What if IRBs do not make decisions on the basis of[horizontal ellipsis]evidence? The answer at present is nothing will happen. That is because IRBs as currently designed are autocratic, are unaccountable, and many act in idiosyncratic ways that are impervious to evidence or sensibleness.1^(p517)

Obtaining institutional review board (IRB) approval is required by federal regulations for conducting research involving human subjects. Charged with the responsibility of protecting the rights and welfare of research subjects, IRBs determine level of participation risk, assess adequacy of informed consent, ensure adherence to research protocols, and monitor for adverse outcomes. Although we recognize and support the need for IRB approval of research, we believe that a better balance is needed to reduce the unintended consequences to investigators of burdens associated with IRB review and approval that unduly affect the research process.

Institutional review boards have been described as unpredictable, inconsistent, inefficient, oblivious to their original intention, and dysfunctional.2-4 Although oversight of human subject research by parties other than the investigators is necessary,5-7 many investigators who have submitted a protocol for review to their local IRB can probably empathize with at least some of these negative descriptors of IRBs.

Despite the best intentions of many IRBs, the mounting number and scope of regulations for conducting research involving human subjects and the steadily changing nature of these regulations has made the review process unwieldy, time-consuming, mired in paperwork, contradictory, and, in many cases, ineffective. Concerns about delays, variations in review, and loss of sight of their original goals have been noted about IRBs in a variety of settings.8,9

With regard to delays in IRB review, many investigators have noted an increasing challenge to doing research associated with protracted time to IRB approval when initial submissions are lost in the system or do not receive timely IRB review. At many institutions, IRBs meet only monthly and have strict timelines for reviewing proposals submitted that can result in investigators waiting 2 or more cycles for their application to be reviewed. Some institutional IRBs meet less frequently during the summeror holidays, resulting in extremely long delays in review. This problem is even more acute outside academic institutions. In local community hospitals or similar local nonacademic entities that require IRB review at their level, delays can be extreme because of the limited number of IRB meetings and lack of flexibility in dealing with pending IRB issues outside usual meeting times. Other sources of delay include requirements by some IRBs for a full scientific review of the protocol before the IRB review, extremely slow reviewer responses times, inability of the IRB to form a quorum even at a scheduled meeting, and full meeting agendas. It is not uncommon for delays of 3 to 9 months for these and other unexplained reasons. Institutional review board delays in reviewing protocols are 1 of the most common concerns of investigators, some even documenting high financial costs associated with delays in starting studies owing to IRB delays in review.9

Studies of IRB reviews of identical protocols from multicenter trials have revealed striking inconsistencies in the way IRB members review protocols. Variation in review by different IRBs is such a problem in the ethical review process that it has become a source of burden to investigators. In a review of 52 studies of the costs of IRB review, wide variations in the process and outcome of IRB review were noted.9 Several studies in which different IRBs reviewed identical protocols received startling differences in responses. Commonly, investigators were asked for different revisions, different issues were noted, and competing revisions were requested.

Investigators have expressed major concerns about the generalizability of research findings given the introduction of selection bias based on some IRB review processes.10 Requirements for lengthy and complex consent forms that include passages irrelevant to the research being conducted may differentially discourage elderly individuals and those with lower educational attainment, low reading comprehension, or lower socioeconomic status from being in research. The result is lack of diversity in the recruited sample and, hence, lack of applicability of the results of the study to the excluded populations. For multicenter research in which local IRB approval is required in addition to that from a central IRB, there are negative consequences of requiring this double review.10 When local IRB review was required, there were lower rates of participation by sites than when it was not, the effort required to go through the process of local IRB review was substantial even when the study had already received approval at the central site, and there were major delays in data collection. Moreover, the requirement for local IRB approval resulted in the enrollment of fewer minorities.10

Another consequence of burgeoning documentation requirements associated with IRB review is the length and complexity of consent forms required for no more than minimal risk studies.11 These documents are written in a format that is dictated by legal concerns and are thought by many to be designed to protect institutions rather than the research participants.4 These documents, which can be as long as 15 pages, are often rejected by potential subjects who refuse to read them because they are perceived to be intimidating and overwhelming. Lengthy consent forms seen as daunting and off-putting often result in the retraction of initial interest by potential subjects and, ultimately, in their refusal to participate in the research study. This lack of interest directly affects subject recruitment and the costs of conducting the study. Many IRBs require common language in every consent form regardless of the potential risks of the research. Sometimes, this language is irrelevant to the project and confusing or threatening to potential participants.11 As 1 investigator commented in a qualitative study, "Informed consent has clearly transmogrified from protection of subjects to protection of the institution. We are required to use language that is incomprehensible to most research subjects. This absurdity has been compounded by HIPAA[horizontal ellipsis]. In short, the informed consent process has become a travesty."11^(p70)

