Institutional review board

International ethics review committees

Numerous other countries have equivalent regulations or guidelines governing human subject studies and the ethics committees that oversee them. However, the organizational responsibilities and the scope of the oversight purview can differ substantially from one nation to another, especially in the domain of non-medical research. The United States Department of Health and Human Services maintains a comprehensive compilation of regulations and guidelines in other countries, as well as related standards from a number of international and regional organizations.

Naming and composition

Although "IRB" is a generic term used in the United States by the FDA and HHS, each institution that establishes such a board may use whatever name it chooses. Many simply capitalize the term "Institutional Review Board" as the proper name of their instance. Regardless of the name chosen, the IRB is subject to the US FDA's IRB regulations when studies of FDA-regulated products are reviewed and approved. At one time, such a committee was named the "Committee for the Protection of Human Subjects".

Originally, IRBs were simply committees at academic institutions and medical facilities to monitor research studies involving human participants, primarily to minimize or avoid ethical problems. Today, some of these reviews are conducted by for-profit organizations known as independent or commercial IRBs. Anyone, including private individuals, can pay a commercial IRB for review. The responsibilities of these IRBs are identical to those based at academic or medical institutions, and within the US, they are governed by the same US federal regulations.

While its composition varies, it often includes a balance of academia and non-academia members. This serves to provide a greater scope of understanding which helps ensure ethics in research. In the US, regulations set out the board's membership and composition requirements, with provisions for diversity in experience, expertise, and institutional affiliation. For example, the minimum number of members is five, at least one scientist, and at least one non-scientist. The guidance strongly suggests that the IRB contain both men and women, but there is no regulatory requirement for gender balance in the IRB's membership. The full requirements are set out in 21 CFR 56.107.

As IRBs are normally staffed with employees, who have to be paid, there are costs to operating them. In 2001, the cost of operating an IRB typically ranged from about $75,000 to $770,000 ($136,000 to $1,400,000, after accounting for inflation) per year, depending on the volume of research reviewed.

Convened and expedited reviews

Unless a research proposal is determined to be exempt (see below), the IRB undertakes its work either in a convened meeting (a "full" review) or by using an expedited review procedure. When a full review is required, a majority of the IRB members must be present at the meeting, at least one of whom has primary concern for the nonscientific aspects of the research. The research can be approved if a majority of those present are in favor.

An expedited review may be carried out if the research involves no more than minimal risk to subjects, or where minor changes are being made to previously approved research. The regulations provide a list of research categories that may be reviewed in this manner. An expedited review is carried out by the IRB chair, or by their designee(s) from the board membership. In the US, research activity cannot be disapproved by expedited review.

Pharmaceutical trials and good clinical practice

The International Conference on Harmonisation sets out guidelines for registration of pharmaceuticals in multiple countries. It defines Good Clinical Practice (GCP), which is an agreed quality standard that governments can transpose into regulations for clinical trials involving human subjects.

Here is a summary of several key regulatory guidelines for oversight of clinical trials:

• Safeguard the rights, safety, and well-being of all trial subjects. Special attention should be paid to trials that may include vulnerable subjects, such as pregnant women, children, prisoners, the elderly, or persons with diminished comprehension.
• Obtain trial protocol(s)/amendment(s), written Informed Consent Form(s) (ICFs) and consent form updates the investigator proposes for use in the trial, subject recruitment procedures (e.g., advertisements), written information to be provided to subjects, investigator's brochure, available safety information, information about payments and compensation available to subjects, the investigator's current curriculum vitae and/or other documentation evidencing qualifications, and any other documents the IRB may need to fulfill its responsibilities.
• Review both the amount and method of payment to subjects to assure neither presents problems of coercion or undue influence on the trial subjects. Payments to a subject should be prorated and not wholly contingent on completion of the trial by the subject. Information regarding payment to subjects, including the methods, amounts, and schedule of payment to trial subjects, should be set forth in the written informed consent form and any other written information to be provided to subjects. The way payment will be prorated should be specified.
• Ensure that the proposed trial is reviewed within a reasonable time and document opinions and decisions in writing, clearly identifying the trial, the documents reviewed and recorded dates for approvals, required modifications prior to approval, disapproval of a proposed trial, or termination/suspension of any prior approval. Continuing review of ongoing trials is required at intervals appropriate to the degree of risk to human subjects, but at least once per year.

The reviewers may also request that more information be given to subjects when, in their judgment, the additional information would add meaningfully to the protection of the rights, safety and/or well-being of the subjects. When a non-therapeutic trial is to be carried out with the consent of the subject's legally acceptable representative, reviewers should determine that the proposed protocol and/or other document(s) adequately address relevant ethical concerns and meets applicable regulatory requirements for such trials. Where the protocol indicates prior consent of the trial subject or the subject's legally acceptable representative is not possible, the review should determine that the proposed protocol and/or other document(s) adequately addresses relevant ethical concerns and meets applicable regulatory requirements for such trials (i.e., in emergency situations).

Adapting IRB review to social science

While the Belmont Principles and U.S. federal regulations were formulated with biomedical and social-behavioral research in mind, the enforcement of the regulations, the examples used in typical presentations regarding the history of the regulatory requirements, and the extensiveness of written guidance have been predominantly focused on biomedical research.

Numerous complaints by investigators about the fit between the federal regulations and its IRB review requirements as they relate to social science research have been received. Broad complaints range from the legitimacy of IRB review, the applicability of the concepts of risk as it pertains to social science (e.g., possibly unneeded, over-burdensome requirements), and the requirements for the documentation of participants' informed consent (i.e., consent forms). Researchers have tried to determine under what instances participants are more likely to read informed consent forms, and ways to improve their efficacy in the social sciences. IRB members have been criticized for assuming that surveys about past trauma has a re-traumatizing effect. Social scientists have criticized biomedical IRBs for failing to adequately understand their research methods (such as ethnography). For this reason, some large research institutions have set up multiple specialized IRBs, and may have one committee that exclusively oversees social science research.

