For weeks now, health and community leaders in Colorado have engaged in a sometimes tense debate about how best to allocate scarce supplies of the coronavirus vaccine in order to save the most lives.
One argument goes that the state should prioritize those most at risk of death – people who are older and people who have high-risk medical conditions. Another argues that the vaccine should be used to target outbreak-prone groups, such as those living in prisons or college dorms, in order to reduce the spread of the virus and provide communitywide protection.
Both strategies can find support from previous, successful vaccination campaigns for other infectious diseases. What’s been missing from the debate has been research to help figure out the best strategy for the current pandemic.
“I don’t know that we have good insight that says this is how you control a pandemic caused by SARS-CoV-2 virus,” said Dr. Eric France, chief medical officer of the Colorado Department of Public Health and Environment, using the scientific name for the coronavirus that causes COVID-19.
But, it turns out, a possible answer has been sitting for months in a study published by University of Colorado Boulder researchers.
In September, CU professor Daniel Larremore, doctoral student Kate Bubar and five other researchers published a paper, “Model-informed COVID-19 Vaccine Prioritization Strategies by Age and Serostatus.” The paper was published on the website medRxiv, where scholars often post work before it has been peer-reviewed and accepted by a more traditional journal – a practice that has become common amid the express-lane speed of pandemic science.
Larremore is not a typical public health expert. He’s a professor of computer science, who also works with CU’s BioFrontiers Institute. He and his team created a computer model to estimate what impact different coronavirus vaccine allocation strategies would have on deaths.
And the conclusion?
“It’s clear that the recommendation should be that those most vulnerable people should get vaccinated first,” Larremore said.
This is exactly what the state now plans to do, though how Colorado ended up taking the approach that Larremore’s research says it should is somewhat fortuitous. Larremore said no one from CDPHE called to ask him about the study, and, when asked, France did not cite any particular studies that state health officials looked at when drafting the allocation plan.
Larremore’s research was used to help formulate guidance from the National Academies of Science, Engineering and Medicine, and that guidance informed Colorado’s planning. But, when Colorado released its initial allocation plan in October, it gave high priority to people living in outbreak-prone settings.
The state later – controversially – stripped those groups out of the plan. Gov. Jared Polis said the state, while also vaccinating health care workers and first responders, would focus on vaccinating those most at risk of death. France echoed that, saying Colorado decided “to take the traditional approaches.”
“Identify the populations that are most likely to die or have severe illness or be hospitalized and make sure they’re protected,” he said.
Larremore said his team’s research began in late spring in anticipation of exactly the debate taking place now. Both of the approaches – protect vulnerable individuals vs. reduce community spread; the direct approach vs. the indirect approach – are common-sense ideas. Vaccinating children for the flu has worked well to prevent grandparents from getting sick, for instance, because kids are big flu spreaders and the flu vaccine often is less effective in older adults.
But they are also in conflict with one another, especially when vaccine is scarce. So that’s where the modeling projections can help, he said.
Larremore’s team looked specifically at the second phase of vaccine rollout, assuming that health care workers would go first. And it modeled several different scenarios. What if the vaccine is more effective in young people than in adults? What if community spread of the virus is rampaging – or not? What if the rollout is slow or fast?
That led them to realize there is no one right answer except: It depends.
Countries like New Zealand or Taiwan that have the virus well under control have more options, Larremore said. With so little spread, they can use limited supplies to vaccinate those most likely to be spreaders if the virus does get reintroduced into the country. Countries that have enough vaccine to roll out their program rapidly also have more options.
Larremore said he thinks of it like a race between the virus and the vaccine.
“Two things competing for space in people’s bodies,” he said. “Whichever one gets there first wins that person.”
In a situation like we have now – fast viral spread and not a lot of doses to go around, though the vaccines appear to be highly effective across all age groups – that makes it imperative for health officials to win the race to the most vulnerable.
Larremore’s team found, in most scenarios, that prioritizing older adults leads to the greatest reduction in deaths and saves the most years of life.
“I was a little bit surprised at how many different scenarios all pointed to the conclusion that we should prioritize the most vulnerable,” he said. “… Part of the reason is this is a disease that is so much more deadly for older adults and the most vulnerable.”
