What an IRB Does and Who Sits on It

Part of Human Subjects Protection for Clinical Research Staff: Who Protects Participants.

About 7 minutes. Free, in your language, no account needed.

Why this matters

Almost everything you do with participants happens inside a boundary drawn by an Institutional Review Board (IRB), called an IEC or ethics committee in many countries (ICH E6(R3) says IRB/IEC). Knowing what it can do and how it decides tells you when to call it and what you cannot do without it.

ICH E6(R3) is the international baseline in this course. The US federal regulations (HHS 45 CFR 46, FDA 21 CFR 50 and 56, HIPAA 45 CFR 164) are taught as the detailed worked example, and passages that name them are US law. Outside the United States, follow your national law and ethics committee requirements. State or local rules may add requirements.

What the IRB is for

ICH E6(R3) states the purpose plainly: the IRB/IEC safeguards the rights, safety and well-being of all trial participants, with appropriate consideration for trials that intend to include vulnerable participants.

Under 45 CFR 46.109(a) and 21 CFR 56.109(a), an IRB has authority to approve research, to require modifications in it to secure approval, or to disapprove it. Under 46.113 and 56.113, an IRB can suspend or terminate approval of research that is not being conducted in accordance with its requirements or that has been associated with unexpected serious harm to subjects. It must state its reasons and report promptly.

What the IRB reviews

Where applicable, ICH E6(R3) has the IRB/IEC review the protocol and amendments, consent and assent materials and the consent process, the investigator's brochure, other participant information, advertisements and the recruitment process, compensation plans, safety updates and investigator qualifications.

The IRB also has authority to observe, or have a third party observe, the consent process and the research.

Who sits on it

The membership rules are designed to bring different viewpoints into one room.

In the United States, under 45 CFR 46.107 and 21 CFR 56.107:

Outside experts may assist but may not vote.

ICH E6(R3) says the IRB/IEC should consist of members who collectively can evaluate the science, medical aspects and ethics of the trial, with at least five members recommended, at least one whose primary area of interest is not in medical sciences and at least one who is independent of the institution or site. Only members independent of the investigator and sponsor vote. The investigator and site staff may provide information but do not take part in the decision or vote.

If you are asked to answer questions at a meeting, you answer them and leave the decision to the committee.

How it decides

Research that is not reviewed by an expedited procedure is reviewed at convened meetings with a majority of members present, including at least one member whose primary concerns are nonscientific. Approval requires a majority of those present.

Under its written procedures, the IRB ensures prompt reporting of changes in research, and that changes are not started without IRB approval, except where necessary to eliminate apparent immediate hazards to subjects.

In the United States, research that requires review by the convened IRB gets continuing review at intervals appropriate to the degree of risk, not less than once per year (46.109(e)). The 2018 Common Rule lists exceptions in 46.109(f), covered in Module 3. The FDA rule keeps the not-less-than-annual standard for the research it covers (56.109(f)). Under ICH E6(R3), the IRB/IEC conducts continuing review at intervals appropriate to the degree of risk.

The approval criteria in one place

In the United States, the Common Rule lists what the IRB must determine before it approves research (46.111(a)(1) to (7)); FDA's IRB rule has a parallel list (56.111). The IRB must find that:

  1. risks to subjects are minimized
  2. risks are reasonable in relation to anticipated benefits, if any, and the importance of the expected knowledge
  3. selection of subjects is equitable
  4. informed consent will be sought as required
  5. informed consent will be documented, or documentation appropriately waived
  6. when appropriate, the plan makes adequate provision for monitoring the data to ensure safety
  7. when appropriate, there are adequate provisions to protect privacy and confidentiality

When some or all subjects are likely to be vulnerable to coercion or undue influence, additional safeguards must be included (46.111(b)). For research HHS conducts or supports, Subparts B, C and D add requirements for pregnant women, fetuses and neonates, prisoners and children (Module 5). Each item is taught elsewhere in this course. Outside the United States, follow your national law and your ethics committee's requirements; the ICH E6(R3) list of what the IRB/IEC reviews, above, is the international baseline.

Limits on other officials

Institutional officials may review research that an IRB has approved, and they may add their own review. They may not approve research that the IRB has not approved (46.112; 56.112). An administrator cannot approve a study the IRB has not approved.

Timing

No participant should be enrolled before the IRB/IEC has issued its documented approval or favorable opinion.

In practice

A coordinator is asked by a PI to start prescreening conversations on Friday because the sponsor wants early numbers. The IRB meets the following week. She checks the file: no documented approval yet. She tells the PI that nobody can be enrolled before approval and offers to use the days to prepare the consent materials. The PI agrees.

What this looks like when it goes wrong

A site's PI, who is also on the hospital's research committee, is asked to attend the IRB meeting on her own study. She answers questions, then stays in the room and takes part in the vote. The study is approved. Later, a reviewer notices that someone with a conflicting interest voted on the project.

Providing the information the IRB requests is allowed. Taking part in the rest of the review, including the vote, is not. A study approved with a conflicted vote creates a problem for the site and for the participants who enrolled under that approval.

Key takeaways