model Provide education for those responsible for managing institutional risk Develop guidance that clearly defines the responsibilities of the local IRB when using alternative review methods
Control and accountability Possible damage to the institution’s reputation
●
Clearly define and document roles
Loss of resources Insufficient staff and funding remain for compliance audits and other local responsibilities
●
Educate administrators on the IRBs continuing importance and need for funding Concentrate on other aspects of human research protection
●
●
●
Taken from Alternative Models of IRB Review Workshop Summary Report [20].
60
Kirschner et al
Table 3. Alternative models for the IRB Organization...
has its own IRB uses another institution’s IRB uses a central IRB relies on a lead IRB is part of a group that formed its own IRB uses a commercial IRB [“commercial” vs “central” not defined] uses some combination of above options
Adapted from presentation give by M. Speers, Executive Director, Association for the Accreditation of Human Research Protection Programs, November, 2006 [21].
ate their income from clients with a direct financial interest in obtaining approval, they are affected by the very problem they are expected to curtail” [1]. As the opening quotes illustrated, some scholars have fundamental disagreements about for-profit enterprises and their role in human-subjects research. Comparing local IRBs and CIRBs distinguishes the details about what type of IRBs are used from the broader issues about research integrity, public trust, and protection of human subjects. Both local IRBs and CIRBs charge external companies a fee for the cost of the IRB review. CIRBs are often in the position to process the application more quickly than local IRBs, especially if they are reviewing similar protocols, such as for oncology or psychiatric clinical drug trials. Some proponents of CIRBs have focused on the expertise that is developed by working with similar protocols. Although the local IRB may not have the experience of solely working with pharmaceutical protocols, the diversity of backgrounds and volunteer nature of the committees are intended to add a layer of protection. There is variability in the quality of local IRBs, as well as variability in quality of central IRBs. Table 1 briefly summarizes the pros and cons in selecting a CIRB or local IRB for a particular project. Institutions that have their own IRB may consider the use of a CIRB for certain classes of projects (drug clinical trials) in which the sponsor mandates the use of CIRB to be included as a study site. Other institutions that use local IRBs may consider using CIRBs for the same reason. It usually is not a decision made by the principal investigator solely.
CONFLICTS OF INTEREST Both types of IRBs can have real and perceived conflicts of interest. The direct financial incentives of CIRBs are one area. According to one source, WIRB made $20 million in profit in 2004 [15]. Dr. Bowen contests this figure but does not give the actual figure. Moreover, public trust and perceptions of conflicts can have far-reaching impact on the reputation and credibility of institutions. Both local and central IRBs have policies on conflicts of interest. For example, the WIRB website has this paragraph about conflict of interest:
THE CONTROVERSY SURROUNDING CENTRAL IRBS
What is a conflict of interest? Sometimes a principal investigator or sub investigator may have a conflict of interest. Conflicts of interest can take many forms. The most common conflict of interest arises when a doctor who provides you with medical care is also acting as the researcher overseeing your participation in the research study. In this case, there can be a conflict in his or her duty to make medical decisions in your best interest and his or her duty to follow the research protocol. Another common type of conflict of interest is financial. For example, the principal investigator might own stock in the sponsoring company, be a paid consultant for the sponsoring company, receive speaking fees from the sponsoring company, or be paid by the sponsoring company to conduct the research. The investigator may hold a patent on the drug or device being studied and stand to benefit if the drug or device is approved for general use. You should feel free to ask if either the principal investigator or research staff has a conflict of interest. If the principal investigator or any of the research staff do have a conflict of interest, discuss how they intend to manage that conflict. Only enter the study if you are satisfied with the safeguards that have been put in place to deal with the potential conflict [17]. It appears that the research participant is responsible for asking whether any research personnel have a conflict of interest. They may not be aware that they also should ask about any conflict of interest on the part of any IRB member who reviewed and approved the protocol. Local IRBs, usually located in universities, have conflict-of-interest policies and a process of review that may include an ad-hoc committee, the Principal Investigator’s dean, and the university’s Vice President/Vice Chancellor for Research. The most recent guidance document from the U.S. Department of Health and Human Services offers recommendations and points for consideration but indicates alternative approaches may suffice if all applicable statutes and regulations are satisfied [18]. Scrutiny for conflict of interest and research goes beyond the research team and IRB. Many institutions and/or senior administrators at the institution can have financial interests in companies that wish to conduct research within their institutions. The institution stands to gain if a drug or device gets to market and proves to be lucrative. In the Gelsinger case, the University of Pennsylvania owned stock in the company involved in the gene therapy [8]. The plan for managing conflict of interest should be public so that anyone considering participating in a study in which there is a conflict of interest on the part of a research team member, institutional administrator, or the institution itself can learn of the conflict and how it will be managed.
PM&R
NATIONAL MEETINGS TO EXPLORE ALTERNATIVE MODELS In 2005, The National Institutes of Health, the Office for Human Research Protections, the Association of American Medical Colleges, the American Society of Clinical Oncology, and Department of Veterans Affairs held a workshop on alternative models of IRB review. Their summary report presents issues to consider when deciding to use an alternative model, such as a central IRB: “access to expertise, the number of participating sites, the need for consistency among participating study sites, the level of risk, the sponsorship and type of study, the size of the institution’s research program, how quickly approval is needed, the type of disease involved in a medical study, the potential for conflicts of interest, liability issues, and available resources” [19]. In addition, it identifies challenges and suggested strategies for addressing the ethical concerns (Table 2). A subsequent national conference was held a year later with the purpose to “enhance the protection of human subjects of research by encouraging the use of alternative Institutional Review Boards (IRB) models under appropriate circumstances” [20]. The participants, regulators, funders, medical educators and researchers, and stakeholder groups explored in depth the issues identified in the 2005 workshop. One presenter listed the various alternative models of IRB (Table 3) [21]. Ideally, both the local IRB and CIRB will negotiate a shared responsibility that improves the protections of human subjects without adversely affecting the efficiency of the review process. The 2006 workshop executive summary concludes with some general points to consider: ●
●
●
There should be a detailed agreement regarding the sharing of regulatory liability between the CIRB and institution; Outsourcing disengages the principal investigator and research team in the protection of human subjects and may drain resources from the local IRB; and There should be clear guidance on the use of alternative models of IRBs and harmonization among federal laws and requirements.
