Changing people鈥檚 behavior until a vaccine could be developed prevented roughly 800,000 COVID-19 deaths in the U.S., far more than many scientists predicted was possible, according to new CU Boulder and UCLA research.
But the authors stress that interventions like lockdowns and school closures came at great cost鈥攐ne that could be reduced in future pandemics if the country had a better infrastructure for gathering public health data.
The paper was published in the spring edition of .
鈥淥ur work shows that behavior change can be a powerful force for slowing the spread of a dangerous and infectious respiratory disease for a long time,鈥� said co-author Stephen Kissler, an assistant professor of computer science at CU Boulder. 鈥淏ut with COVID-19, it came at a tremendous economic, social and human cost.鈥�
As an epidemiologist at the beginning of the pandemic, I certainly did not think that we could have kept it up for as long as we needed to in order to avert as many deaths as we did.鈥�
鈥揝tephen Kissler
How vaccines and behavior change听work听together
For the study, Kissler, a mathematical epidemiologist, teamed up with Andrew Atkeson, a professor of economics at UCLA, to answer a fundamental but unanswered question:
How many deaths from COVID-19 were prevented by behavioral interventions like masking and social distancing, combined with vaccines?
The authors gathered national serology data from blood samples to estimate how many people had been infected or vaccinated at various points from February 2020 to February 2024 and mortality data from the Centers for Disease Control.
Then, they used computer models to mathematically recreate the pandemic as it happened, factoring in the role of behavioral changes.
By tinkering with the model inputs to simulate different scenarios they were able to ask questions like: How many people would have died if no one had done things like wear masks or practice social distancing? Or how many people would have died if the vaccines never came?
The study found that vaccines and behavioral changes were inextricably linked.
鈥淲ithout vaccines, behavior alone would have postponed infections, but in the end, nearly everyone would have been infected and subject to a high infection fatality rate from that first infection,鈥� they wrote. 鈥淲ithout a behavioral response, vaccines would have come too late to save lives.鈥�
They found that, thanks largely to behavioral changes, 68% of Americans were able to get vaccinated before ever being infected.
Had they gotten COVID for the first time without being vaccinated, their risk of dying would have been as much as four times higher.
Absent the two-pronged strategy, 1.98 million people in the U.S. would have died of COVID in the last four years vs. the 1.18 million that did succumb to the disease. The pandemic would have ripped through the country in the first eight months, taking 60,000 lives per day at its peak.
The authors estimate that the slowdown in vaccine uptake during the second half of 2021, as the virus mutated, cost an additional 273,000 preventable deaths from the Delta and Omicron variants.
The power of changing behavior
Kissler and Atkeson were surprised at just how big an impact behavior changes had, as pre-pandemic studies forecast their impact would be minimal.
One study found that behavioral interventions reduced transmission by less than 50% during the 1918 influenza epidemic, and people only changed their behavior for about 15 weeks. Another study forecast that if a pandemic hit the U.S., and vaccines came more than 120 days after the first case, they would essentially be useless because most people would have already been infected.
In reality, only one in five Americans (and a much smaller proportion of people over age 65) had been infected by the time COVID-19 vaccines were made available, 11 months after the first case.
By the numbers
- 68%听of听Americans got vaccinated before being infected
- 1.98 million听would have died without vaccines or behavior changes
- 273,000 preventable deaths occured due to declining vaccine uptake in late 2021
- 60,000 would have died per day at the pandemic peak without behavioral changes
鈥淎s an epidemiologist at the beginning of the pandemic, I certainly did not think that we could have kept it up for as long as we needed to in order to avert as many deaths as we did,鈥� said Kissler.
That said, the authors worry that if another pandemic were to emerge, Americans would be less willing to stay home.
鈥淢y concern is that the next pandemic will be deadlier, but people will ignore it, because they will say, 鈥極h, we overdid it during COVID,鈥� 鈥� said Atkeson.
He and Kissler acknowledge that some behavioral changes made during COVID, such as requiring masks on outdoor hiking trails, ended up being unnecessary. But knowledge about transmission was still lacking.
They say U.S. policy makers should prioritize developing a more centralized national infrastructure for gathering data on how people move around and interact to spread viruses and what behaviors work best to contain them.
鈥淲e stumbled through (COVID) without a plan. We were basically flying blind,鈥� said Atkeson.
Improvements听already in the works
The researchers noted that the CDC Center for Forecasting and Outbreak Analytics is already working to develop a system similar to the National Weather Service to provide forecasts on the country鈥檚 epidemiological conditions. 听
They point to the recent Mpox (formerly known as monkeypox) outbreak, which was contained relatively swiftly with targeted behavioral interventions and vaccines, as a great success.
鈥淭he behavioral changes needed to prevent disease spread can look very different from one community to the next and one virus to the next,鈥� said Kissler. 鈥淚f we have more granular, timely data to work with we can do a much better job of targeting interventions so we don鈥檛 have to just stop everything.鈥�
