What Is an Adverse Event in Research?

What exactly counts as an adverse event in research?

A participant develops a headache after a study visit. Another becomes dizzy during a procedure. A laboratory test returns an abnormal result. Someone enrolled in a behavioral study experiences significant psychological distress. Are these adverse events, even if nobody yet knows whether the research caused them?

The term adverse event, often abbreviated as AE, can be confusing because “adverse” sounds as though the research must have harmed the participant. That is not what the term necessarily means. In commonly used research-safety frameworks, identifying an adverse event and deciding what caused it are separate questions.

There is also no single definition that applies identically across every research setting. The U.S. Office for Human Research Protections (OHRP), for example, notes that the HHS human-subject regulations do not define or use the term and that definitions differ among government and non-government entities. Researchers therefore need to understand both the general concept and the specific definition governing their study.

An adverse event is an occurrence, not a conclusion about causation

How adverse events are identified and understood

The definition is deliberately broader than “side effect”

OHRP uses adverse event broadly to mean an untoward or unfavorable medical occurrence in a human subject, including an abnormal sign, symptom, or disease, that is temporally associated with research participation, whether or not it is considered related to that participation. OHRP further explains that adverse events can encompass physical and psychological harms.

This means that an AE is not limited to a participant saying, “I had a side effect.” A researcher might identify an event through a participant's report, clinical examination, laboratory result, observation, medical record, or another source specified by the study's safety-monitoring procedures.

For clinical trials involving investigational products, the applicable wording may be narrower. ICH E6(R3), for example, defines an adverse event as an unfavorable medical occurrence in a trial participant who has been administered the investigational product and explicitly states that the event does not necessarily have a causal relationship with the treatment.

Temporal association is not the same as causation

The word temporally matters. It means that the event occurred during a relevant period associated with participation or exposure. It does not mean that one event caused the other.

Suppose a participant develops influenza while enrolled in a six-month clinical study. Depending on the protocol and applicable definition, the illness may be captured as an adverse event even if there is no plausible connection to the intervention. Classification as an AE can therefore be the beginning of a safety assessment rather than its conclusion.

This is why researchers should not exclude an event simply because they initially believe it is unrelated. Whether an AE must have been caused by the research, and how causality affects subsequent obligations, requires a separate assessment of causality and reporting.

An adverse event can be a sign, symptom, disease, or abnormal finding

An AE does not have to be a formally diagnosed disease. Depending on the governing definition and study protocol, it may appear as an unfavorable sign or symptom, such as dizziness, rash, nausea, pain, or clinically relevant psychological distress. It may also appear as an abnormal physical examination or laboratory finding. OHRP expressly includes abnormal signs, symptoms, and diseases within its broad definition.

Whether every abnormal measurement should be recorded as an AE is more study-specific. Protocols and safety-monitoring plans may define when an abnormal laboratory value becomes an AE, for example because it is clinically significant, represents a new condition, worsens from baseline, or requires intervention. Researchers should therefore use the protocol's operational definition rather than assuming that every value outside a laboratory reference range automatically constitutes a separate AE.

A worsening pre-existing condition may become an adverse event

A condition that already existed before participation is not necessarily a new adverse event merely because it remains present during the study. The distinction often becomes important when the condition changes.

Study protocols commonly treat a clinically meaningful worsening of a pre-existing condition as an AE. For example, NIH-supported trial materials have distinguished unchanged pre-existing conditions from conditions that worsen in severity, frequency, or character during the relevant study period. The exact rule, however, should come from the protocol and applicable safety framework rather than from a generic definition alone.

Adverse events are not confined to drug trials

The terminology is strongly associated with clinical trials, but participant harm is not exclusively pharmacological. OHRP notes that adverse events encompass physical and psychological harms and may occasionally occur in social and behavioral research as well as biomedical research.

For example, research procedures involving traumatic experiences may be associated with psychological distress. Whether a particular episode satisfies the study's formal AE definition will depend on the applicable protocol and oversight framework. Regardless of terminology, researchers conducting sensitive studies should have procedures for responding to unexpected participant distress rather than waiting to debate labels while a participant needs assistance.

Several different questions come after identifying an AE

Once an event has been identified, researchers may need to characterize it along several dimensions. These dimensions should not be collapsed into a single judgment.

These distinctions matter because an event can occupy different combinations of categories. An event may be unrelated but serious, related but non-serious, expected or unexpected, and subject to different reporting requirements. The question of what makes an adverse event serious therefore needs to be assessed separately from whether an AE occurred at all.

An adverse event is not automatically an unanticipated problem

In the OHRP framework, adverse event and unanticipated problem are overlapping but different concepts. OHRP considers an unanticipated problem to involve an incident, experience, or outcome that is unexpected, related or possibly related to the research, and suggests that the research places participants or others at greater risk of harm than was previously known or recognized. Some unanticipated problems do not involve an adverse event at all.

Researchers should therefore avoid using the two labels interchangeably. The distinction between an unanticipated problem and an adverse event can determine how a safety occurrence is evaluated and handled.

A participant develops dizziness during a research study

The important point is that the team does not need to prove that the intervention caused the dizziness before recognizing the occurrence as an AE under this hypothetical protocol. Nor does the label itself prove causation. The appropriate person must subsequently assess whether the event is related to the research.

Common mistakes when interpreting adverse events

What to do when something unfavorable happens to a participant

When you learn that a participant has experienced an unfavorable medical or psychological occurrence, avoid making the first decision based solely on whether you think the research caused it. Start with the definition and procedures specified for your study.

If a participant becomes acutely ill or injured, classification and paperwork should not displace appropriate attention to the participant's immediate welfare. Researchers should follow the study's emergency and safety procedures for illness or injury occurring during research, then complete the necessary assessment and documentation.

Before deciding how to classify a participant-safety event

Frequently asked questions about adverse events in research

Recognize the event first, then determine what it means

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.