Skip to content
Scintria
ToolsGrant writerFunding
HomeLiteratureClinical trialsEvidence LabProtocol packData analysisLearnProjectsLibraryReferencesApplications
Start free
ScintriaStart free

Learn / Epidemiology / Descriptive epidemiology, surveillance and outbreaks

Investigating an outbreak

15 minute read · 6-question quiz

Download this lesson as a colour Word or PDF study guide with the figures and answers. Included in paid plans; reading and quizzes stay free.Sign in

What you will learn

  • List and explain the steps of an outbreak investigation

  • Write a case definition and calculate attack rates

  • Identify the source with a food-specific attack-rate table and relative risk

An outbreak is the occurrence of cases in excess of the number normally expected. Investigating one is a core skill for district health teams and field epidemiology programmes. The steps are a checklist, not a strict order: several happen at once, and control measures start as soon as there is a reasonable idea of the source.

  1. 1

    1. Prepare

    Assemble the team, laboratory support, supplies and permissions; agree roles.

  2. 2

    2. Establish the existence of an outbreak

    Compare current cases with the expected number using surveillance data. Rule out artefacts such as a new test or a change in reporting.

  3. 3

    3. Verify the diagnosis

    Review clinical findings and laboratory results; collect specimens early.

  4. 4

    4. Construct a case definition

    Use clear clinical, time, place and person criteria. Start sensitive, then tighten (suspected, probable, confirmed).

  5. 5

    5. Find cases systematically

    Look at health facility registers, ask clinicians, search the community, interview. Build a line list.

  6. 6

    6. Describe by time, place and person

    Epidemic curve, spot map, attack rates by group.

  7. 7

    7. Develop hypotheses

    What source, mode of transmission and risk factors explain the pattern?

  8. 8

    8. Evaluate hypotheses

    Compare cases with non-cases: a cohort analysis if the group is defined, otherwise a case-control study. Add laboratory and environmental evidence.

  9. 9

    9. Implement control measures

    Do not wait for the final answer: isolate cases, treat water, close the premises, vaccinate.

  10. 10

    10. Communicate

    Brief the authorities, inform the public, write a report, plan follow-up.

The case definition

A case definition is a standard set of criteria for deciding whether a person counts as a case. It includes clinical features, and limits on time, place and person, e.g.:

Suspected case: any resident of or visitor to Kyotera District with acute watery diarrhoea (three or more loose stools in 24 hours) with onset between 1 and 20 March 2026. Confirmed case: a suspected case with Vibrio cholerae O1 or O139 isolated from stool.

Early in an investigation use a sensitive definition so you do not miss cases; later, a specific one so you do not dilute the analysis. National and WHO definitions exist for notifiable diseases.

Line list

A table with one row per case: ID, age, sex, location, date of onset, symptoms, exposure history, laboratory result, outcome. It is the working tool of the investigation.

Worked example: a wedding feast

A wedding in a village is followed by gastroenteritis in 55 of 100 guests. Interviewing everyone about foods, the team tabulates attack rates by exposure for each food.

For the chicken salad:

IllNot illTotalAttack rate
Ate chicken salad4555090%
Did not eat10405020%

Relative risk = 90% / 20% = 4.5. Those who ate the salad had 4.5 times the risk. The odds ratio is 45 × 40/(5 × 10) = 36. Other foods might show:

FoodAttack rate if ateAttack rate if did not eatRR
Chicken salad90%20%4.5
Rice56%54%1.0
Cassava55%55%1.0
Cake60%50%1.2

Only the chicken salad stands out. Ten ill people did not eat it: perhaps they ate it without realising, or have another cause (background cases). A good analysis explains the exceptions. The team would send salad samples for testing, and examine how the salad was prepared and stored.

Chicken salad and illnessIllNot illAte salad455Did not eat1040Attack rate: 90% vs 20%Relative risk = 4.5Odds ratio = 36
Foodborne outbreak analysis in a two-by-two table.

Cohort or case-control?

If you can list everyone exposed (guests at a wedding, a school, a ward), do a retrospective cohort: compute attack rates and relative risks. If the exposed population is large and undefined (a town's water supply), do a case-control study: compare cases with a sample of healthy controls and compute odds ratios.

The pitfalls

Waiting for laboratory confirmation before acting; defining cases too narrowly; forgetting the non-cases; blaming the most memorable food (recall bias); presenting counts without denominators; failing to feed back to the community. Control first, publish later.

A pocket checklist

Before leaving the field: Do you have a line list, an epidemic curve, a map, attack rates, a hypothesis tested with comparison data, laboratory results, control measures in place and a plan to tell the community?

Summary

  • Confirm the outbreak and the diagnosis, define cases, find cases, describe, hypothesise, test, control and communicate.

  • Start with a sensitive case definition and tighten it as understanding grows.

  • Compute food- or exposure-specific attack rates and relative risks to find the source.

  • Use a cohort design when the exposed group is defined; case-control otherwise.

  • Act on control measures early; do not wait for certainty.

Check your understanding

Question 1
1

The first purpose of a case definition is to:

Question 2
2

Early in an outbreak the best case definition is:

Question 3
3

In the wedding example the attack rate among guests who ate the salad was:

Question 4
4

A food with an RR near 1.0 is probably:

Question 5
5

When should control measures begin?

Question 6
6

For an outbreak in a large town linked to its water supply, the appropriate analytic design is likely:

Draft study guide. Spotted an error or something unclear? Tell us at hello@scintria.com. This is educational material, not clinical or regulatory advice.

Person, place and timeSurveillance: watching for disease systematically