CONCLUSION The use of CIRBs over local IRBs is relatively new, especially in drug studies apart from cancer drugs. The decision to use a CIRB over a local IRB should be a transparent decision that offers the best protection for human subjects. Trust, transparency, and clear communication between the institution, local IRB, and CIRB are critical. Identification and disclosure of any perceived or real conflicts of interest should be disclosed in a way that potential research subjects understand and have the opportunity to carefully consider. A detailed contract indicating the
Vol. 2, Iss. 1, 2010
61
areas of shared authority and responsibility of the institution, the local IRB, and the CIRB should be negotiated. Regardless of which model of IRB is selected, the institutional home of the research has an ethical responsibility to ensure that all reasonable efforts are expended to protect human subjects, which, in turn, protects the institution.
REFERENCES 1. Emanuel EJ, Lemmens T, Elliott C. Commercial IRBs have a fundamental conflict of interest. PLoS Med 2006;3:942-943. 2. The National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research. The Belmont Report. April 18, 1979. Available at: http://www.dhhs.gov/ohrp/humansubjects/ guidance/belmont.htm. Accessed December 22, 2009. 3. Jones J. Bad Blood: The History of the Tuskegee Syphilis Experiment. New York, NY: Free Press; 1993. 4. Brody BA. The Ethics of Biomedical Research: An International Perspective. New York, NY: Oxford University Press; 1998. 5. Advisory Committee on Human Radiation Experiments. Available at: http://www.gwu.edu/⬃nsarchiv/radiation. Accessed December 22, 2009. 6. Kent DM, Mwamburi DM, Bennish ML, Kupelnick B, Ioannidis JPA. Clinical trials in sub-Saharan Africa and established standards of care. JAMA 2004;292: 237-242. 7. Gelsinger P, Shamoo AE. Eight years after Jesse’s death, are human research subjects any safer? Hast Cent Rep 2008;38:25-27. 8. Press briefing: Hopkins issues internal review committee report on research volunteer’s death, 2001. Available at: www.hopkinsmedicine. org/press/2001/JULY/010716.htm. Accessed December 22, 2009. 9. Gilbert S. Trials and tribulations. Hast Cent Rep 2008;38:14-18. 10. Carpenter D, Zucker EJ, Avorn J. Drug-review deadlines and safety problems. N Engl J Med 2008;358:1354-1361. 11. Angell M. The Truth About Drug Companies: How They Deceive Us and What To Do About It. New York, NY: Random House; 2004. 12. Bass A. Side Effects: A Prosecutor, a Whistleblower, and a Bestselling Antidepressant on Trial. Chapel Hill, NC: Algonquin Books; 2008. 13. Brody H. Hooked: Ethics, the Medical Profession, and the Pharmaceutical Industry. Lanham, MD: Rowman & Littlefield Publishers, Inc.; 2007. 14. Elliott C. The soul of a new machine: Bioethicists in the bureaucracy. Camb Q Healthc Ethics 2005;14: 379-384. 15. Evans D, Smith M, Willen L. Big Pharma’s Shameful Secret. Bloomberg Markets. November 2, 2005. Available at http://www.ahrp.org/ infomail/05/11/03.php. Accessed July 30, 2009 16. Bowen A. Research ethics boards: Error and misconception. PLoS Med 2006;3:e457-e458. 17. WIRB website Info for research subject. Available at: http://www. wirb.com/content/research_subjects.aspx. Accessed August 18, 2009. 18. U.S. Department of Health and Human Services. Final Guidance Document: Financial Relationships and Interests in Research Involving Human Subjects: Guidance for Human Subject Protection; 2004. Available at: http://www.dhhs.gov/ohrp/humansubjects/finreltn/fguid.pdf. Accessed November 12, 2009. 19. National Institute of Health, Office for Human Research Protections, Association of American Medical College, & American Society of Clinical Oncology. Alternative models of IRB review: Workshop Summary Report, 2005. Available at: http://www.hhs.gov/ohrp/sachrp/ documents/AltModIRB.pdf . Accessed July 30, 2009. 20. National conference on alternative IRB models: Optimizing human subject protection. November 19-21, 2006. Available at: http://www.aamc. org/research/irbreview/irbconf06rpt.pdf. Accessed November 11, 2009. 21. Speers M. Shared authority and responsibilities. 2006 National Conference on Alternative IRB Models: Optimizing Human Subject Protection, November 19-21, 2006, Marriott Wardman Park Hotel, Washington, DC.
62
Kirschner et al
THE CONTROVERSY SURROUNDING CENTRAL IRBS
Commentary from John Whyte, MD, PhD In recent years, there have been increasing complaints that our institutional research board (IRB) system fails to balance protection of human subjects with preservation of the scientific enterprise. Previous ethical lapses have led to ever more voluminous and complex review processes that diminish research productivity by delaying study approval and requiring diversion of staff to regulatory rather than scientific tasks. Moreover, as consent documents become longer and more complex, it’s not at all clear that they actually make the consent process more “informed,” particularly for subjects with mental health or cognitive limitations. Clearly a simpler and more streamlined system that still maintains rigorous ethical oversight would be desirable for both research participants and investigators. Central IRBs (CIRBs) have the potential to streamline and standardize the process because they may have staff members that are highly experienced in the review process, have sufficient review volume in specialized research areas to develop expertise, and can substitute a single review for multiple independent reviews of multicenter studies most typical of research sponsored by industry. Moreover, they can eliminate the need, common in multicenter projects, to have all IRBs re-review changes that are requested by each individual IRB. Weighing against these advantages, as noted in the accompanying commentary, are the potential for conflict of interest and the lack of local institutional knowledge. These themes invite further consideration, both in general, and as they apply to areas of rehabilitation research with which I am most familiar. There is no doubt that financial conflict of interest exerts biases in the research enterprise even when there is no conscious attempt to distort results [1]. Thus, the relationship between central IRBs and their sponsors is of concern. But it is worth noting that conflict of interest comes in many forms and is surely not absent from locally reviewed projects. Investigators may themselves have relationships with sponsoring entities as speakers or as expert consultants. Even in the absence of a financial relationship, an investigator who achieves notoriety through involvement with a new treatment technique may be reluctant to find such a technique ineffective. In rehabilitation research, there are many emerging treatments that can be implemented outside of a traditional Food and Drug Administration approval process. For example, many medical technologies such as exercise machines, and orthotic and prosthetic devices, need only show that they are similar to existing products and pose no clear new risks to be marketed [2]. This produces a situation in which investigators may be studying products that are already being heavily promoted by the manufacturer. In some cases, the investigator may gain access to such equipment at low cost in part because of his/her role in studying it further. In other cases, the fact that an institution has such cutting-
edge technology is a marketing advantage that they don’t wish to undercut. And because there is no consensus standard of care for many rehabilitation problems, one cannot use a deviation from this standard as a clue that research may be taking place. Although ethical viewpoints on a given study may happen to vary from institution to institution, there is no inherent reason why “research ethics” should be a local enterprise. And one might wish that all of the issues relevant to review were generic and easily distilled into a written application. But for better or for worse, local IRBs get to know individual investigators, come to understand how closely they are supervised by their departmental administration, how seriously local administrative officials take the issue of conflict of interest, etc. The IRB may be aware that an investigator is overseeing many projects with very small funded effort or may raise concerns about a current project on the basis of an event in a previous project involving that investigator or department. In addition, the ethics of a study often are intertwined with the processes of the local clinical system. For example, review of a study with aphasic subjects may be considered quite differently if the IRB knows that the institution has a large cohort of patients with aphasia, has highly trained recruiters who know how to assess language comprehension, and has experience in developing appropriate graphic materials to support the informed consent process. Given the apparent advantages as well as risks of a CIRB process, a combined review, as alluded to in the accompanying article, seems most desirable if it can be efficiently coordinated. One form this might take is the identification of issues of concern by local IRBs, but the delegation of solutions to those issues to a CIRB. In other words, one or more IRBs could identify a consent form as too complex without each suggesting specific and different edits. Individual IRBs could identify concerns about methodology and safety, but the CIRB could manage the response to those concerns. This would allow simultaneous review by many local IRBs, forwarding of a full set of concerns to the CIRB for resolution, and return of a modified protocol and consent forms to all IRBs. The remaining question is whether each IRB would need the authority to approve or reject each project as modified by the C IRB or whether the quality of the CIRB’s management of these concerns could be monitored across a set of projects with the renewal of the relationship with the CIRB dependent on satisfaction with the process. Clearly, the IRB review process itself is a worthy subject of ongoing research. One of the few areas of consensus is that we have not yet achieved the optimal balance between efficiency and subject protection.
PM&R
REFERENCES 1. Kjaergard LL, Als-Nielsen B. Association between competing interests and authors’ conclusions: Epidemiological study of randomised clinical trials published in the BMJ [see comment]. BMJ 2002;325:249.
Vol. 2, Iss. 1, 2010
63
2. U. S. Food and Drug Administration. Medical Device Exemptions 510(k) and GMP Requirements. Available at: http://www.accessdata. fda.gov/scripts/cdrh/cfdocs/cfpcd/315.cfm. Accessed November 25, 2009.
Commentary from R. Norman Harden, MD If you accept the proposition (as I do) that medical research has value, then methodologies must evolve to assure that this work is performed in a humane and ethical way. There is inherent risk in research for the subject and for the sponsoring institution; bureaucracies have naturally evolved to minimize this risk to the subject (human or animal) and to the institution (reputational, legal). Thus, what have come to be known as institutional review boards (IRBs) emerged to ostensibly “protect” subjects from risk and abuse, but equally to protect the research institution from litigation and bad press. In the mid-1900s medical research was a local phenomenon usually confined to urban academic centers; logically local IRBs (LIRBs) coevolved. Predictably, the government regulatory agencies began to provide more and more oversight and to eventually organize guidelines under the auspices of the Food and Drug Administration (FDA). Unfortunately this FDA oversight is essentially passive. As the need for speed and large numbers of subjects became necessary (as the result of FDA’s “encouragement” to use cumbersome and archaic statistical methods in pharmaceutical trials), multicenter trials became the norm, and the number of centers required to feed the gaping (statistical) maw grew exponentially. Small nonacademic sites began to emerge to meet this need (and to exploit what became a profitable business opportunity), and central IRBs (CIRBs) coevolved to provide ethical oversight to these private forprofit sites. The use of multiple idiosyncratic LIRBs in multicenter trials proved to be a logistical nightmare and began to add unacceptable variance to the data. As LIRBs proved to be slow, cumbersome, unresponsive, and inflexible, the industry, in a predictable Darwinian fashion, gravitated toward sites using uniform and “nimble” CIRBs. These “clinical research organizations” (CRO) had several advantages, such as recruiting from “virgin” populations (as opposed to the sometimes cut-throat competition for favored diagnoses in the cities by urban academic centers, eg postherpetic neuralgia). These professional CROs proved to have multiple advantages for pharmaceutical-sponsored
research (such as the ability to develop lists of local people with a specific diagnosis that would willingly and rapidly enlist in trial after multicenter trial, ie, ”professional subjects”). Of course the validity of using these CROs and such populations is very questionable; nonetheless, the speed and responsiveness of CROs was irresistible to industry for rapidly organizing and executing large multicenter phase 3- or 4-type trials, and academic research organizations were phased out. This evolution is now essentially complete, and this business model is set. So the question is not really “are CIRBs good and should we use them” but rather “do we want to play at all, do we want this business?” If we conclude that we want these research dollars (as other funds become rare), then we must embrace CIRBs. There is no option: the attempts to hybridize the situation and use a few academic research organizations (primarily for the credibility they lend to a project) are a logistical and statistical nightmare, and companies are increasingly unwilling to do this. In fact, many companies now will not allow LIRBs at all. LIRBs do not play well with CIRBs, and legal considerations mitigate against any such collaboration. So how is one to assure that the inevitable CIRB is behaving in an ethical and proper way? There are demonstrably “good” and “bad” CIRBs (just as there are good and bad LIRBs), and currently the onus of vetting these organizations falls to the principal investigators and their legal teams. This is predictably inadequate in many cases. One answer, unfortunately, is active federal oversight, credentialing, loading CIRBs with nonpaid monitors at all levels. Another approach is to develop independent, publically run, not-for-profit CIRBs. An ideal option is to allow study design to evolve to modern statistical methods that vastly decrease the numbers of subjects required and will naturally allow companies to return to academic centers and LIRBs (and save time and money). So again, the question is not “should we” but ”how do we” assure that these necessary entities operate ethically and with integrity?
