Researchers warn that the next major pandemic may not be too far away. In a new paper out Monday, they estimate that a pandemic as deadly as covid-19 is expected to arrive within the next six decades, while a pandemic on the scale of the 1918 Spanish flu could be expected to recur every 400 years. They also argue that the annual probability of these extreme events may increase over time, given that emerging and re-emerging diseases have become more common in recent decades.
The world is currently embroiled in year two of the covid-19 pandemic, which has so far killed at least 4.4 million people and likely far more. Outside of HIV/AIDS, which has killed at least 36 million since its emergence in the 1980s but isn’t always considered a pandemic, covid-19 is the deadliest pandemic since the Spanish flu, which killed somewhere between 20 million to 100 million people (most estimates hover around 50 million).
Pandemics aren’t as rare as some might think, though. The last pandemic before covid-19 was only a decade earlier—the 2009 swine flu—and a pandemic has occurred on average every 20 years in the past century. But researchers at Duke University and elsewhere say there hasn’t been much statistical work done to estimate the probability of these major outbreaks of disease—a knowledge gap they hoped to address in their new paper, published Monday in the journal PNAS.
“First, I should say that we are not making predictions about the future. We are characterizing the likelihood of large epidemics occurring based on historical data,” study author William Pan, associate professor of global environmental health at Duke University, told Gizmodo in an email.
The team looked at major recorded epidemics of plague, cholera, novel influenza strains, and other pathogens, stretching back 350 years, to come up with their estimates. They especially focused on outbreaks of emerging or re-emerging diseases that killed at least 10,000 people. They smushed together epidemics that occurred in different places around the same time, such as concurrent plague outbreaks in the 17th and 18th centuries. And they also excluded outbreaks of diseases after they became manageable through medicines like antibiotics or vaccines, as well as currently ongoing epidemics like HIV/AIDS, malaria, and covid-19 (in practice, this meant no epidemics after 1945 were included in their primary analysis).
Overall, there was plenty of variation in how frequently major epidemics occurred over the years, though they did decline over time. But the team says their novel method of statistical modeling, previously used to estimate the risk of extreme climate events like floods, can pin down a rough probability of how likely pandemics of varying scale are to occur on an annual basis. For something like covid-19, they estimate the likelihood of a covid-like event happening in any given year is now around 2%, meaning that it would be expected to happen sometime in the next 59 years. Importantly, this doesn’t mean that the next covid is 59 years away, only that, within 59 years, we should see it happen again. For something like the Spanish flu, they say something similar would be expected to happen every 400 years, give or take a few decades.
According to other research, Pan said, the likelihood of a pandemic happening should drop substantially the more severe it is (so a Spanish flu event should be exceedingly rare). But his team’s work seems to show that this likelihood doesn’t decrease that fast relative to severity after all, so even catastrophic pandemics can happen with alarming regularity.
The work underpinning the team’s math is based on assumptions, as all models are. So at the end of the day, these numbers are just estimates. But the authors argue that their baseline forecasts may undersell the problem, if anything. They point out that small-scale outbreaks of emerging and re-emerging diseases have increased in recent decades. And when they accounted for this increase in their modeling, they concluded that the annual probability of extreme epidemics could rise by threefold in the coming decades. That could mean that a pandemic on the scale of the Spanish flu would arise every 127 years on average, not 400.
Though the team didn’t look into why these outbreaks are becoming more common, Pan cites other research showing that environmental changes have led to more contact between humans and animals that can carry these exotic germs. Poverty, poor sanitation, and lack of good health care systems can then allow diseases to keep spreading, as can a lack of cooperation between countries in monitoring these threats.
The basic message here is that large-scale pandemics are relatively likely to happen, the authors say. And because of that, we should be doing more to prevent them or to blunt their impact when they do arrive. “We obviously show the potential threat of global pandemics, but the real implication here is how do we invest more effectively in global health and pandemic preparedness?” Pan said.
Pan said scientists should be studying the ongoing global response to covid-19, in order to determine the approaches that should be followed or avoided in the future, while recognizing that some interventions may not work for all future pandemic threats (masks may not be needed, for instance, depending on how the hypothetical pandemic spreads). We also need more teamwork between countries, ideally aided by existing structures like the United Nations and World Health Organization.
“But it’s not just about emergency response—it is also about lifting the poorest and most vulnerable countries up—we need to make sure we achieve the Sustainable Development Goals,” Pan said, referring to the UN’s blueprint for reducing the impact of extreme poverty, inequality, and other major threats like climate change by 2030.
Not included in the team’s paper is their math on the likelihood of the Big One—a pandemic deadly enough to end humanity. But they did create such an estimate: According to their model, a pandemic that could kill all humans is likely within the next 12,000 years. On the bright side, there are plenty of other things that could kill us all before then, like an asteroid, artificial superintelligence, or nuclear holocaust.