Why the next pandemic could be stopped by math, not borders

What do LinkedIn, the global airline network and the human brain have in common? They are all complex systems built from networks of connections.

At the University of Virginia, data science professor YY Ahn researches these connected systems to learn how understanding one system can help us understand them all. Mapping and analyzing these connections is part of the research area called network science, which is becoming foundational for disciplines like artificial intelligence, epidemiology, systems biology and more.

We sat down with Ahn to learn more about how network science could change the way we prepare for—or maybe even avoid—the next pandemic.

You’ve mentioned that we keep dismantling effective systems that stop the spread of infectious diseases. Why is that?

In July 2025, the CDC made reporting of Cyclospora and five other pathogens optional for FoodNet, its foodborne-illness surveillance network. Although we can’t isolate this decision as the sole reason for the cyclospora outbreak we are seeing, this is a common pattern we see over and over again. You create a prevention unit after the panic of an outbreak, and then, over time, more and more people see it as a waste. Then the cycle begins again.

When we hear about an outbreak, the first reaction is often to close borders and “don’t let people in.” What the public is far less enthusiastic about is sending people toward the source of the outbreak to contain it there and learn how it spreads.

Why is going toward the source of a disease outbreak so important?

It’s critical to go to the source because that’s where the exponential growth is unfolding and where we have the most leverage to control it. Once it hits the airline hubs, it becomes really difficult to control. Epidemiologists, especially network epidemiologists, understand this, but it’s not very intuitive or commonly known to most people.

You never really hear about all the success stories where outbreaks were contained very early. You’ve already controlled it, so it does not become huge news. That’s the dilemma for disease prevention called the “preparedness paradox.” The more successful you are at prevention, the less people and the system may see you as necessary.

What if the disease has already started spreading?

One of the best-known findings in network science is that once a disease starts spreading through the airline network, it’s practically impossible to curb it by closing airports.

Even if you shut down 90% of air traffic, you only delay the arrival somewhere else a little bit. During the 2009 H1N1 pandemic, travel to and from Mexico dropped by around 40%, and that bought maybe just a few days. Because airports are wired and optimized to make it as easy as possible to move people from one place to another in the world, they form a network that is also extremely good at spreading any disease from one corner of the world to everywhere in the world.

The general rule is that we need to control it at the source, as early as possible, before it explodes. SARS in 2003 started in China and seeded an outbreak as far away as Toronto, but we could still control it through contact tracing because it was less contagious and more severe. People got visibly, seriously ill, so they could be identified and isolated before they infected many others. But COVID-19 was too contagious, with lots of asymptomatic infections.

We also published a paper about contact tracing during the pandemic. With network theory, you can identify two very different directions of contact tracing. If you have a patient with a contagious disease, you can go forward (trying to reach the person that the patient spread the disease to) and backward (trying to reach the person from whom the patient got the disease).

We found that if you follow a link backward from a patient to whoever infected them, you land on a heavily connected person far more often than chance would suggest—which is usually where the superspreading happened. This was a combination of network theory, empirical data and simulations.

How does this tie into the work you’ve done on misinformation and how it spreads?

Recently, I was part of a paper that modeled how misinformation may have affected the course of the COVID pandemic. The idea was that those who were misinformed may change their behavior, particularly their attitude toward vaccination, and that can lead to greater spread.

In another paper, we also looked at how COVID was spreading before it was really identified in China. This was in November and December 2019, and we examined how rumors of the new disease prompted people’s behavior, like buying masks. People were buying masks based on rumors before the government announced there was this new disease spreading.

Although we do not have full social network data, you can still glean how that information spread in social networks. It turned out that at the very beginning of this “cryptic” transmission phase, people who purchased masks before the official announcement was made were hospital staff, more educated people and people who learned about this mysterious disease through their family and friend ties. This shows both the power of social networks in spreading information and the inequality that played out in real time during the pandemic.

Who’s behind this story?

Robert Egan

Bachelor’s in mathematical biology, Master’s in creative writing. Well-traveled with unique perspectives on science and language.

Full profile →