The Controversy Surrounding Central Institutional Review Boards Feature Editor Introduction: Kristi L. Kirschner, MD For this column, we are tackling an issue that many in rehabilitation may not be familiar with—namely, central institutional review boards (CIRBs). As rehabilitation research continues to grow, particularly in collaboration with pharmaceutical and device manufacturers, CIRBs will probably play an increasingly prominent role. Given the likely unfamiliarity of the issues posed by CIRBs, what follows is a brief commentary (by me, and my colleagues Teresa A. Savage, PhD, RN, and Debjani Mukherjee, PhD, of the Rehabilitation Institute of Chicago Donnelley Family Disability Ethics Program and Northwestern University Feinberg School of Medicine Medical Humanities and Bioethics Program) that lays out the issues, followed by expert commentaries and responses by Dr. John Whyte and Dr. Norman Harden. As always, we invite your comments and responses to this column!
INTRODUCTION A debate has ensued about the use of central or commercial institutional review boards (CIRBs) for research. At the heart of the matter is research integrity, public trust, and, of course, protection of human subjects. The 2 quotes below exemplify the debate: At a time when commercial interests threaten the safety of research participants and the integrity of medical research, it is remarkable that North American regulatory agencies have not seen any problem with entrusting the rights and well-being of human research participants to a lightly regulated commercial enterprise. . ..The protection of research participants is a critically important public mandate, and it merits a truly independent regulatory structure [1]. Making a distinction about the quality and merits of IRBs based on tax status is an antiquated piece of ideology reminiscent of Orwell’s Animal Farm—“for-profit bad, not-for-profit good.” The distinction obscures what we should care about and directly measure—the quality of IRB review of protocols and the monitoring of the safety of research participants [1]. In this column, we discuss the use of central or commercial (CIRBs) and summarize the historical context and debate. Throughout this feature, we welcome the reader to consider how the role of CIRBs may impact science and clinical care in the United States.
Contributors: Teresa A. Savage, PhD, RN Rehabilitation Institute of Chicago and Donnelley Family Disability Ethics Program, Chicago, IL Disclosure: nothing to disclose Debjani Mukherjee, PhD Rehabilitation Institute of Chicago, Donnelley Family Disability Ethics Program and Northwestern University Feinberg School of Medicine, Chicago, IL Disclosure: nothing to disclose
Feature Editor: K.L.K. PM&R, and Medical Humanities and Bioethics, Northwestern University Feinberg School of Medicine, Chicago, IL. Address correspondence to: K.L.K.; e-mail: [email protected] Disclosure: nothing to disclose
BACKGROUND Investigative journalists have largely led the way in elucidating research projects gone awry. When we open a newspaper or read an online account, questions are raised about how the projects were approved, who was monitoring the safety and effects, and how conflicts of interest may have impacted the process. The IRB is a committee developed to oversee research that involves human subjects, with the charge of protecting the rights and welfare of research participants. The need for IRBs was recognized in the wake of egregious examples of abuse in research (eg, Nazi physicians during World War II [2], the Tuskegee Syphilis Study [3], and the Willowbrook hepatitis vaccine study [4]), and mandated by Title 45 CFR 46 (Code of Federal Regulations) Part 46. The requirements for IRBs are set forth by law and regulated by the Office of Human Research Protections within the U.S. Department of Health and Human Services. PM&R 1934-1482/10/$36.00 Printed in U.S.A.
Guest Discussants: J.W. Moss Rehabilitation Research Institute, Albert Einstein Healthcare Network, Elkins Park, PA Disclosure: nothing to disclose R.N.H. Rehabilitation Institute of Chicago and Department of Physical Medicine and Rehabilitation, Northwestern University, Evanston, IL Disclosure: 2A, 7B, 8B Disclosure Key can be found on the Table of Contents and at www.pmrjournal.org
© 2010 by the American Academy of Physical Medicine and Rehabilitation Vol. 2, 57-63, January 2010 DOI: 10.1016/j.pmrj.2009.12.013
57
58
Kirschner et al
Historically, IRBs were set up by universities and medical institutions involved in human-subjects research. During the last 40 years, there has been the addition of for-profit or commercial IRBs, which are also subject to the same federal laws and regulations. [Although most CIRBs are commercial, there are a few exceptions, such as the National Cancer Institute IRB (http://www.ncicirb.org). For the purposes of this column, we are concentrating on for-profit, commercial (or central) IRBs and will refer to them as CIRBs.] CIRBs are being used increasingly for pharmaceutical and medical device clinical trials. From 2000 to 2006, clinical trials in the United States increased nearly 50% to 59,000 drug trials [5]. Research moved from academic medical centers to clinical research organizations or private physician offices. One of the oldest and most respected CIRBs, the Western IRB (WIRB) “reviews more than half of all trials of new drugs submitted for FDA approval. . .” and some academic medical centers are outsourcing their IRB work to CIRBs [5]. Cancer researchers have been the most vocal advocates for the use of CIRBs because they focus on the complexity of the trials, the duplication of reviews, and the ease of coordinating multisite, multi-investigator research. Although a growing number of academic and governmental institutions use CIRBs, some core ethical issues have been raised, including protection of human subjects, actual and perceived conflicts of interest, and balancing the specialized knowledge and speed of CIRBs, with the local knowledge and due process of local IRBs. Regardless of the type of IRB used, harm to research subjects has been documented, and regulations or mandates have been developed. For example, on January 15, 1994, then-President Clinton formed the Advisory Committee on Human Radiation Experiments. These experiments took place for 40⫹ years after World War II and the Cold War. The Advisory Committee on Human Radiation Experiments issued a detailed report of the multiple abuses that took place, including infringement of informed consent and the abuse of vulnerable populations [6]. In the more recent past, there are a number of questions raised about erosion of protections with current practices involving U.S. researchers who engage in clinical trials in developing countries [7]. Additionally, several high-profile deaths of research subjects have raised issues to consider. The deaths include Jesse Gelsinger at the University of Pennsylvania in 1999 [8], the death of a healthy volunteer Ellen Roche in an asthma study at Johns Hopkins in 2001 [9], and the death of Jolee Mohr, a 35-year-old woman in 2007 at the University of Chicago involved in a gene therapy study for rheumatoid arthritis [5]. Detailed reviews of these deaths revealed concerns about safety with the protocols, adherence to protocol, and the informed-consent processes. Some evidence also exists that drugs and devices that are approved by the Food and Drug Administration (FDA) have a high incidence of postmarketing safety concerns. Examples of drugs that have been withdrawn from the market after
