What Counts as Research Risk, and How Do You Know Whether It Is Acceptable?

When Does Something in a Study Count as a Research Risk?

A blood draw can cause bruising. An interview about bereavement may cause distress. A leaked dataset could expose a participant's health status, immigration status, political activity, or other sensitive information. Even a seemingly harmless survey might create social or professional consequences if someone's responses can be traced back to them.

These possibilities are not ethically equivalent, but they raise the same basic question: what harm could occur because someone participates in the research?

Identifying a possible harm is only the beginning. Researchers must also consider how likely that harm is, how serious it could be, whether it can be prevented or reduced, and whether the remaining risk is justified. That last judgment is why simply declaring a study “low risk” is not much of an ethics assessment.

Research Risk Is the Possibility of Harm Resulting From Research Participation

How Research Risk Is Identified and Judged

Risk Is a Possibility of Harm, Not the Harm Itself

The Belmont Report makes a useful distinction: risk refers to the possibility that harm may occur. If a participant experiences a severe allergic reaction during an experimental intervention, the reaction is the harm. Before it occurs, the possibility of that reaction is a risk.

This distinction matters because researchers usually make ethical decisions before knowing exactly what will happen. Risk assessment therefore deals with uncertainty. You are asking what could plausibly go wrong and how much ethical weight that possibility deserves.

Two Dimensions Matter: Probability and Magnitude

Calling something “high risk” or “low risk” can conceal two different questions. The Belmont Report specifically distinguishes the probability of harm from its magnitude.

A relatively common but minor harm and a very rare but catastrophic harm present different ethical problems. Neither can be evaluated adequately by probability alone. A one-in-a-million possibility of temporary embarrassment is not ethically equivalent to a one-in-a-million possibility of death.

This becomes especially important when considering very unlikely but very serious harms. Low probability does not automatically make a consequence ethically negligible.

Research Risk Extends Beyond Physical Injury

Researchers sometimes associate “risk” primarily with clinical trials, invasive procedures, drugs, or medical devices. That is too narrow. The Belmont Report explicitly recognizes psychological, physical, legal, social, and economic harms. Contemporary research may also create reputational harms or harms arising from privacy and confidentiality failures.

For that reason, social science, educational, qualitative, digital, and observational research should not be assumed to be risk-free simply because nobody receives a drug or undergoes a procedure. An interview may disclose trauma. A workplace survey may expose criticism of a supervisor. Location data may reveal behavior participants never intended others to know.

The relevant types of research risk depend on what the study asks participants to do, what information is collected, who could gain access to it, and what could happen as a result.

Risk Must Result From the Research

Not every bad thing that could happen to a participant is a research risk. The relevant question is whether the research creates, increases, or exposes the participant to the possibility of harm.

This distinction is explicit in the U.S. Common Rule. When evaluating risks and benefits, an institutional review board considers those arising from the research rather than risks participants would face from therapies they would receive even without participating.

Suppose a study recruits patients who already undergo a painful procedure as part of ordinary clinical care. The underlying illness and the clinically required procedure do not automatically become research risks simply because researchers study those patients. If the protocol requires an additional biopsy solely for research purposes, however, the risks associated with that additional biopsy are attributable to research participation.

Risk Is Not the Same as Burden or Inconvenience

A two-hour interview, repeated questionnaires, travel to a research site, or a tedious cognitive task may impose substantial demands even when the probability of actual harm is small. Conversely, an activity requiring almost no participant effort could create a serious confidentiality risk.

Risk, burden, and inconvenience can overlap, but they are not interchangeable. Keeping risk, burden, and inconvenience conceptually separate helps researchers describe what participation actually entails rather than placing every undesirable feature of a study under the label “risk.”

First Ask Whether the Risk Is Necessary

Before asking whether a risk is justified, ask whether participants need to face it at all. This is one of the most important steps in ethical study design.

The Belmont Report states that risks should be reduced to those necessary to achieve the research objective. Similarly, the U.S. Common Rule requires risks to be minimized through procedures consistent with sound research design that do not unnecessarily expose participants to risk. The 2024 Declaration of Helsinki requires measures to minimize predictable risks and burdens.

This creates an important sequence: minimize first, justify what remains second. A favorable potential benefit does not excuse an avoidable risk.

Acceptable Does Not Mean Risk-Free

Ethical research is not necessarily research in which nothing unpleasant or harmful could possibly happen. Requiring zero risk would make much valuable human research impossible. The ethical requirement is more demanding in another way: risks should be identified, minimized, managed, disclosed where appropriate, and justified.

