It’s never an easy business to predict which flu viruses will make people sick the following winter. And there’s reason to believe two of the four choices made last winter for this upcoming season’s vaccine could be off the mark.

Twice a year influenza experts meet at the World Health Organization to pore over surveillance data provided by countries around the world to try to predict which strains are becoming the most dominant. The Northern Hemisphere strain selection meeting is held in late February; the Southern Hemisphere meeting occurs in late September.

The selections that officials made last week for the next Southern Hemisphere vaccine suggest that two of four viruses in the Northern Hemisphere vaccine that doctors and pharmacies are now pressing people to get may not be optimally protective this winter. Those two are influenza A/H3N2 and the influenza B/Victoria virus.

The strain selection committee concluded the H3N2 and B/Victoria viruses needed to be updated because the ones used in the Northern Hemisphere vaccine didn’t match the strains of those viruses that are now dominant. Influenza epidemiologist Dr. Danuta Skowronski described the significance of those two changes in one word: “mismatch.”

“I think the vaccine strain selections by the WHO committee are obviously important for the Southern Hemisphere but they’re also signals to us because they’re basing their decisions on what they see current predominating on the global level,” said Skowronski, who is with the British Columbia Center for Disease Control in Vancouver.

Flu vaccine is a four-in-one or a three-in-one shot that protects against both influenza A viruses—H3N2 and H1N1—and either both or one of the influenza B viruses, B/Victoria and B/Yamagata. Most flu vaccine is made with killed viruses, and most vaccine used in the United States is quadrivalent—four-in-one.

There was great uncertainty around which version of H3N2 to choose for the Northern Hemisphere vaccine when the committee met last February—there was a lot of variation between the strain the U.S. was seeing and the H3N2 viruses sickening people in Canada and Europe. There was so much uncertainty, in fact, that the committee delayed making the choice of the H3N2 strain for a month to try to get a clearer picture.

In the end, the committee selected a version of the virus that was causing a wave of late season illness in the United States. (Canada also had a late season surge of H3N2 activity, but caused by a different version of the virus.)

“That H3N2 wave was late and it was evolving at the time that they met in February,” Skowronski said of the strain selection committee. “And there was a diverse mix of H3 viruses. And it wasn’t clear to them, I guess, [which strain] … would emerge the clear winner.”

It appears the virus that was ultimately selected is not the H3N2 that dominated during the Southern Hemisphere’s winter 2019 season.

Scott Hensley, an associate professor of microbiology at the University of Pennsylvania, said the variant of H3N2 viruses that just swept through the Southern Hemisphere is more likely to be the main cause of H3N2 infections for the Northern Hemisphere this winter than was the case in the U.S. late last winter and into the early spring.

But that version of H3N2 is difficult to grow in eggs, which is the way the vast majority of flu vaccines is made, he noted, suggesting that fact may have influenced the thinking

In recent years the H3N2 component has generally been the least effective part of the vaccine. If H3N2 viruses predominate this coming flu season, a vaccine mismatch could add to the severity of the season. But if those viruses play a smaller role this winter, the impact of a mismatch will be less significant, making it hard to predict if this choice is going to turn out to be a problem.

Flu circulation “remains difficult to predict and flu viruses are constantly breaking rules that we try to establish for them,” Hensley said, adding that flu vaccines “often protect against severe disease even when … mismatched.”

The selection of a new B/Victoria virus for the Southern Hemisphere 2020 shot also concerns Skowronski. There was almost no influenza B activity in the 2018-2019 flu season and it’s been several years since B/Victoria viruses have caused much illness. As a result, there may not be a lot of immunity to those viruses in the population, she said.

B/Victoria flu viruses are especially hard on children, Skowronski said.

Given the possibility that a couple of the components of the vaccine might not be well-matched to circulating flu viruses, Skowronski said it will be important for doctors to realize vaccinated patients may still contract influenza. For those who are at high risk of developing severe illness, rapid treatment with flu antiviral drugs should be considered.

She also suggested older people or people who have underlying health problems—in other words, those who are likely to develop a severe case of flu if they contract the virus—should take steps to avoid being around sick people.

The sliver of good news: The officials meeting at the WHO last week concluded that the H1N1 and the flu B/Yamagata components of the Southern Hemisphere vaccine didn’t need to change, suggesting they are representative of the strains of those viruses we’re likely to encounter this winter.

Republished with permission from STAT. This article originally appeared on September 30, 2019