THE CONTROVERSY SURROUNDING CENTRAL IRBS
FDA approval, sometimes years later, include rofecoxib, valdecoxib, trovafloxacin, mibefradil, terfenandine, troglitazone, cerviastatin, alosetron, and tegaserod. Other drugs have had black box warnings added after approval, such as selective serotonergic reuptake inhibitors (for suicide risk with adolescents) and rosiglitazone (for congestive heart failure and myocardial infarction risk). Some of the devices with postmarking safety concerns include drug-eluting stents, biventricular pacers, and implantable defibrillators. Since the passage of the Prescription Drug User Fee Act in 1992 (which allowed the pharmaceutical industry to pay a user fee to expedite the review of the drug), there has been marked increase in rates of postapproval safety problems in drugs that were approved in the 2 months before the approval deadline imposed by the Prescription Drug User Fee Act as compared with drugs approved in other months of the review cycle (odds ratio of 5.5 for drug withdrawals, blackbox warning odds ratio 4.4, and discontinuation of at least one dosage from 3.3—all P ⱕ .02) [10]. In the analysis of many scholars who study the problems of drug safety in the United States, there have been substantial concerns raised about the undue influence of money in the erosion of protections of human subjects, particularly when pharmaceutical companies and device makers are involved [8,11-14]. The erosions of protections can occur in a number of ways, including: (1) independence of the researcher (financial ties, ownership of data, writing manuscripts, reporting harms); and (2) independence of the review bodies (eg, the IRB, FDA).
COMPLIANCE ISSUES Federal regulations on the conduct of research apply to both local IRBs and CIRBs. In addition to protecting research participants, local IRBs also are motivated to uphold their high standards to protect the reputation of their institution. CIRBs also desire to protect their reputations, although any missteps or transgressions by the CIRB may not be public information, unlike for local IRBs. One needs only to search the Office of Human Research Protections Compliance Oversight webpage to find institutions that have been out of compliance. In the late 1990s and early 2000s, there were a number of university IRBs that had research activities suspended by the NIH Office for Protection from Research Risk (now Office for Human Research Protections). Duke University, Rush University, University of Rochester, and University of Illinois at Chicago are a few whose research was suspended and all active protocols had to be re-reviewed. There have been lawsuits against IRBs, such as in the melanoma vaccine case in Tulsa, Oklahoma, and the efalizumab (Raptiva) case, a psoriasis medication [15]. At the Medical College of Georgia, a psychiatrist stole more than $10 million of research money and allowed untrained employees to administer experimental drugs. The university IRB was unaware of the
PM&R
Vol. 2, Iss. 1, 2010
59
Table 1. Pros and cons of a CIRB and local IRB Pros CIRB
Specialized expertise in reviewers Large, well-trained IRB staff
Consistency in reviews Perception of quick approval thereby decreasing costs and speeding up scientific discoveries Local IRB
Naive community member more representative of human subject Can coordinate with other aspects of local human research protection program, eg radiation safety, conflicts of interest On-site monitoring geographically easier; investigation of adverse event reports geographically easier
Cons Lack of local context and expertise Creates more distances between principal investigator/research team and human subject protection processes and procedures Potential coercion by drug company to agree or be excluded There are no data supporting decreased costs
Lack of specialized expertise re certain conditions or drugs Fewer staff limit number of studies reviewed per meeting, thereby increasing interval between submission and approval Responsible for on-site research even when CIRB is IRB of record
study; the CIRB (WIRB) was criticized for not conducting any on-site audits, which may have uncovered the misconduct [15]. Dr. Angela Bowen, founder and former CEO of the WIRB, takes issue with the accuracy of the report by Evans et al, but she did not set the record straight in her response [16].
Local IRBs are criticized for being slow, inefficient, and too focused on unimportant minutiae; CIRBs are criticized for their “fundamental conflict of interest. They are in a clientprovider business relationship with the commercial entities whose studies they review. Because commercial IRBs gener-
Table 2. Identifying and addressing key challenges Key challenges Assurance of review quality ● Loss of safety net of redundant review ● Insufficient attention to local concerns ● Inappropriate consent forms
Suggested strategies ● ● ● ● ● ● ● ● ● ●
Sensitivity to local context Failing to understand culture, concerns, ethnic groups
●
●
●
Liability (institutional and individual)
● ● ● ● ● ●
Identify independent indicators of quality Provide performance data Consider local IRB representation on the central IRB Provide evidence of benefits Ensure that all policies and procedures are transparent Conduct reciprocal visits Consult references Consider a demonstration project Create detailed agreements Take steps to build public trust Develop means of accessing and using local information Develop strong communication Show that the institution is fulfilling its responsibilities Develop solid policies and procedures Establish a crisp agreement between the local facility and the alternative model Ensure “due diligence” in the selection of the alternative model Provide education for those responsible for managing institutional risk Develop guidance that clearly defines the responsibilities of the local IRB when using alternative review methods
Control and accountability Possible damage to the institution’s reputation
●
Clearly define and document roles
Loss of resources Insufficient staff and funding remain for compliance audits and other local responsibilities
●
Educate administrators on the IRBs continuing importance and need for funding Concentrate on other aspects of human research protection
●
●
●
Taken from Alternative Models of IRB Review Workshop Summary Report [20].
