Story at a glance
- The coronavirus patterns of spread show signs of overdispersion.
- Overdispersion is when something is distributed with more variability than expected.
- This may lead to superspreader events where numerous new cases arise from a few infectious people.
People interested in how researchers study the spread of disease may have recently heard the term “overdispersion.” It’s a term used by experts to describe a phenomenon in which something has dispersed with a wider range of variability than you may expect. For example, it may not spread very much at all from one individual but at a high rate from another individual.
This is where signs of overdispersion may give more insight than estimates of R0, or the reproduction number that estimates the average number of new infections from one case. If overdispersion is occurring, it could mean that a few individuals are spreading the disease to many others, but that may not be represented in R0 if most people do not spread it very much.
Experts are concerned about overdispersion in COVID-19 cases because that makes it less predictable and thus harder to stop the spread. Some early research estimates that as few as 10 percent of infected individuals may lead to 80 percent of new cases.
What is overdispersion?
In research, overdispersion is a value estimated from formulas that model the spread of disease within a population, often denoted with the letter “k.” The overdispersion value gives you an idea of how many individuals are responsible for most of the transmission of a disease. More broadly, it is a measure of variability in the data.
Models, even when based on real data, are always inaccurate to some degree but can reveal patterns that can illuminate what’s happening in the data. In a sense, models are like analogies that can only go so far to explain the “behavior” of a disease in real time. At a zoomed-out scale, like at the population or country level, it’s useful for explaining disease dynamics to describe a disease as spreading or behaving a certain way when it enters a population.
WHAT YOU NEED TO KNOW ABOUT CORONAVIRUS RIGHT NOW
But when zooming in and looking at it from on the ground, infectious diseases are not individual actors and don’t have any agency, so they cannot react or behave like people do. It’s a simplification that we make to try to better understand reality. So when experts say that a disease is overdispersed, they often mean that there are people in the population who are acting as “superspreaders” who have unusually high transmission rates that lead to a number of new cases higher than R0.
What this means for coronavirus
There have been recent examples of superspreader events, such as cases that have been linked back to a German river cruise where people were singing. There were also reports of a wedding and birthday party in Long Island that led to the infection of 56 people.
There are many assumptions here; we are assuming that some infectious people are superspreaders and that when they are in a large group of people they can spread the disease easily. One thing we don’t know yet is how to determine who those “superspreader” individuals are, and whether we can determine who they are ahead of time. Another thing we don’t know is whether there was something inherent in how their bodies reacted to the virus that made them more infectious or was their behavior more the reason for high transmission of the virus. For example, researchers could examine if they simply produce a lot more virus than other people, expel more virus when they speak or breathe and whether timing matters.
Models don’t give us all the answers and we won’t know what the full picture may look like until after the pandemic has ended. Although models may look for patterns in how disease spread occurs, underlying that is human action that leads to spread as well as the characteristics of the pathogen that can make it easily transmitted. In the real world, R and k values are meaningless. They only apply to specific circumstances that are outlined in the model they pertain to. The most accurate model is the real world, and it’s too complex to fully capture with math.
With COVID-19, we can see after the fact that the coronavirus may have dispersed more than expected in some situations. But we are the ones who set up the conditions for the dispersion to happen. What we can control about a disease is how we let it access our bodies as transport and if we introduce it to our friends.
In order to do the math, it’s necessary to lump people together into case counts, other numbers and networks of contacts. And it’s easier to figure out after a superspreader event that it was an event and trace it back to who was infectious at the time. But it’s important to remember that we are the actors in this situation.
For up-to-date information about COVID-19, check the websites of the Centers for Disease Control and Prevention and the World Health Organization. For updated global case counts, check this page maintained by Johns Hopkins University or the COVID Tracking Project.
You can follow Chia-Yi Hou on Twitter.
BREAKING NEWS ABOUT THE CORONAVIRUS PANDEMIC IN AMERICA