March 30, 2021 -- Even as millions of Americans flock to be vaccinated against COVID-19, there's a pressing question on the flip side of the lines and hard-to-find appointments — how many people will decide, ultimately, not to get a vaccine and what's that going to cost us?
Even with highly effective vaccines, current levels of hesitancy in the United States could require us to continue to keep workplaces and schools closed and to keep wearing masks through at least the end of 2022 to keep the pandemic under control.
The model also projected that thousands more people — both vaccinated and unvaccinated — could die and be hospitalized over the coming months because some people remain wary of the shots.
In the United Kingdom, where hesitancy is low, more than 80% of people recently surveyed about their willingness to get a vaccine said they had already been vaccinated or planned on it. In Germany, those who say they probably or definitely would get vaccinated is up to nearly 70%. In France, around half of people have said they would get a COVID-19 vaccine. The ICL team modeled these countries as "low," "medium," and "high" hesitancy scenarios, respectively, and they compared those results with an ideal scenario — 98% of people 15 and older are vaccinated.
The ideal scenario in the model assumes that a few people can't take the vaccines because of allergies or preexisting health problems, but that everyone wants to get the vaccine.
Model Predicts Winter Surges
The model predicts several future surges in COVID-19 cases — mostly over the winter months — continuing into 2024, as people resume travel and return to offices and schools, restaurants, and entertainment venues.
Even in a scenario like that in the U.K., which has highly effective vaccines and hesitancy is low — perhaps between 10% and 20% — the model shows that daily deaths will be nearly 9 times higher at the peak of the first surge because people opt out on vaccines. Those deaths will be concentrated among people who choose not to get the shots, but they will also affect those who are vaccinated because even the best vaccines are not 100% effective.
Over the next 2 years — through the end of 2022 — the model predicts about 30% more COVID-19 deaths in a low-hesitancy country like the U.K., about 350% more deaths in a medium-hesitancy country like Germany, and about 770% more deaths in a country with high hesitancy like France.
Where does the United States fall in that mix? The ICL team didn't model the US specifically, but recent surveys suggest that America is most closely aligned with Germany in the "medium" hesitancy position.
With medium hesitancy, the model predicts 1,744 deaths from COVID-19 per million people — with about 168 per million of those deaths in vaccinated people and 1,576 in unvaccinated people, through the end of 2022.
Tricky to Model
There are some things to keep in mind about this model, though, said Rupali Limaye, PhD, director of behavioral and implementation science at the Johns Hopkins Bloomberg School of Public Health in Baltimore, Maryland.
She says hesitancy is hard to model because it clusters. Whether someone is likely to get the vaccine depends on a slew of factors, including what their friends post on Facebook, their race, age, gender, political affiliation, religion, and even their level of education.
"People that tend not to get vaccines tend to hang out with each other," she said.
For the purposes of the model, the ICL team only factored age into the hesitancy equation. They assumed that in every other way, people were equal.
Limaye said the numbers in the ICL model look reasonable, but in real life, the deaths and hospitalizations as a result of hesitancy won't be evenly spread. Some communities will be hit much harder by hesitancy than others.
Just knowing the numbers doesn't help you intervene where it's most needed.
Limaye, who wasn't involved in the ICL modeling, has worked with Johns Hopkins colleague Shaun Truelove, PhD, to try to make similar estimates for the US.
Truelove, a data modeler, said the ICL report painted a picture that was perhaps too pessimistic.
First, he said, vaccine hesitancy has declined in the US and other countries.
"More and more people are getting vaccinated. We've got more data on safety, more data on effectiveness; and we also, at the same time, have had this re-emergence of the pandemic with the variants," Truelove said. "Because of all that, willingness also goes up."
He did point out that willingness could drop if people lose confidence in some of the newer vaccines, like AstraZeneca's.
The model looks forward almost 2 years. Truelove said that's a long time when you're talking about such a fluid situation. When he's making models, he doesn't like to go any farther out than about 6 months.
Also, he said, models aren't forecasts, but they can be useful to help scale potential impacts and help guide public policy decisions.
The study authors say they wholeheartedly agree on that point.
"Getting vaccinated is an individual choice; however, this choice has social consequences," said Daniela Olivera Mesa, a PhD student at Imperial College London, in a news release. "Our work demonstrated that vaccine hesitancy can have a substantial health impact that affects both the vaccinated and unvaccinated populations. Building trust in vaccines is an important public health priority to control COVID-19."