60
Kirschner et al
Table 3. Alternative models for the IRB Organization...
has its own IRB uses another institution’s IRB uses a central IRB relies on a lead IRB is part of a group that formed its own IRB uses a commercial IRB [“commercial” vs “central” not defined] uses some combination of above options
Adapted from presentation give by M. Speers, Executive Director, Association for the Accreditation of Human Research Protection Programs, November, 2006 [21].
ate their income from clients with a direct financial interest in obtaining approval, they are affected by the very problem they are expected to curtail” [1]. As the opening quotes illustrated, some scholars have fundamental disagreements about for-profit enterprises and their role in human-subjects research. Comparing local IRBs and CIRBs distinguishes the details about what type of IRBs are used from the broader issues about research integrity, public trust, and protection of human subjects. Both local IRBs and CIRBs charge external companies a fee for the cost of the IRB review. CIRBs are often in the position to process the application more quickly than local IRBs, especially if they are reviewing similar protocols, such as for oncology or psychiatric clinical drug trials. Some proponents of CIRBs have focused on the expertise that is developed by working with similar protocols. Although the local IRB may not have the experience of solely working with pharmaceutical protocols, the diversity of backgrounds and volunteer nature of the committees are intended to add a layer of protection. There is variability in the quality of local IRBs, as well as variability in quality of central IRBs. Table 1 briefly summarizes the pros and cons in selecting a CIRB or local IRB for a particular project. Institutions that have their own IRB may consider the use of a CIRB for certain classes of projects (drug clinical trials) in which the sponsor mandates the use of CIRB to be included as a study site. Other institutions that use local IRBs may consider using CIRBs for the same reason. It usually is not a decision made by the principal investigator solely.
CONFLICTS OF INTEREST Both types of IRBs can have real and perceived conflicts of interest. The direct financial incentives of CIRBs are one area. According to one source, WIRB made $20 million in profit in 2004 [15]. Dr. Bowen contests this figure but does not give the actual figure. Moreover, public trust and perceptions of conflicts can have far-reaching impact on the reputation and credibility of institutions. Both local and central IRBs have policies on conflicts of interest. For example, the WIRB website has this paragraph about conflict of interest:
THE CONTROVERSY SURROUNDING CENTRAL IRBS
What is a conflict of interest? Sometimes a principal investigator or sub investigator may have a conflict of interest. Conflicts of interest can take many forms. The most common conflict of interest arises when a doctor who provides you with medical care is also acting as the researcher overseeing your participation in the research study. In this case, there can be a conflict in his or her duty to make medical decisions in your best interest and his or her duty to follow the research protocol. Another common type of conflict of interest is financial. For example, the principal investigator might own stock in the sponsoring company, be a paid consultant for the sponsoring company, receive speaking fees from the sponsoring company, or be paid by the sponsoring company to conduct the research. The investigator may hold a patent on the drug or device being studied and stand to benefit if the drug or device is approved for general use. You should feel free to ask if either the principal investigator or research staff has a conflict of interest. If the principal investigator or any of the research staff do have a conflict of interest, discuss how they intend to manage that conflict. Only enter the study if you are satisfied with the safeguards that have been put in place to deal with the potential conflict [17]. It appears that the research participant is responsible for asking whether any research personnel have a conflict of interest. They may not be aware that they also should ask about any conflict of interest on the part of any IRB member who reviewed and approved the protocol. Local IRBs, usually located in universities, have conflict-of-interest policies and a process of review that may include an ad-hoc committee, the Principal Investigator’s dean, and the university’s Vice President/Vice Chancellor for Research. The most recent guidance document from the U.S. Department of Health and Human Services offers recommendations and points for consideration but indicates alternative approaches may suffice if all applicable statutes and regulations are satisfied [18]. Scrutiny for conflict of interest and research goes beyond the research team and IRB. Many institutions and/or senior administrators at the institution can have financial interests in companies that wish to conduct research within their institutions. The institution stands to gain if a drug or device gets to market and proves to be lucrative. In the Gelsinger case, the University of Pennsylvania owned stock in the company involved in the gene therapy [8]. The plan for managing conflict of interest should be public so that anyone considering participating in a study in which there is a conflict of interest on the part of a research team member, institutional administrator, or the institution itself can learn of the conflict and how it will be managed.
PM&R
NATIONAL MEETINGS TO EXPLORE ALTERNATIVE MODELS In 2005, The National Institutes of Health, the Office for Human Research Protections, the Association of American Medical Colleges, the American Society of Clinical Oncology, and Department of Veterans Affairs held a workshop on alternative models of IRB review. Their summary report presents issues to consider when deciding to use an alternative model, such as a central IRB: “access to expertise, the number of participating sites, the need for consistency among participating study sites, the level of risk, the sponsorship and type of study, the size of the institution’s research program, how quickly approval is needed, the type of disease involved in a medical study, the potential for conflicts of interest, liability issues, and available resources” [19]. In addition, it identifies challenges and suggested strategies for addressing the ethical concerns (Table 2). A subsequent national conference was held a year later with the purpose to “enhance the protection of human subjects of research by encouraging the use of alternative Institutional Review Boards (IRB) models under appropriate circumstances” [20]. The participants, regulators, funders, medical educators and researchers, and stakeholder groups explored in depth the issues identified in the 2005 workshop. One presenter listed the various alternative models of IRB (Table 3) [21]. Ideally, both the local IRB and CIRB will negotiate a shared responsibility that improves the protections of human subjects without adversely affecting the efficiency of the review process. The 2006 workshop executive summary concludes with some general points to consider: ●
●
●
There should be a detailed agreement regarding the sharing of regulatory liability between the CIRB and institution; Outsourcing disengages the principal investigator and research team in the protection of human subjects and may drain resources from the local IRB; and There should be clear guidance on the use of alternative models of IRBs and harmonization among federal laws and requirements.
