What Is an Adverse Event?

An adverse event is anything unwelcome that happens to a person's health while they are taking part in a trial or taking a medicine, whether or not the medicine had anything to do with it. A headache counts. A rash counts. A broken wrist from a fall on the ice counts. The phrase deliberately makes no claim about cause. It means that something happened, and that it has been written down.

That is the key difference from a side effect, a phrase which quietly assumes the treatment caused it. A companion article here explains why molecules cause unwanted effects in the first place — the receptors, the neighbouring tissues. This one is about the bookkeeping: how trials record what happens to people, how they sort it, and how to read the tables that come out at the end.

Why write down things the treatment did not cause?

Because at the time, nobody knows which ones it caused.

Think of an aircraft's flight recorder. It does not decide which readings are important. It records every one of them, all the time, and the investigators decide afterwards which mattered. If it only recorded what the engineers expected to go wrong, it would be useless on exactly the day something unexpected happened.

Trials work the same way. If researchers only wrote down events they already suspected were linked to the treatment, they would only ever find what they were looking for. A new kind of harm, the one nobody predicted, would be filtered out at the moment it first appeared, by someone sincerely believing it was a coincidence.

So the international definition used in trials is deliberately wide. An adverse event is any untoward medical occurrence in a patient or trial participant given a medicinal product, which does not necessarily have to have a causal relationship with that treatment. It covers any unfavourable and unintended sign, symptom or disease, including an abnormal laboratory result, whether or not anyone thinks it is related 1.

That is why a trial report can list a sprained ankle, a tooth abscess and a car accident alongside nausea. It looks odd. It is the system working as intended.

Adverse event, adverse reaction, side effect

Three terms get used as though they meant the same thing. They do not, and the whole difference between them is about cause.

The same guideline that defines these terms recommends that side effect should no longer be used in this setting, because it has been used to mean too many things, including favourable effects, and should not be treated as meaning the same as adverse event or adverse reaction 1. When you see a list headed side effects, it is worth asking which of the first two it really is.

How do researchers tell whether the treatment caused it?

There are two ways, one for individuals and one for groups.

For an individual, the doctor running the trial makes a judgement. Did the event start soon after the treatment began? Did it fade when the treatment stopped? Did it come back when the treatment restarted? Is there a known way the treatment could cause it? Is there a more obvious explanation? None of these settles the question alone, and the judgement is recorded along with the event.

For a group, the answer comes from the comparison. If headaches turn up in eight per cent of people on the treatment and eight per cent of people on placebo, the headaches are background: things that happen to people anyway, in any few months of ordinary life. If they turn up in twenty per cent against eight, something is going on.

The comparison can be startling. In one trial, people who had stopped taking a cholesterol-lowering statin because of side effects were given a year's worth of monthly bottles in a random order. Some bottles held the statin, some held identical dummy tablets, and some were empty, meaning a month with nothing at all. Nobody knew which month was which. About 90 per cent of the symptoms people reported during the statin months were also reported during the placebo months 3. The symptoms were real. Most of them were not caused by the drug.

That is why a list of adverse events from a study with no comparison group is so hard to interpret. Every event on it might be the treatment, or might just be life, and there is nothing to subtract.

How are adverse events graded?

Recording that something happened is only the start. Trials also record how bad it was, on a shared scale, so that one hospital's moderate means the same as another's. The most widely used scale is the Common Terminology Criteria for Adverse Events, drawn up by the United States National Cancer Institute and now used well beyond cancer research. It grades most events on five levels 2.

Each specific kind of event, whether nausea, a low blood count or a skin reaction, has its own detailed description at every grade, so that grading is less a matter of opinion and more a matter of checking a description. The fifth version is the one most recent trials report; a sixth version appeared in 2025.

Why serious and severe are not the same

This is the distinction people most often get wrong, including people who write about trials for a living.

Severe is about intensity: how bad the event was. Serious is about outcome: what the event did to the person. The guideline is blunt about it. A headache can be severe and still be of relatively minor medical significance, and that is not the same as serious 1.

An adverse event counts as serious if it results in death, is life-threatening, requires admission to hospital or a longer stay, causes persistent or significant disability, or is a birth defect. Medical judgement can add other important events that could lead to one of those outcomes without treatment. And life-threatening here means the person was actually at risk of dying at the time, not that the event might have been fatal had it been worse 1.

Serious events set off fast reporting. When a serious and unexpected reaction happens during a trial, regulators must be told promptly, because a single case can change how the whole programme is run 1.

What about reports after a medicine is approved?

After approval, the recording carries on in a looser form. Doctors, pharmacists, patients and manufacturers can send reports of suspected problems to national monitoring schemes. In the United States they go into a database now called the FDA Adverse Event Monitoring System, formerly known by the initials FAERS 4.

These databases are how rare harms get spotted once millions of people are taking a medicine. They are also easy to misread. The FDA's own dashboard warns that for any given report there is no certainty that the suspected drug caused the event. A report may reflect the illness being treated, or another medicine taken at the same time. Reports are not medically verified, some are duplicates, and the counts cannot be used to work out how often an event happens 4.

So a headline announcing that thousands of reports mention a particular medicine is describing a signal, something worth investigating. It is not describing a finding. Popular medicines collect more reports simply because more people take them, and news coverage of a suspected problem reliably brings in a fresh wave of reports.

How to read an adverse event table

Most trial papers include a table of adverse events. It looks forbidding and is not. Four habits make it readable.

1. Read across, not down. The figure for the treatment group means little on its own. Compare it with the same row for the placebo or comparison group.
2. Find the serious events and the dropouts. The rows for serious adverse events, and for people who stopped because of an adverse event, usually say more than the long list of mild ones.
3. Check the size. A zero in a trial of thirty people is not reassuring. Rare events need thousands of participants before they have a fair chance to appear.
4. Check the length. Events that take a long time to develop cannot show up in a trial that lasted twelve weeks.

And if there is no comparison group at all, remember that every line of the table is a mixture of what the treatment did and what would have happened anyway, and nothing in the table can pull the two apart.

Why such a plain phrase matters

Adverse event is a humble phrase, and it carries a whole way of thinking. Write everything down. Decide later. Compare against a group that did not get the treatment. Keep intensity and consequence separate. Treat a report as a question rather than an answer.

Most compounds sold as research peptides have never been through any process that records adverse events at all. There are no trial tables, no graded events and no comparison groups. That does not mean nothing happens. It means nobody has been writing it down in a way anyone could learn from.