Many IRBs now have assumed responsibility for overseeing implementation of the Health Insurance Portability and Accountability Act (HIPAA) Privacy Rule. There is evidence that IRB administration of the HIPAA rule has adversely affected the conduct of research.12 In a study of 1527 epidemiologists throughout the country, fewer than 25% felt that the rule had promoted participants' privacy. Responding epidemiologists reported that the number of IRB protocols for which the HIPAA rule had an adverse impact on research participants was substantially higher than those for which the impact was positive. Most respondents indicated that the HIPAA rule had made research more difficult (score of 4 in a 5-point scale, with 5 indicating a "great deal of added cost and time to study completion").12

In an editorial in Science, scholars from the University of Illinois decried the "mission creep" that has beset IRBs that have been hit with steadily increasing requirements for document and paperwork involving oversight of issues that have little to do with protecting human subjects involved in risky research.4 They contend that excessive requirements for documentation and monitoring of even minimal risk research may actually decrease the protection of people involved in research. Paperwork overload, inclusion of untrained members of IRBs, requirements to check consent forms for standard language that may be inappropriate for the study involved, and concerns for protecting the institution have, in some cases, overshadowed the protection of human subjects. Pushed increasingly by federal agencies for even more documentation and no flexibility in interpreting regulations, IRBs have become driven by concerns of being cited for noncompliance.3,4

In a recent literature review, Silberman and Kahn9 concluded that there is sufficient evidence to document burden and costs associated with IRB review. Reform is needed to strike a better balance between IRB oversight and its costs and burdens of conducting human subject research.

To this end, we offer the following recommendations to streamline the IRB review and approval process: (1) establish central IRBs for multicenter studies; (2) eliminate lengthy and complex consent forms; (3) update outdated regulations; (4) eliminate excessive paperwork and documentation requirements; and (5) improve the timeliness of IRB review. At the federal level, both the US Department of Health and Human Services13 and the Office for Human Research Protections14 support the use of central IRBs to improve the efficiency of conducting multicenter clinical trials. Most recently, in January 2013, the Clinical Trials Transformation Initiative concluded that central IRBs would most likely improve quality and efficiency of multicenter clinical trials and be of benefit to clinical research.15 Taking the lead in this area, the National Cancer Institute established a central IRB in 2001 to review phase III oncology trials.16 A national level review was conducted on all National Cancer Institute-sponsored trials before distribution to local IRBs, facilitating the expeditious review at the local IRB level. The use of central IRBs reduces time for protocol review and approval, decreases IRB staff effort, and reduces net cost.17 Furthermore, we recommend eliminating mandatory "boiler plate" language within consent forms and replacing it with simple, yet specific, wording that describes the research, as well as the true and reasonable potential risks and benefits to human subjects. This will shorten the number of pages of the consent form and also help to increase their readability. We believe that this change will also facilitate overall recruitment, especially of diverse and aging populations, which often refuse to consider participating after seeing the lengthy and confusing consent form.

Lastly, regulations in place to regulate human subject research are in desperate need of updating. The Common Rule, in place since 1991, established baseline ethical standards for human subject research in the United States. Recognizing the need for updating, in 2011, the US Department of Health and Human Services asked for public input to improve the rules and regulations. Areas of input addressed ethics, safety, oversight of human subject research, and updating of current forms and processes.18 Reducing the voluminous paperwork and documentation currently required for human subject research is another burden that must be addressed in the revisions being considered. We applaud the efforts by the federal government to update the current regulations and agree that regulations must reflect the current research environment. If enacted, this would be the biggest change since 1981 to the US IRB regulations.

We recognize and value the important work of IRBs to protect the rights and welfare of human research subjects. However, change is needed to streamline the process and reduce burden at all levels.

Cynthia Arslanian-Engoren

Debra K. Moser

REFERENCES