CONCLUSION The use of CIRBs over local IRBs is relatively new, especially in drug studies apart from cancer drugs. The decision to use a CIRB over a local IRB should be a transparent decision that offers the best protection for human subjects. Trust, transparency, and clear communication between the institution, local IRB, and CIRB are critical. Identification and disclosure of any perceived or real conflicts of interest should be disclosed in a way that potential research subjects understand and have the opportunity to carefully consider. A detailed contract indicating the
Vol. 2, Iss. 1, 2010
61
areas of shared authority and responsibility of the institution, the local IRB, and the CIRB should be negotiated. Regardless of which model of IRB is selected, the institutional home of the research has an ethical responsibility to ensure that all reasonable efforts are expended to protect human subjects, which, in turn, protects the institution.
REFERENCES 1. Emanuel EJ, Lemmens T, Elliott C. Commercial IRBs have a fundamental conflict of interest. PLoS Med 2006;3:942-943. 2. The National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research. The Belmont Report. April 18, 1979. Available at: http://www.dhhs.gov/ohrp/humansubjects/ guidance/belmont.htm. Accessed December 22, 2009. 3. Jones J. Bad Blood: The History of the Tuskegee Syphilis Experiment. New York, NY: Free Press; 1993. 4. Brody BA. The Ethics of Biomedical Research: An International Perspective. New York, NY: Oxford University Press; 1998. 5. Advisory Committee on Human Radiation Experiments. Available at: http://www.gwu.edu/⬃nsarchiv/radiation. Accessed December 22, 2009. 6. Kent DM, Mwamburi DM, Bennish ML, Kupelnick B, Ioannidis JPA. Clinical trials in sub-Saharan Africa and established standards of care. JAMA 2004;292: 237-242. 7. Gelsinger P, Shamoo AE. Eight years after Jesse’s death, are human research subjects any safer? Hast Cent Rep 2008;38:25-27. 8. Press briefing: Hopkins issues internal review committee report on research volunteer’s death, 2001. Available at: www.hopkinsmedicine. org/press/2001/JULY/010716.htm. Accessed December 22, 2009. 9. Gilbert S. Trials and tribulations. Hast Cent Rep 2008;38:14-18. 10. Carpenter D, Zucker EJ, Avorn J. Drug-review deadlines and safety problems. N Engl J Med 2008;358:1354-1361. 11. Angell M. The Truth About Drug Companies: How They Deceive Us and What To Do About It. New York, NY: Random House; 2004. 12. Bass A. Side Effects: A Prosecutor, a Whistleblower, and a Bestselling Antidepressant on Trial. Chapel Hill, NC: Algonquin Books; 2008. 13. Brody H. Hooked: Ethics, the Medical Profession, and the Pharmaceutical Industry. Lanham, MD: Rowman & Littlefield Publishers, Inc.; 2007. 14. Elliott C. The soul of a new machine: Bioethicists in the bureaucracy. Camb Q Healthc Ethics 2005;14: 379-384. 15. Evans D, Smith M, Willen L. Big Pharma’s Shameful Secret. Bloomberg Markets. November 2, 2005. Available at http://www.ahrp.org/ infomail/05/11/03.php. Accessed July 30, 2009 16. Bowen A. Research ethics boards: Error and misconception. PLoS Med 2006;3:e457-e458. 17. WIRB website Info for research subject. Available at: http://www. wirb.com/content/research_subjects.aspx. Accessed August 18, 2009. 18. U.S. Department of Health and Human Services. Final Guidance Document: Financial Relationships and Interests in Research Involving Human Subjects: Guidance for Human Subject Protection; 2004. Available at: http://www.dhhs.gov/ohrp/humansubjects/finreltn/fguid.pdf. Accessed November 12, 2009. 19. National Institute of Health, Office for Human Research Protections, Association of American Medical College, & American Society of Clinical Oncology. Alternative models of IRB review: Workshop Summary Report, 2005. Available at: http://www.hhs.gov/ohrp/sachrp/ documents/AltModIRB.pdf . Accessed July 30, 2009. 20. National conference on alternative IRB models: Optimizing human subject protection. November 19-21, 2006. Available at: http://www.aamc. org/research/irbreview/irbconf06rpt.pdf. Accessed November 11, 2009. 21. Speers M. Shared authority and responsibilities. 2006 National Conference on Alternative IRB Models: Optimizing Human Subject Protection, November 19-21, 2006, Marriott Wardman Park Hotel, Washington, DC.
62
Kirschner et al
THE CONTROVERSY SURROUNDING CENTRAL IRBS
Commentary from John Whyte, MD, PhD In recent years, there have been increasing complaints that our institutional research board (IRB) system fails to balance protection of human subjects with preservation of the scientific enterprise. Previous ethical lapses have led to ever more voluminous and complex review processes that diminish research productivity by delaying study approval and requiring diversion of staff to regulatory rather than scientific tasks. Moreover, as consent documents become longer and more complex, it’s not at all clear that they actually make the consent process more “informed,” particularly for subjects with mental health or cognitive limitations. Clearly a simpler and more streamlined system that still maintains rigorous ethical oversight would be desirable for both research participants and investigators. Central IRBs (CIRBs) have the potential to streamline and standardize the process because they may have staff members that are highly experienced in the review process, have sufficient review volume in specialized research areas to develop expertise, and can substitute a single review for multiple independent reviews of multicenter studies most typical of research sponsored by industry. Moreover, they can eliminate the need, common in multicenter projects, to have all IRBs re-review changes that are requested by each individual IRB. Weighing against these advantages, as noted in the accompanying commentary, are the potential for conflict of interest and the lack of local institutional knowledge. These themes invite further consideration, both in general, and as they apply to areas of rehabilitation research with which I am most familiar. There is no doubt that financial conflict of interest exerts biases in the research enterprise even when there is no conscious attempt to distort results [1]. Thus, the relationship between central IRBs and their sponsors is of concern. But it is worth noting that conflict of interest comes in many forms and is surely not absent from locally reviewed projects. Investigators may themselves have relationships with sponsoring entities as speakers or as expert consultants. Even in the absence of a financial relationship, an investigator who achieves notoriety through involvement with a new treatment technique may be reluctant to find such a technique ineffective. In rehabilitation research, there are many emerging treatments that can be implemented outside of a traditional Food and Drug Administration approval process. For example, many medical technologies such as exercise machines, and orthotic and prosthetic devices, need only show that they are similar to existing products and pose no clear new risks to be marketed [2]. This produces a situation in which investigators may be studying products that are already being heavily promoted by the manufacturer. In some cases, the investigator may gain access to such equipment at low cost in part because of his/her role in studying it further. In other cases, the fact that an institution has such cutting-
