Adverse Event (AE)

An adverse event is any untoward medical occurrence in a patient or clinical investigation participant administered an intervention, whether or not the event has a causal relationship with the treatment under study.^[3] The term encompasses all negative health experiences that occur during the study period, from minor laboratory abnormalities to life threatening conditions.

Systematic collection and reporting of adverse events is a regulatory requirement for regulated clinical trials (for example, drug trials under a US IND and EU trials of medicinal products) and forms the foundation of safety evaluation for new therapies, medical devices, and vaccines. Adverse event data directly influence regulatory decisions about product approval, labeling requirements, and post market surveillance obligations. For clinical researchers, accurate AE documentation protects participant welfare by helping ensure that safety concerns are identified and addressed promptly. For systematic reviewers and meta analysts, AE data from multiple trials can be pooled to identify safety signals that individual studies are too small to detect.^[2] Incomplete or inconsistent AE reporting undermines the evidence base for clinical decision making.^[1][2]

During a clinical trial, investigators record the adverse events experienced by participants, as the protocol specifies, from the time of enrollment through the end of the follow up period. Recording occurs regardless of whether the event is thought to be related to the study intervention; this broad capture reduces the chance that an important safety signal is missed.

Each AE is characterized along several dimensions. Severity is often graded as mild (causing no limitation of usual activities), moderate (causing some limitation), or severe (causing inability to carry out usual activities). Seriousness is a distinct classification based on whether the event results in death, is life threatening, requires hospitalization or prolongation of existing hospitalization, causes persistent disability, results in a congenital anomaly, or requires medical intervention to prevent one of these outcomes.^[3] The distinction between severity and seriousness is important: a severe headache is not necessarily a serious adverse event, while an event that requires hospitalization is serious even when its symptoms are mild.^[3]

Investigators also assess the relationship between the AE and the study intervention, using categories such as unrelated, unlikely, possibly related, probably related, and definitely related. This assessment is inherently subjective, but standardized guidance helps improve consistency across sites. The Medical Dictionary for Regulatory Activities (MedDRA) provides a standardized terminology for coding AEs, supporting consistent classification across studies and regulatory submissions.

Serious adverse events (SAEs) trigger expedited reporting requirements, but who must report, to whom, and how quickly depends on the jurisdiction and on each party's role. In US drug trials conducted under an IND, for example, the investigator must report any serious adverse event to the sponsor immediately, whether or not it seems related to the drug.^[5] The sponsor then reports qualifying serious, unexpected suspected adverse reactions to the FDA within 15 calendar days of determining that the information qualifies, or within 7 calendar days of initial receipt when the reaction is unexpected and fatal or life threatening (21 CFR 312.32(c)).^[4] Reporting to the ethics committee or IRB is a separate obligation governed by that committee's own requirements. Other jurisdictions set their own routes and deadlines, so trial teams follow the rules that apply to their study. Safety data are summarized in tables that compare AE incidence rates between treatment and control arms, enabling reviewers to identify patterns.

Adverse event monitoring is required by regulation in drug, biologic and device trials (for example, 21 CFR Part 312 for US IND trials and Regulation (EU) No 536/2014 for EU trials of medicinal products), is expected in interventional clinical trials generally, and is also recommended in observational studies that evaluate therapeutic interventions. Researchers should establish clear AE collection procedures before the study begins, including standardized definitions, reporting timelines, and data capture forms. AE reporting is relevant to systematic reviewers who assess the safety of interventions and to health technology assessment bodies that weigh benefits against harms.

A common mistake is inconsistent AE grading across study sites, which makes it difficult to compare safety profiles in multicenter trials. Site investigators should receive standardized training on severity and seriousness classifications before enrollment begins. Another error is failing to distinguish between solicited adverse events (those specifically asked about using checklists or diaries) and unsolicited events (those spontaneously reported by participants). Studies that actively solicit AEs typically report higher rates, and pooling solicited and unsolicited AE data without accounting for this difference can produce misleading comparisons.

Underreporting is also a significant problem, particularly for adverse events perceived as minor or expected. If investigators do not consistently document all events, the published safety data will underestimate the true burden of treatment related effects. Finally, some researchers report AEs only as percentages without providing the actual number of events and participants, which prevents readers from assessing the precision of the estimates.

In a Phase III trial of a new oral diabetes medication involving 3,200 participants, the safety monitoring team records 1,847 adverse events over 52 weeks. Of these, 412 are classified as possibly or probably related to the study drug. Fourteen serious adverse events are reported, including three cases of acute pancreatitis in the treatment arm and none in the placebo arm. This imbalance triggers a safety signal review by the data monitoring committee, which recommends adding pancreatitis to the informed consent form and implementing enhanced monitoring for pancreatic enzymes.