With the approaching release of the movie
Contagion, I thought it would be appropriate to post my cheat sheet on how to investigate a disease outbreak. Aspiring disease detectives take notes!
What do you think of when you hear the word “outbreak”? Maybe you envision a population decimated by a terrible, novel, and incurable disease like in the aformentioned movie
Contagion or you think of
Dustin Hoffman roaming around California in a blue biocontainment suit with
Rene Russo trying to protect folks from a tiny monkey and narrowly preventing an airstrike by the US military?
Hollywood has done their best to capture what an outbreak is…but here are the facts. An
outbreak, or
epidemic, occurs when there are more cases of disease than would normally be expected in a specific time and place. The disease may be something doctors have already seen before just in a new form or abnormally high numbers, such as
foodborne or
healthcare-associated infections, or it may be an
emerging disease that we don’t know much about like
SARS. Either way, we need to investigate to determine why it is happening and how to prevent other people from getting sick or dying.
Outbreaks are usually noticed by an astute clinician, such as those who first noticed
AIDS in New York City and San Francisco, but there are also many high tech disease detection systems available to help us spot any increase in illness.
PulseNet is a laboratory network that uses PFGE (pulsed-field gel electrophoresis) to help identify foodborne outbreaks by monitoring the genetic make-up of the bacteria causing what may otherwise look like unrelated illnesses. In the recent events of the Salmonella outbreak in ground turkey, PulseNet and the
National Antimicrobial Resistance Monitoring System helped identify the cause of the outbreak as well as determine how widely it had spread. Programs such as
Biosense and
First Watch monitor the chief complaint or reason that someone called 9-1-1 or went to the hospital (aka syndromic surveillance). We also monitor news media for reports of outbreaks and websites such as
Google Flu trends, which tracks circulating viruses and illnesses. With new technology ordinary citizens can also increasingly report outbreaks in their communities too.
The Magic Formula
So how do you figure out the who, what, when, and where of a disease outbreak? We usually teach our
disease detectives a 10-step process for
investigating outbreaks, which I’ve condensed into 5 steps below.
When we investigate outbreaks they are often in urgent situations with patients and their families wanting immediate answers and the news media asking why we are not working hard and fast enough. This can be a lot of pressure when you are trying to make sure you gather all the clues and piece them together properly. If you’re wrong, not only do more people die, but you may implicate the wrong product, microbe, or disease transmission (such as the false accusation of Spanish cucumbers as the cause of the recent
European E. coli outbreak). For that reason it’s important you follow each step.
Step 1: Determine an outbreak is occurring (seems obvious, I know)
Too often an initial suspicious outbreak is just lots of different illnesses that are being confused for the same thing or an error from the lab or other monitoring system. So the first step involves lots of listening and then asking some basic questions:
- What are the signs and/or symptoms?
- Is this an increased number for this area, time of year, or age group? Are the cases linked to a common source or agent? Keep in mind that a change in disease monitoring, laboratory diagnostics, reporting requirements or even increased publicity might affect the number of cases reported without accurately reflecting a true disease outbreak.
- How many cases are there?
- The 5 “W’s” of any investigation apply here as well. Who? What? Where? When? Why?
The most critical piece here is often the “what is the problem” also referred to as
verifying the diagnosis. This is where the Sherlock Holmes part of being a disease detective gets turbocharged with leading edge laboratories that should best be called CSI-CDC. The CDC labs were the first labs to identify the agents responsible for a long list of diseases such as
Legionella,
Hantavirus Pulmonary Syndrome,
Ebola,
West Nile virus in the US,
SARS, and most recently the novel H1N1
influenza virus.
Step 2: Now that we have confirmed an outbreak, we need a case definition (not very sexy but critical)
Photo courtesy of Kendra Helmer, USAID
All of the information we gathered in Step 1 allows us to piece together the person, place, and time aspects of the outbreak to develop a
case definition for who we think is likely to be part of the outbreak. This is much easier if we have a lab test, but for a new disease – like AIDS or SARS – we often have to use clinical data instead (data aquired from patient observation and treatment). With a case definition we can ask if there are commonalities among all the cases. For example, are people of a certain age, race, ethnicity, location, occupation, underlying medical condition, or travel itinerary affected by this?
The case definition will create a standardized method for identifying past, present and future cases. It should include clinical information, characteristics of those with the disease, and geographic and timing information about the cluster of cases. A good case definition starts out very inclusive so as not to miss any potential cases that do not have the typically identified symptoms. As the investigation continues for an unknown disease, this definition may become more restricted to ensure we focus on persons who truly have the illness of concern. When the
SARS outbreak occurred in 2003, the case definition continually changed as we learned more about the disease, its manifestations, and transmission patterns. It started very general as a type of pneumonia, and then the definition became more specific when the transmission method and symptoms were further identified and finally a lab test was developed by CDC.
Step 3: Get a clue (often better known as throwing darts)
This is the same step we learned in grade-school for coming up with a scientific hypothesis. You must develop a question or educated guess of how something works in order to test whether you’re correct. That is essentially what we do in an investigation. We use information about those who are ill, in addition to knowledge about existing diseases, and some intuition to determine a plausible hypothesis.
This is when listening and close observation comes in handy to identify the public health misadventure that led to the outbreak: lots of dead crows surrounding the West Nile Virus outbreak in NYC, rodents associated with people with Hantavirus Pulmonary syndrome, or Hepatitis C cases that went to a specific clinic that was reusing needles.
Step 4: The leap of faith (testing the hypothesis)
While the earlier steps are often about the
art of
epidemiology, this step is all about the
science. Once we establish who is ill we need to find the factor that is causing them to become infected or poisoned.
A common way of conducting epidemiological studies (studying the patterns and causes of disease) is through a case-control study, where identical surveys or lab data are collected from those who are ill and those who are not. We then see if there are factors that are significantly (not due to chance) different in cases vs. controls. For example, in a multi-state outbreak of a rare but serious parasitic eye infection,
Ancathamoeba Keratitis, an interview was conducted with both cases and controls regarding a myriad of factors that scientists thought might be implicated in the occurrence of the disease mainly among contact lens wearers. After the questionnaires were completed, an analysis was run and sure enough, certain variables were strongly associated with an increased number of cases. One brand of contact lens solution was
recalled from the market because a significantly higher number of cases used this solution.
Step 5: Take Action!
Accurately identifying the risk factor allows us to put in place the appropriate prevention method. This public health protection step is as old as John Snow (the so-called “father” of Epidemiology) taking the handle off the
Broad Street pump in 1854 after recognizing it as the source of the cholera outbreak (and long before we had identified the cholera bacterium). In some cases it may be an “easy” fix such as recalling a product or altering manufacturing practices, but it may require a long-term outreach and education measure to promote behavior change such as the consistent use of condoms to prevent sexually transmitted diseases or smoking cessation to prevent lung cancer.
While this five-step process may appear fairly clear and logical, many investigations will throw us curve balls in the process. For example, on many international outbreaks the logistical and security challenges can present significant hurdles. I’ve been on a
monkeypox outbreak in Zaire that suddenly ended when a civil war spilled over into the area where we were working, giving us sufficient time to leave with only our specimens and data sheets. Even on domestic outbreaks, it can take a while to get everybody working together. However, no matter what the circumstances, public health officers will persevere “until we get our man.”
For more information about CDC’s current investigations or to learn more about real-life disease detectives, please visit
http://www.cdc.gov/eis/index.html.