How to Investigate a Disease Outbreak, Step by Step

By Jenny Lam

Image Credit: Wikimedia Commons

Outbreaks of disease occur much more frequently than one may expect. According to the CDC, over 500 outbreaks of foodborne illnesses were recorded each year during the 1990s, and that’s only considering one type of illness. Properly dealing with outbreaks is one of many ways epidemiology, the study of the distribution and patterns of diseases, finds its place in public health. Due to the COVID-19 pandemic, the topics of epidemiology and outbreak investigations have only gotten more relevant. There are many steps involved in conducting an outbreak investigation, but because these investigations must be done quickly and accurately to ensure public safety, many of the steps presented below are done at the same time.

First, field investigators must adequately prepare for a field investigation by making sure they have enough scientific knowledge, equipment, and laboratory materials to carry out the investigation, along with having an action plan. Investigators don’t do this alone; they also have to put together a team of people that will help them investigate the scene, which often varies depending on the situation. For example, an outbreak investigation that is suspected to involve bioterrorism might include the FBI, while one that deals with a foodborne illness might include the FDA.

Investigators need to also make sure what they are investigating is indeed an outbreak. By comparing the observed number of cases with the expected data from previous months, they can figure out if a cluster of cases is large enough to be considered unusual, and therefore be classified as an outbreak. However, if a disease is deemed severe or infectious enough, an investigation can still be launched, even if the reported cases aren’t high enough to be considered an outbreak. To ensure that a disease has been identified properly, investigators will also review lab results and clinical data, as well as talk to patients to get a better picture of the situation at hand to later identify a possible cause of the outbreak.

The next step is to establish a case definition, a set of criteria that defines whether one has the disease of interest or not. A case definition includes four components: the clinical criteria (symptoms of disease), characteristics of affected people, location, and time sequence. For example, if an outbreak happened in a hospital among the elderly, the case definition might include the common symptoms of the affected people, the location of the hospital, and might define an age range that includes the affected people. With this, investigators can categorize people who are confirmed or suspected to have the disease according to the case definition, make it easier to learn more about the outbreak and analyze it. To identify cases and record information, public health officials may request reports from hospitals with similar cases of the disease or visit the affected healthcare facilities to obtain information. If only a restricted population was affected, such as people on a specific plane, they might survey the entire group and ask case-patients if they know anyone else with the same condition.

Image Credit: Wikimedia Commons @ CDC

To characterize the outbreak, descriptive epidemiology is used to organize the data in terms of person, place, and time. Summarizing the data of an outbreak this way can give insight into the trend of the disease over time, the source of the outbreak, and predict what types of people are at risk for contracting the disease.

To investigate time, a histogram is usually created to monitor the number of cases over a period of time. This type of histogram—one that plots the number of cases against the time of onset of illness—is known as an epidemic curve, or epi curve. One can learn a lot about an outbreak just by looking at the shape of an epi curve. For instance, a type of outbreak where people are exposed to the same source over a short period of time (called a point-source epidemic) will look like a steep hill with a gradual downward slope, whereas an outbreak where exposure to the cause of the disease (agent) is sporadic over time (intermittent common-source epidemic) will have an irregularly jagged curve to represent the scattered and irregular nature of this type of outbreak. Epi curves can also convey other types of information, such as how long the incubation period is and when the period of exposure might have been.

To investigate place, maps are used to mark where affected individuals were during a given period of time. By looking at clusters of people at a specific place, investigators narrow down the location or item that caused the disease. Lastly, taking note of the characteristics of affected people gives insight into what kind of population is at risk for the disease. Investigators may look into characteristics of the host, such as age, sex, race, and medical status, and possible factors that could be related to the exposure, such as the host’s daily activities and medications.