In 2003, the US Office for Human Research Protections (OHRP), in conjunction with the Oral History Association and American Historical Association, issued a formal statement that taking oral histories, unstructured interviews (as if for a piece of journalism), collecting anecdotes, and similar free speech activities often do not constitute "human subject research" as defined in the regulations and were never intended to be covered by clinical research rules. In 2017, the federal government announced that effective January 2018, the regulations would no longer cover "Scholarly and journalistic activities (e.g., oral history, journalism, biography, literary criticism, legal research, and historical scholarship), including the collection and use of information that focus directly on the specific individuals about whom the information is collected."

Other US federal agencies supporting social science have attempted to provide guidance in this area, especially the National Science Foundation. In general, the NSF guidelines assure IRBs that the regulations have some flexibility and rely on the common sense of the IRB to focus on limiting harm, maximizing informed consent, and limiting bureaucratic limitations of valid research.

Adapting IRB review to big data research

Aspects of big data research pose formidable challenges for research ethics and thus show potential for wider applicability of formal review processes. One theme is data breaches, but another with high difficulty is potentially dangerous predictive analytics with unintended consequences, via false-positives or new ways to invade privacy. A 2016 article on the hope to expand ethics reviews of such research included an example of a data breach in which a big data researcher leaked 70,000 OkCupid profiles with usernames and sexual orientation data. It also gave an example of potential privacy invasion and government repression in which machine learning was used to build automated gaydar, labeling strangers as "probably gay" based on their facial photographs. Analogies with phrenology and Nazis identifying people as "probably part-Jewish" based on facial features have been made to show what can go wrong with research whose authors may have failed to adequately think through the risks of harm.

Relationship with citizen science

Generally speaking, citizen science, whether conducted by a single private individual or a group of individuals, is not required to follow the IRB process. This is true even if some of the individuals involved are professional researchers or are also employed at institutions that normally review all research conducted by the institution.

However, many academic journals require proof of IRB approval for all human-subject research, even when it is not legally required, which means that citizen scientists may be unable to publish scientific papers describing their findings. Citizen scientists who expect to need IRB approval for publication or to comply with the terms of a research grant can pay a commercial IRB company. In the US, a standard initial review often costs a few thousand dollars; a review to determine that the project is less expensive.

The IRB-based approach to ethics assumes that human-subject research is conducted by an institution employing researchers, and that the institution and researchers have far more power and knowledge than the participants. The researchers and the participants are seen as distinct groups, and the concern is to prevent the researchers from exploiting the participants as a means to an end. This leads to IRBs issuing requirements such as having researchers explain the research project and obtain informed consent. However, this model does not always fit citizen science projects, especially when the participants are themselves the experts and researchers. In such cases, a requirement to explain the project means participants would absurdly be informing themselves of their own plans. In a citizen science project, the boundaries between the researcher and the participant are blurred. Similarly, many institutionally-driven research programs are limiting or prohibiting the return of results to individuals, especially for genetic or medical studies, for fear that some participants could be harmed if they misunderstand the results. In this restrictive model, the participant never finds out their test results, or they can only find out their test results if the researchers carefully explain the results to them. But in a citizen science project, learning the results is a highly desired reason for participating, and, since the researchers are themselves participants, it would be impossible to prevent them from obtaining the results.

Criticism

Conflicts of interest

While the IRB approval and oversight process is designed to protect the rights and welfare of the research subjects, it has been the subject of criticism, by bioethicists and others, for conflicts of interest resulting in lax oversight. In 2005, the for-profit Western Institutional Review Board claimed to conduct the majority of reviews for new drug submissions to the FDA. In a 2006 study of 575 IRB members at university medical centers, over one-third reported industry financial ties, and over one-third admitted they "rarely or never" disclosed conflicts of interest to other board members.

In 2009 the US Government Accountability Office (GAO) set up a series of undercover tests to determine whether the IRB system was vulnerable to unethical manipulation. In one test, a fake product "Adhesiabloc" was submitted to a number of IRBs for approval for human tests. The product, company, and CVs of the supposed researchers were all fictitious and documents were forged by the GAO. The product was deliberately formulated to match some "significant risk" criteria of the FDA and was described by GAO as a "gel that would be poured into a patient's stomach after surgery to collect the bits and pieces left over from an operation." Despite this, one IRB approved the device for human testing. Other IRBs to whom the device was submitted rejected the application, one of them saying it was "the riskiest thing I've ever seen on this board". However, none of the IRBs approached detected that the company and product were fake. The GAO also set up a fake IRB and obtained requests for approval from companies. They succeeded in getting assurance approval from the HHS for their fake IRB. At the time, the US HHS has only three staff to deal with 300 IRB registrations and 300 assurance applications per month. HHS stated that it would not be worthwhile to carry out additional evaluation even if they had the staff to do it.

Mortality Costs From Delays

In 2026, an article in the Journal of Empirical Research on Human Research Ethics. found current ethics approval systems for medical research in the US, paradoxically result in preventable deaths by delaying patient access to experimental therapies. The article reviewed peer-reviewed evidence that demonstrates that ethics delays impose mortality costs that vastly exceed the harms they prevent, particularly for terminal illnesses. Subsequent responses by Steel (largely in agreement with the criticism) and to Tsan (defending the status quo) were evaluated by Pearce concluding "...IRB delays result in more human mortality than they protect. The current system is not logically defensible, and it is long past time that we fixed it."