Posters · Glossary
What is a case definition?
A case definition is a set of standard criteria used to decide whether a person counts as having a particular disease or condition in surveillance, an outbreak investigation or a study. It usually combines clinical signs, laboratory results and, for outbreaks, limits on time, place and person, often grading cases as confirmed, probable or suspected.
Every rate, ratio and trend on an epidemiology poster depends on who was counted, and two studies of the same disease can disagree simply because they counted differently. Printing the case definition is what lets a reader compare a board with anyone else's numbers.
Nuwan Madhusanka · Co-founder
5 min read · Published
| Level | Typical criteria | When it is used | Trade off |
|---|---|---|---|
| Confirmed | Laboratory evidence meeting the defined test criteria, usually with clinical signs | Reporting certain cases, comparing across places, analysing risk factors | Most specific, but misses people who were never tested |
| Probable | Clinical signs plus an epidemiological link to a confirmed case, or supportive but not definitive tests | Surveillance and outbreaks where testing is incomplete | Captures more true cases at the cost of some misclassification |
| Suspected or possible | Clinical signs alone, sometimes a partial picture | Early outbreak detection and triage | Most sensitive, includes many people without the disease |
What a case definition contains
The CDC's Principles of Epidemiology course describes a case definition as a set of standard criteria for classifying whether a person has a particular disease, syndrome or other health condition. The clinical part lists the signs and symptoms, and often a laboratory test. For an outbreak investigation it adds restrictions by time, place and person, such as onset between two dates, attendance at a particular event or residence in a district, and sometimes an age group. Once written, the same definition is applied to everyone, whether or not they appear to be linked to the outbreak, so that counting is consistent and not shaped by what investigators expect.
Levels of certainty
Most definitions grade cases by how much evidence supports them. The CDC course illustrates this with confirmed, probable and possible or suspected categories, where confirmed cases usually meet laboratory criteria and probable cases meet clinical criteria with an epidemiological link. Australia's national surveillance case definitions, published through the Communicable Diseases Network Australia, set out for each notifiable condition whether confirmed cases alone or confirmed and probable cases should be notified. The World Health Organization's COVID-19 case definitions use suspected, probable and confirmed. A poster should say which levels were counted in each analysis.
Surveillance and outbreak definitions differ
Surveillance definitions are written to be stable over years and applicable across places, so that trends and comparisons mean something. They rarely change and seldom include time or place limits. Outbreak definitions are temporary and local: they start broad, to find as many possible cases as possible, then narrow as the cause becomes clear and testing becomes available. That is why an outbreak report may present a sensitive definition for case finding and a stricter one for analysing risk factors. A board that mixes the two without saying so will produce numbers that cannot be compared with routine surveillance. When a poster compares its counts with national figures, it should use the national surveillance definition, or show how many of its cases would meet it.
Sensitivity against specificity
A broad definition catches more true cases but also sweeps in people with other illnesses, which dilutes associations with the real exposure. A strict definition is cleaner but may miss cases that were never tested, underestimating the size of the problem and biasing towards people with better access to care. Neither is right in general. The choice depends on the purpose: early detection favours sensitivity, and analysing causes or comparing rates favours specificity. The STROBE statement asks observational studies to define all outcomes and give diagnostic criteria, which is where that choice should be stated.
Common mistakes
The most common error is not printing the definition at all, so readers cannot tell what the counts mean. Others are changing the definition partway through a study without reporting when, mixing confirmed and probable cases in a rate without saying so, and using a laboratory only definition in a setting where few people are tested. A subtler problem is defining cases with the exposure built in, such as requiring attendance at the suspected event, and then reporting a strong association with that event, which the definition itself guaranteed. Finally, some boards give the definition in words that cannot be applied, such as compatible illness, without saying which symptoms or tests qualify; another team reading the poster should be able to classify a patient the same way.
Where it shows up in the poster builder
A case definition is a methods item, and the protocol block, on 38 of the 40 layouts, holds up to five labelled entries, so the definition can take one entry of its own, separate from the data source and analysis. Where the records flow matters, ten layouts carry a six node flow block with one aside, which suits a path from notifications through exclusions to confirmed cases. The epidemiology example prints its case definition in methods and reconciles a records flow from 9,412 notifications to 8,236 confirmed cases.
Questions people ask
Who writes case definitions?
For notifiable diseases, national or state public health agencies publish standard surveillance definitions, often agreed by expert networks. For an outbreak, the investigating team writes a working definition, usually adapting the national one with local time and place limits. For research studies, investigators define cases in the protocol, ideally using an established definition so results can be compared.
Can a case definition change during an outbreak?
Yes, and it often should, for example when a laboratory test becomes available or the likely source is identified. Record each version with the date it applied and report how many cases were classified under each. When analysing the whole outbreak, apply the final definition to all cases retrospectively where the data allow, so the counts are consistent.
Is a case definition the same as a diagnosis?
No. A diagnosis is a clinician's judgement about an individual patient and can use any available information. A case definition is a fixed rule for counting in a population, applied the same way to everyone. A person can be diagnosed with a condition by their doctor and still not meet a surveillance case definition, or the reverse.
Should controls in a case control study have a definition too?
Yes. Controls should be defined as clearly as cases, including how they were selected and what excludes them, such as having symptoms of the condition. Poorly defined controls can include undetected cases, which weakens any association. The STROBE statement asks case control studies to describe the sources and methods of case ascertainment and control selection.
How detailed should the case definition be on a poster?
Detailed enough that another investigator could apply it: the clinical criteria, the test and result required for confirmation, the time window and any place or person limits. That usually fits in two or three sentences. If the full definition is long, give the confirmed case criteria on the board and link the complete version through a QR code.
What is a case definition in non infectious disease research?
The same idea applies to chronic diseases, injuries and exposures. A diabetes study might define cases by a laboratory threshold or a medication record, and an injury study by hospital codes. Defining cases through routine records needs particular care, because coding practices vary between sites and change over time, which can create false trends.
Make one with posters
The button opens the generator with this use case already described. Change the wording to match your own.
Create a poster with OneCraftRelated questions
- What is a rate ratio?What is a rate ratio: the incidence rate in one group divided by the rate in another, using person time. How it differs from a risk ratio and an odds ratio.
- Cohort vs case-control studyCohort vs case control study: a cohort follows people from exposure to outcome, a case control study starts at the outcome and looks back. What each estimates.
- What is the protocol block on a research poster?What is a protocol on a research poster: methods set out as short labelled entries instead of paragraphs. What each entry holds, and when prose works better.
Step by step in the builder: Choose the template, theme and size.
Written and checked by the OneCraft team. Last checked .