This is why not every research risk has to be eliminated. Some risks are inherent in answering a legitimate research question. The relevant issue is whether exposing participants to the remaining risk can be ethically defended.

There Is No Universal Numerical Cutoff for Acceptable Risk

Researchers sometimes want a formula: below a particular probability, risk is acceptable; above it, risk is unacceptable. Research ethics generally does not work that way. The Belmont Report acknowledges that risk-benefit balancing rarely permits precise quantitative calculation.

Context changes the judgment. The seriousness and probability of possible harms matter, but so do the characteristics of the participants, the procedures involved, available alternatives, possible benefits, safeguards, and the scientific and social value of the knowledge sought.

Regulatory concepts such as minimal risk can affect review requirements and ethical judgments, but “minimal risk” is not a synonym for “automatically acceptable.” A poorly designed study may expose people to even small risks without sufficient scientific or social value to justify them.

Scientific Quality Is Part of the Ethical Equation

A study cannot justify participant risk merely by promising that something interesting might be learned. The expected knowledge must be credible enough to matter.

CIOMS states that imposing research risks requires social and scientific value. The Declaration of Helsinki likewise requires scientifically sound and rigorous design likely to produce reliable, valid, and valuable knowledge. This connects methodology directly to ethics: if a study is so poorly designed that it is unlikely to answer its research question, exposing participants to its risks becomes harder to justify.

In other words, methodological weakness can become an ethical problem. Peer reviewers have been saying this in considerably less dramatic language for decades.

Acceptability Requires More Than Informed Consent

Voluntary informed consent is essential in many forms of human-participant research, but consent does not transform an otherwise unacceptable study into an acceptable one. Participants cannot simply authorize researchers to disregard unnecessary or disproportionate risks.

Independent ethics review therefore examines the protocol itself. Under the U.S. Common Rule, for example, an IRB must determine that risks are minimized and reasonable in relation to anticipated benefits, if any, and the importance of the knowledge reasonably expected to result.

CIOMS similarly requires risks to be minimized and appropriately balanced against potential individual benefit and the social and scientific value of the research. Its framework also distinguishes procedures that may directly benefit participants from procedures that offer no such prospect, an important distinction when a study contains several different interventions.

Risk Acceptability Can Change Over Time

Ethical approval is based on the information available when a protocol is reviewed. That does not freeze the assessment for the entire study.

The Declaration of Helsinki requires risks and burdens to be continuously monitored, assessed, and documented. New adverse events, unexpectedly high distress, a confidentiality breach, emerging evidence about an intervention, or a safer alternative can change what was originally considered reasonable.

When a new risk becomes apparent after research begins, researchers should not simply rely on the original approval. The new information may require additional safeguards, protocol modification, renewed consent, temporary suspension, or, in some circumstances, stopping the research.

How a Seemingly Simple Interview Can Create Research Risk

The lesson is not that the interview is automatically acceptable or unacceptable. It is that the ethical assessment begins with the actual pathways to harm rather than the superficial label attached to the method. “It is only a survey” and “it is only an interview” are descriptions of data-collection techniques, not risk assessments.

Common Mistakes When Thinking About Research Risk

How to Decide Whether the Risks in Your Study Are Ethically Defensible

Do not begin by trying to prove that your study is “low risk.” Begin by mapping the ways participation could plausibly lead to harm. That change in starting point tends to produce a more useful assessment because it forces you to examine the study as participants may actually experience it.

For each meaningful risk, consider its probability, possible magnitude, who is exposed, what creates the risk, and what safeguards could reduce it. Then examine the remaining risks in relation to possible participant benefits and the scientific and social value of the knowledge expected from the study.

A formal risk-benefit assessment develops this reasoning in greater detail. For the initial design stage, however, one principle is particularly useful: do not ask only, “How risky is this study?” Ask, “What creates each risk, what can I do about it, and what justifies what remains?”

Before Describing the Risks in Your Protocol, Check These

Frequently Asked Questions About Research Risk

Acceptable Research Risk Is Risk That Has Been Identified, Minimized, and Justified

How to Cite This Guide

This guide is intended to be read, shared, and used in research, teaching, and academic work. If you draw on its ideas, explanations, or other content, please acknowledge the source by citing the guide. Doing so gives appropriate credit and helps your readers locate the original resource.