edge technology is a marketing advantage that they don’t wish to undercut. And because there is no consensus standard of care for many rehabilitation problems, one cannot use a deviation from this standard as a clue that research may be taking place. Although ethical viewpoints on a given study may happen to vary from institution to institution, there is no inherent reason why “research ethics” should be a local enterprise. And one might wish that all of the issues relevant to review were generic and easily distilled into a written application. But for better or for worse, local IRBs get to know individual investigators, come to understand how closely they are supervised by their departmental administration, how seriously local administrative officials take the issue of conflict of interest, etc. The IRB may be aware that an investigator is overseeing many projects with very small funded effort or may raise concerns about a current project on the basis of an event in a previous project involving that investigator or department. In addition, the ethics of a study often are intertwined with the processes of the local clinical system. For example, review of a study with aphasic subjects may be considered quite differently if the IRB knows that the institution has a large cohort of patients with aphasia, has highly trained recruiters who know how to assess language comprehension, and has experience in developing appropriate graphic materials to support the informed consent process. Given the apparent advantages as well as risks of a CIRB process, a combined review, as alluded to in the accompanying article, seems most desirable if it can be efficiently coordinated. One form this might take is the identification of issues of concern by local IRBs, but the delegation of solutions to those issues to a CIRB. In other words, one or more IRBs could identify a consent form as too complex without each suggesting specific and different edits. Individual IRBs could identify concerns about methodology and safety, but the CIRB could manage the response to those concerns. This would allow simultaneous review by many local IRBs, forwarding of a full set of concerns to the CIRB for resolution, and return of a modified protocol and consent forms to all IRBs. The remaining question is whether each IRB would need the authority to approve or reject each project as modified by the C IRB or whether the quality of the CIRB’s management of these concerns could be monitored across a set of projects with the renewal of the relationship with the CIRB dependent on satisfaction with the process. Clearly, the IRB review process itself is a worthy subject of ongoing research. One of the few areas of consensus is that we have not yet achieved the optimal balance between efficiency and subject protection.
PM&R
REFERENCES 1. Kjaergard LL, Als-Nielsen B. Association between competing interests and authors’ conclusions: Epidemiological study of randomised clinical trials published in the BMJ [see comment]. BMJ 2002;325:249.
Vol. 2, Iss. 1, 2010
63
2. U. S. Food and Drug Administration. Medical Device Exemptions 510(k) and GMP Requirements. Available at: http://www.accessdata. fda.gov/scripts/cdrh/cfdocs/cfpcd/315.cfm. Accessed November 25, 2009.
Commentary from R. Norman Harden, MD If you accept the proposition (as I do) that medical research has value, then methodologies must evolve to assure that this work is performed in a humane and ethical way. There is inherent risk in research for the subject and for the sponsoring institution; bureaucracies have naturally evolved to minimize this risk to the subject (human or animal) and to the institution (reputational, legal). Thus, what have come to be known as institutional review boards (IRBs) emerged to ostensibly “protect” subjects from risk and abuse, but equally to protect the research institution from litigation and bad press. In the mid-1900s medical research was a local phenomenon usually confined to urban academic centers; logically local IRBs (LIRBs) coevolved. Predictably, the government regulatory agencies began to provide more and more oversight and to eventually organize guidelines under the auspices of the Food and Drug Administration (FDA). Unfortunately this FDA oversight is essentially passive. As the need for speed and large numbers of subjects became necessary (as the result of FDA’s “encouragement” to use cumbersome and archaic statistical methods in pharmaceutical trials), multicenter trials became the norm, and the number of centers required to feed the gaping (statistical) maw grew exponentially. Small nonacademic sites began to emerge to meet this need (and to exploit what became a profitable business opportunity), and central IRBs (CIRBs) coevolved to provide ethical oversight to these private forprofit sites. The use of multiple idiosyncratic LIRBs in multicenter trials proved to be a logistical nightmare and began to add unacceptable variance to the data. As LIRBs proved to be slow, cumbersome, unresponsive, and inflexible, the industry, in a predictable Darwinian fashion, gravitated toward sites using uniform and “nimble” CIRBs. These “clinical research organizations” (CRO) had several advantages, such as recruiting from “virgin” populations (as opposed to the sometimes cut-throat competition for favored diagnoses in the cities by urban academic centers, eg postherpetic neuralgia). These professional CROs proved to have multiple advantages for pharmaceutical-sponsored
research (such as the ability to develop lists of local people with a specific diagnosis that would willingly and rapidly enlist in trial after multicenter trial, ie, ”professional subjects”). Of course the validity of using these CROs and such populations is very questionable; nonetheless, the speed and responsiveness of CROs was irresistible to industry for rapidly organizing and executing large multicenter phase 3- or 4-type trials, and academic research organizations were phased out. This evolution is now essentially complete, and this business model is set. So the question is not really “are CIRBs good and should we use them” but rather “do we want to play at all, do we want this business?” If we conclude that we want these research dollars (as other funds become rare), then we must embrace CIRBs. There is no option: the attempts to hybridize the situation and use a few academic research organizations (primarily for the credibility they lend to a project) are a logistical and statistical nightmare, and companies are increasingly unwilling to do this. In fact, many companies now will not allow LIRBs at all. LIRBs do not play well with CIRBs, and legal considerations mitigate against any such collaboration. So how is one to assure that the inevitable CIRB is behaving in an ethical and proper way? There are demonstrably “good” and “bad” CIRBs (just as there are good and bad LIRBs), and currently the onus of vetting these organizations falls to the principal investigators and their legal teams. This is predictably inadequate in many cases. One answer, unfortunately, is active federal oversight, credentialing, loading CIRBs with nonpaid monitors at all levels. Another approach is to develop independent, publically run, not-for-profit CIRBs. An ideal option is to allow study design to evolve to modern statistical methods that vastly decrease the numbers of subjects required and will naturally allow companies to return to academic centers and LIRBs (and save time and money). So again, the question is not “should we” but ”how do we” assure that these necessary entities operate ethically and with integrity?