Case series

A case series is a study that collects and describes a group of individuals who share a common exposure, clinical feature, or outcome, reporting them together without a separate comparison group. It documents what happened to those specific people; it does not include a group that lacked the exposure to compare against.

In the evidence hierarchy, a case series sits above a single case report — which describes one individual — but well below any design that includes a control arm. It is entirely observational: the researcher records the patterns in the assembled cases and cannot test whether those patterns hold more broadly.

Case series arise frequently where a comparison group is impossible or premature to construct — early reports of an unusual reaction, a rare population, an emerging substance or practice. They often represent the first evidence on record for a phenomenon, and that is their primary role in the research sequence.

What this design can establish

A well-conducted case series can show that something happened across multiple individuals under similar conditions — that an outcome is not a one-person anomaly. If a single case report can be dismissed as coincidence, a consistent series demonstrates that the observation recurs.

It can describe a range of presentations: what a reaction looked like, how variable it was, what timing and features the cases shared or diverged on. This descriptive depth is useful when a phenomenon is too new or too rare for a controlled design to exist.

A case series can also establish temporal sequence — that exposure preceded an outcome in the documented individuals. That ordering is a necessary first step toward a causal claim, even if it does not complete one.

Most practically, case series generate hypotheses. A cluster of consistent adverse events becomes the signal that motivates a controlled study, a pharmacovigilance review, or a formal interaction study.

What it cannot

A case series cannot establish causation. Without a comparison group there is no way to know whether the documented outcome would have occurred in similar people who lacked the exposure. The reader's natural inference — "all these patients took X and developed Y, so X causes Y" — is not supported by this design alone.

It cannot establish frequency. A series of cases says nothing about whether the event occurs in one person in a hundred or one in a million. The cases appear in the series precisely because the outcome happened; everyone for whom it did not is absent from the record.

It cannot identify risk factors reliably. People appear in a case series because they experienced the outcome, not because they were drawn systematically from the population. Any shared characteristic may reflect who gets noticed and reported rather than who is genuinely at higher risk.

The most common wrong conclusion is that a consistent pattern across a series confirms an effect in the general sense. It confirms the effect occurred in those cases. How often, in whom, and under what conditions it occurs remain open questions until a controlled design addresses them.