The study of outbreaks is, in many ways, the study of people.

After all, it’s people who get sick, who transmit a disease, and who can change the way they live in response to outbreaks. So, it makes sense that many of the scientific tools we use to study outbreaks are centered around people and their relationship to the disease.

The better we understand that relationship, the quicker we can intervene to cut ourselves off from the disease and stop its spread. In outbreak science, the discipline that helps us do this is called epidemiology. In this episode, we’ll learn what epidemiology is, how it helps us track diseases, and how it can in the right circumstances, give us the info we need to stop outbreaks.

I’m Pardis Sabeti, and this is Crash Course Outbreak Science! [Theme Music]. At first, it seems obvious that epidemi-ology would be the study of epidemics. The clue’s in the name, right?

While some epidemiologists do study epidemics and outbreaks, the field is actually much broader. In general, epidemiology is the study of the patterns of disease and health conditions within populations, and their causes. In particular, the patterns we’re after are how often a disease occurs in different groups.

For our purposes, we’re going to focus on infectious disease epidemiology, since that’s the kind most relevant for outbreaks, but some epidemiologists also study non-communicable diseases, which are the kind that aren’t infectious, like diabetes, or health conditions like asthma. After all, sometimes these disciplines intersect. For example, a nutritional epidemiologist studying malnutrition may find that infectious bacteria in contaminated water can change a child’s gut bacteria, making them more prone to malnutrition.

Whether the diseases are infectious or not, epidemiologists learn about them by making careful observations about populations, and different groups within them. The evidence from those observations provides clues about the disease and its impacts. We can then make hypotheses, to explain what a pathogen’s origin is or how it spreads.

Like any scientific hypothesis, we make specific predictions and use data to help confirm or rule out our suspicions. Basically, being an epidemiologist can be like being a disease detective. And like detectives, epidemiologists have their own tools of the trade, starting with standard terminology.

These concepts are as vital to epidemiologists as microscopes and test tubes are to other infectious disease scientists. Standard terminology means that whatever the circumstances of an outbreak and wherever in the world they happen, epidemiologists have a common language for talking about them. They make it clear that if we say “a cluster of exposed people”, we mean a group of people who might have been infected and not a party at a nudist beach.

One thing we need to keep consistent in different contexts is what a disease actually looks like in infected people. Epidemiologists call these descriptions cases. A case is simply a person we can identify as infected with the disease that’s being studied.

As we saw when we talked about clinical diagnostics, determining whether someone is a case or not is tricky, so cases need their own definitions. As always, science demands clarity! Case definitions are characteristics that could indicate whether someone has a disease, like its typical symptoms, signs, or clinical test results.

For new diseases, case definitions tend to start out broad and exploratory, and then become more specific as we learn more about them. Outbreaks, by definition, are when the number of cases of a disease exceed what we’d expect in a particular group of people. So naturally, in epidemiology we pay close attention to that number, and we do so in two ways.

We consider the number of existing cases at a single point in time, and the number of new cases that develop over a specific period of time. We call these prevalence and incidence. Both often consider the number of cases as a proportion of the population.

For example, say we were studying the bacterial disease tuberculosis, or TB, in 2012 for the whole world. After analysing the case numbers, we’d say that the prevalence of the disease in 2012 was 169 cases per hundred thousand. That means that for every hundred thousand people on Earth, we could expect about 169 of them to have TB at any given point in the year.

As well as the number of people who have a disease, we’d also want to know the rate at which new cases appear, which is what we call incidence. And that same year, the global incidence of TB was 122 cases per hundred thousand per year, meaning that in 2012, we’d expect 122 people who didn’t have TB to develop it over the course of a year. So, prevalence is the fraction of cases in a population, whether they’re new or not, at a single moment in time.

Meanwhile, we can think of incidence as the number of new cases over a given period of time, or the rate at which new cases appear. Both prevalence and incidence are key indicators for whether an infectious disease outbreak is happening. During an outbreak the incidence will be much higher than usual for a particular group, and whether it’s increasing or decreasing can also tell us whether an outbreak is getting worse, or better.

If the prevalence is high, sometimes it could tell us if a given region could be at risk of an outbreak because a lot of people have a disease. As we’ve seen in previous episodes, it’s also important to know who is at risk. Sometimes, epidemiologists are pretty on the nose when it comes to naming things, so the population at risk during an outbreak is called...

The Population at Risk. We can also describe this group as susceptible, and in general, it describes who could become infected. For a disease like Orchitis, which involves an inflammation of the testicles, as you’d expect, only people with testicles are susceptible and would be in the population at risk.

This is important when considering the incidence and prevalence. Say we calculated the incidence of orchitis in the whole population of the UK and got 12 cases per year per ten thousand people. Considering everyone in the UK might give us a misleading impression because we know about half of that group have no risk of developing orchitis.

Instead, we’d say the incidence for the susceptible population is about twice as high, since only around half of the population is susceptible. Defining the right group of susceptible people is crucial! But doing that isn’t always so straightforward.

For instance, if a person becomes immune to a disease after already having it, technically they’re not susceptible anymore. However, unless we test everyone for antibodies, it’s hard to confirm who is or isn’t immune. Measuring incidence and prevalence is the bread and butter of descriptive epidemiology, which represents the state of a disease or other epidemiological issue through data about it.

Descriptive epidemiology gives us a starting point to understanding the extent of an outbreak. Analytical epidemiology can help us figure out the potential causes of an outbreak. In analytical epidemiology, we compare groups of people who have a disease to similar groups of people who don’t have it and look for differences between the two groups that might explain what causes the disease.

That’s the approach researchers used in 1975, in the rural town of Old Lyme, Connecticut, just off the Connecticut river. They were investigating what we now call Lyme disease. It started when two mothers in the area did some independent sleuthing after their own children were diagnosed with juvenile arthritis, which involved swelling and crippling pain in their joints.

Both moms separately noticed that there were way more kids in their neighborhood that had gotten similar diagnoses in recent years than was likely by pure chance. In other words, the prevalence and incidence of the disease seemed higher than expected! They got in touch with the state health department, who referred them to researchers at Yale University.

Let’s go to the Thought Bubble. First, the researchers put together a case definition. They studied clinical diagnoses in previous years that seemed out of the ordinary, and defined Lyme disease cases as the presence of rashes on the skin, recurring bouts of arthritis in children, or unexplained arthritis in an adult.

They also gathered the details about each case, like their address, age, sex, race, and workplace. Immediately, some patterns stood out. Most of the cases seemed to happen in June and July, the peak of summer.

There were four times as many cases on the east side of the river than on the west side, even though the west side had a bigger population! Lastly, all the cases seemed to be on the edge of the main town, near the woods. That led the researchers to suspect that the disease was being passed on by bites from tiny bugs called ticks.

They hypothesized that one side of the river had more ticks than the other. That hypothesis explained the other patterns, too: kids have long, free summers where they can wander around outdoors, which means kids have more chances of being bitten by ticks than adults. To test this hypothesis, they gathered thirty two people with Lyme disease and for each person, identified someone who lived in the same neighborhood and was the same age and sex who hadn’t caught Lyme disease.

Then, researchers interviewed both groups about their outdoor activities that summer. Sure enough, the Lyme disease group reported experiencing a much higher level of tick bites than the other group! Thanks Thought Bubble!

This helped narrow down the cause of Lyme disease to the ticks. A few years later, the bacteria responsible for Lyme disease was isolated and identified inside tick bodies, which officially solved the puzzle. What the researchers at Yale performed was a case-control study, an analytic epidemiology method which uses two groups of people: those who do have a disease: cases, and those who don’t: controls.

We make sure they’re otherwise similar by matching cases and controls based on characteristics like age, sex, and race. The differences we find between the two groups, like the presence of tick bites in the Lyme disease study, that might explain their different disease rates are called exposures. Instead of looking at a group of people who had already caught the disease, we could perform an analysis by finding a group of people and see who goes on to catch a disease, and then study their exposures.

This kind of analysis, where we first choose a group of people, and then follow them over some period of time, is called a cohort study. Once again, the idea is to identify the exposures that are correlated with the rates at which people catch a disease. Case control and cohort studies are two of the ways epidemiologists can help identify the cause of the outbreak.

The basic idea is that if a particular exposure seems to be associated with a much higher probability of catching the disease, there’s a reasonable chance that it could be what’s responsible for becoming infected. Determining the cause of an outbreak is one of the main goals of epidemiology, since it allows us to come up with interventions, like cutting off contaminated water supplies. In the case of Lyme disease, people are now advised to avoid the tall grass and bushes where ticks lurk, wear insect repellent and check their bodies for ticks after a long hike so the ticks can be found and removed before the bacteria transfers from the tick to the person.

Another way epidemiologists come up with interventions is by working out how a disease spreads. This can be from a contaminated source of food or water, vectors like ticks, or even from a person to their fetus during pregnancy. Often, it tends to happen when an infectious person meets a susceptible person, allowing the pathogen to transmit between them.

One of the ways epidemiologists get data on this kind of transmission is through contact tracing. Contact tracing involves recording who an infected person interacted with and the places they’ve been, usually by interviewing people who are or could have been infected. During the early stages of the Covid-19 pandemic, contact tracing identified that loads of people became infected when in the same poorly ventilated, indoor spaces.

This was a key piece of evidence that the virus spread largely through droplets, and helped epidemiologists decide to recommend wearing masks as an intervention. Recommendations like these, which stop people from getting sick and even save lives, come from the tools of epidemiology. Studying disease prevalence and incidence helps us work out where outbreaks are happening.

When they do happen, studying the patterns of infections and their associated risk factors gives us the tools we need to come up with effective interventions. When making recommendations for what changes we need to stop an outbreak, we want to be able to predict what impact those changes will have, and how the outbreak will develop if left unchecked. That’s where modeling can help us, and it’s what we’ll be looking at next time.

We at Crash Course and our partners Operation Outbreak and the Sabeti Lab at the Broad Institute at MIT and Harvard want to acknowledge the Indigenous people native to the land we live and work on, and their traditional and ongoing relationship with this land. We encourage you to learn about the history of the place you call home through resources like native-land.ca and by engaging with your local Indigenous and Aboriginal nations through the websites and resources they provide. Thanks for watching this episode of Crash Course Outbreak Science, which was produced by Complexly in partnership with Operation Outbreak and the Sabeti Lab at the Broad Institute of MIT and Harvard— with generous support from the Gordon and Betty Moore Foundation.

If you want to help keep Crash Course free for everyone, forever, you can join our community on Patreon.