An Institutional Review Board, or IRB, is an independent committee that reviews and approves research involving human participants before a clinical trial can begin. Its core job is protecting the rights, safety, and well-being of the people who volunteer for studies. Under federal law, an IRB has the authority to approve a study, require changes to it, or reject it entirely.
Every clinical trial regulated by the FDA must go through IRB review. This applies to trials testing new drugs, medical devices, and other therapies. The requirement exists because participants in research face risks that go beyond routine medical care, and an outside group needs to evaluate whether those risks are justified.
The Ethical Principles Behind IRB Review
The modern IRB system traces back to a 1979 document called the Belmont Report, which established three ethical principles for human research. The first, respect for persons, holds that individuals should be treated as autonomous decision-makers and that people with diminished autonomy (such as children or those with cognitive impairments) deserve extra protection. The second, beneficence, requires that researchers actively work to minimize harm and maximize potential benefits. The third, justice, asks who bears the burdens of research and who receives the benefits, ensuring that no single group is unfairly targeted or excluded.
These three principles shape every decision an IRB makes, from evaluating study design to reviewing consent forms.
Who Sits on an IRB
Federal regulations require each IRB to have at least five members with diverse backgrounds. The board must include at least one person whose expertise is scientific and at least one whose background is nonscientific, such as an ethicist, lawyer, or community advocate. At least one member must have no affiliation with the institution running the research, providing an outsider’s perspective.
When a board regularly reviews research involving vulnerable populations (children, prisoners, people with impaired decision-making, or economically disadvantaged groups), it should include members with specific experience working with those communities. This mix of perspectives is intentional: it prevents the review from becoming purely a technical exercise and keeps the focus on what participation actually looks like for a real person.
What the IRB Evaluates
Before approving a clinical trial, an IRB must confirm several things. Risks to participants are minimized by using sound research methods and, whenever possible, procedures the participants would already be undergoing for diagnosis or treatment. The remaining risks must be reasonable compared to the potential benefits to participants and the value of the knowledge the study could produce. The IRB also checks that participant selection is fair, paying close attention to whether vulnerable groups might be subject to coercion or undue pressure to enroll.
Importantly, the IRB focuses only on risks and benefits that come directly from the research itself. It does not weigh broader societal effects, like how the findings might eventually influence public policy. The question is narrower: is this study designed so that the people who volunteer are treated ethically?
Informed Consent Oversight
One of the IRB’s most visible responsibilities is reviewing the informed consent process. Before anyone enrolls in a trial, they must receive a clear explanation of what the study involves. Federal regulations list specific elements that every consent form must cover:
- What the study is about: its purpose, how long participation lasts, which procedures are experimental, and what participants will be asked to do.
- Risks and benefits: any foreseeable discomforts or dangers, along with any potential benefits to the participant or to others.
- Alternatives: other treatments or procedures that might be available outside the study.
- Privacy protections: how the participant’s identity and records will be kept confidential.
- Compensation for injury: for studies involving more than minimal risk, whether medical treatment is available if something goes wrong and how to access it.
- Voluntary participation: a clear statement that enrolling is voluntary, that refusing carries no penalty, and that participants can withdraw at any time without losing benefits they would otherwise receive.
The IRB reviews these forms not just for completeness but for clarity. A consent document filled with technical jargon defeats its purpose. The underlying principle is respect for persons: participants should genuinely understand what they are agreeing to.
Three Levels of Review
Not every study requires the same depth of scrutiny. The IRB system uses three tiers based on how much risk participants face.
Exempt review applies to research that poses very little risk, such as anonymous surveys, educational tests, or observation of public behavior. These studies still go through the IRB, but the review is narrow in scope and does not require a full evaluation of all approval criteria.
Expedited review is available when a study poses no more than minimal risk, meaning the probability of harm is no greater than what people encounter in daily life or during routine medical exams. The study must also fit within one of nine federally defined categories. In an expedited review, one or two experienced IRB members evaluate the protocol rather than the full board. Minor changes to already-approved studies can also go through expedited review.
Full board review is required for any study that does not qualify for exemption or expedited review. This typically means the research involves more than minimal risk. The full committee meets, discusses the protocol, and votes. This is the most thorough level of review and applies to most drug trials, surgical studies, and research involving vulnerable populations.
Ongoing Oversight After Approval
IRB involvement does not end once a trial begins. Federal rules require continuing review at least once a year, though the IRB can set a shorter interval for higher-risk studies. During continuing review, the board examines what has happened since the last approval: any unexpected problems, adverse events, participant withdrawals, and new information from the research team or from monitoring groups.
If a pattern emerges, such as a higher-than-expected rate of serious side effects or an unusual number of participants dropping out, the IRB can conclude that the risks have changed. It can require modifications to the study protocol, demand additional safety measures, or halt the trial altogether. This ongoing authority is what gives the IRB real power beyond its initial stamp of approval.
How IRBs Differ From Safety Monitoring Boards
People sometimes confuse the IRB with a data safety and monitoring board (DSMB), but they serve different functions. The IRB reviews a trial’s design and ethical framework before and during the study. A DSMB, by contrast, looks at the actual data coming out of the trial while it is running, often in an unblinded fashion, to determine whether the study should continue. A DSMB might recommend stopping a trial early because one treatment is clearly superior, or because an unexpected safety signal has emerged.
Not every trial has a DSMB. Small, low-risk studies may rely on the investigator for safety monitoring. Large, multi-site trials testing higher-risk interventions are more likely to use an independent DSMB. When a DSMB exists, its monitoring plan must be approved by the IRB.
Central vs. Local IRBs in Multi-Site Trials
Clinical trials often run at multiple hospitals or research centers simultaneously. Historically, each site submitted its protocol to its own local IRB, meaning a single trial could undergo dozens of separate reviews. Each local board might request different revisions to the consent form or study procedures, creating inconsistencies and significant delays in getting the trial started.
To address this, many multi-site trials now use a single central IRB (sometimes called a single IRB or sIRB). In this model, one IRB reviews the core protocol and consent documents for all participating sites. Individual sites are then reviewed separately for their capacity to carry out the study, but the ethical and scientific evaluation happens once, in a standardized way. This approach reduces redundancy and speeds up site activation, though it can create friction when local institutions are reluctant to fully defer to an outside board.