After having a better picture of the outbreak, investigators will develop hypotheses around different aspects of the situation, such as the source of the agent, the mode of transmission, and exposures that cause the disease. To test these hypotheses, they compare it with established facts and previous data; if there is an obvious and strong explanation to support the hypothesis, there is often no need to carry out the next step, analytic epidemiology. Analytic epidemiology is used when there is less evidence to support a hypothesis and involves quantification of the relationship between variables, such as a disease and an exposure that causes the disease. The case-patients are compared to a comparison group, which is often a group of people who don’t have the disease, to identify differences between the two and find patterns. Two common types of studies used in analytic epidemiology are retrospective cohort studies and case-control studies. The former is used for outbreaks in small, well-defined populations, while the latter is used when the population isn’t well defined and the investigator can’t follow them over a period of time. To assess the possibility that an exposure of interest may be the actual cause of the disease, statistical tests (e.g. chi square test) are run to determine the statistical significance that an exposure is related to the disease.

When analytic epidemiology fails to reveal any useful information, investigators then re-evaluate and refine their hypotheses, such as making it more specific or refocusing the investigation onto another possible source of the outbreak. If a cause for the outbreak is found, they will need to confirm it with laboratory evidence and environmental studies to provide backing for the claim.

quarantine; covid-19
Image Credit: Wikimedia Commons @ Tistip

Although the next step, which is to implement control and prevention measures to ensure public safety, is listed as one of the last, it should be done as early as possible. Control measures are often put into action even before an outbreak investigation is launched. After all, the health of the people take top priority. Control measures aim to stop the disease at different points in the chain of transmission. If investigators find an underlying cause for the outbreak, they may try to directly eliminate it at the source. Other ways to implement control measures include preventing the disease’s mode of transmission (e.g. isolating a group of affected people so the disease can’t infect anyone else) and guarding paths that diseases can travel to infect a patient (e.g. wearing gloves and masks to protect from coming into contact with fluids). Vaccines and other drugs are used to increase the public’s defenses against disease. These measures are experienced firsthand by much of the world during the COVID-19 pandemic, as many quarantine at home and wear masks to reduce the chances of spreading the virus to others.

If active surveillance hasn’t been going on yet at this point, it is initiated to monitor the situation and determine whether the control measures are effective, as well as if the outbreak has spread outside of the original area. If so, the prevention and control measures will have to be extended into the new area.

Finally, the last step is to summarize the findings and outcomes of the investigation into a report and communicate it, either through an oral briefing or in a written report. A report serves as a record of the investigation, a document for legal issues, and as a reference to use for future outbreaks.

Because every outbreak is different, investigators will not always conduct investigations in the same way and this process can change depending on the situation. However, because of the unexpected nature and potential danger of outbreaks, it is necessary to have a systematic plan, like the one discussed here, to efficiently carry out the investigation, get to the source of the outbreak, and ensure that it doesn’t get out of hand.


1. What is the difference between a cluster, outbreak, epidemic, and pandemic?

A cluster is simply an aggregation of cases in a certain area or among a certain group of people. It doesn’t take into account whether the aggregation of cases is more than what is expected from that area or group of people. One of first the tasks of an outbreak investigation is to determine whether a cluster can be considered an outbreak, which is when there are more cases than expected in an area or among a group. The terms epidemic and outbreak are often used interchangeably, but an outbreak usually describes a smaller geographic area than an outbreak. A pandemic is an epidemic that occurs over several countries and continents and affects a large proportion of the population.

2. How do descriptive and analytic epidemiology differ?

Descriptive epidemiology deals with the “whats” (i.e. the facts) of the outbreak, such as what happened, the time and location, and who gets affected. Analytic epidemiology addresses hypotheses for risk factors, exposures, etc. that come up during the investigation. Descriptive studies include case reports and case series, etc., while analytics studies include case-control and cohort studies. Descriptive studies are used to generate hypotheses that are tested using analytic studies.


Image Credit:

Public domain: File:SalmonellaNIAID.jpg - Wikimedia Commons

Public domain: File:Epi Curve 2018 US Salmonella Outbreak.jpg - Wikimedia Commons

No changes were made, File:Quarantine COVID-19.png - Wikimedia Commons, License: Creative Commons — Attribution-ShareAlike 4.0 International — CC BY-SA 4.0