THE vaccine strategy most nations are following – of vaccinating the most vulnerable first rather than those who are likeliest to spread the coronavirus – may be the best way to save lives in the short term. But it is also the strategy with the greatest risk of driving the evolution of variants that can escape vaccine protection, according to a model developed by Julia Gog at the University of Cambridge.

“What is the absolutely worst strategy? You vaccinate all of the vulnerable and none of the ‘mixers’,” Gog said in an online presentation in February.

Gog isn’t calling for a change in vaccine strategy. But her finding reinforces the importance of keeping case numbers down as vaccines are rolled out. “We’ve got to get prevalence down, otherwise we’re [creating] a real risk of producing an escape variant,” she told New Scientist. “What you can’t do is get halfway through vaccination and allow cases to rise. That would be devastating.”

In countries with few or no cases, by contrast, the virus will have far fewer chances to evolve, so the order of vaccination doesn’t matter as far as variant evolution is concerned.

Gog’s simple model is one of the first to evaluate the effect of vaccine strategies while also considering the risk of variants arising. In it, the entire population is divided into vulnerable people with a higher chance of becoming severely ill and “mixers” who are more likely to spread the virus. The likelihood of escape variants appearing is assumed to depend on the number of vaccinated people who become infected, because every time this happens there is a small risk of such variants evolving.

Although the model is simple, Gog thinks the general conclusion that vaccinating the vulnerable first maximises so-called vaccine escape pressure is correct. “It’s bonkers to keep buying your worst enemy lottery tickets and then being surprised if they win the lottery,” she said.

But with the more transmissible B.1.1.7 variant causing a growing number of cases globally, the focus should still be on vaccinating the vulnerable, says Alessandro Vespignani at Northeastern University in Boston. “We have to deal with a variant that is ramping up now,” he says. “We have to think about averting deaths in the next couple of months, rather than down the road.”

There are still too many unknowns for models to give us clear answers on the best vaccine strategy, says Vespignani. For instance, even if variants evolve that are more likely to infect vaccinated people, they might cause only mild disease if they do.

Last year, Vespignani and his colleagues compared what would happen if vaccines were distributed fairly around the world based on population numbers rather than hoarded by rich nations until they have vaccinated their entire populations, as the likes of the US and UK plan to do. The researchers found that equitable distribution would roughly halve the number of global deaths.