Health experts guess wrong on flu vaccine

WASHINGTON — Seasonal influenza is spreading widely throughout the United States, with nearly half the cases caused by strains of the virus that are not directly covered by this year's flu vaccine.

Flu mortality in adults has been higher than in the past two years, but deaths in children — an important marker of severity — have been rare.

Nevertheless, this winter is likely to be one of the few times that public-health experts lose the bet they make each year when they devise the formula for the flu vaccine, something that occurs eight months before the virus starts circulating in the fall. Experts must decide on the formulation then because of the time it takes to produce mass quantities of the vaccine.

"Most years, the prediction is very good," said Joseph Bresee, an influenza epidemiologist at the Centers for Disease Control and Prevention. "In 16 of the last 19 years, we have had a well-matched vaccine."

But probably not this time.

Each year, the vaccine contains representatives of the three huge families of flu virus that are circulating. They are two main types of influenza A, H1N1 and H3N2, and influenza B.

The viruses in the vaccine are either dead or, in the case of the nasal-spray flu vaccine developed four years ago, crippled so they cannot cause illness. What they do is stimulate the body's immune system to mount a defense should the virus be encountered.

The viruses in each of these lineages are constantly changing through mutation. Inevitably, one appears that is different enough from its ancestors that a person protected against them, either through previous infection or vaccination, is not protected against the new variant.

Such an emergent virus easily finds victims because almost nobody has immunity against it.

A version of this scenario — muddied by real life — apparently happened twice this year.

A new strain of H3N2 virus was identified in Brisbane, Australia, last February, a few weeks after the components of this winter's vaccine were chosen. But it was too late to substitute "Brisbane/10" — the short version of its name — for the H3N2 strain that had been in the vaccine since the 2006-07 season, called "Wisconsin."

Even if there had been time, it was not certain the Brisbane strain would take off and spread. It has.

From the start of flu season until the beginning of February, 34 percent of flu viruses taken from patients around the country were Brisbane strains.

At the same time, a strain of influenza B called "Yamagata," which is significantly different from the "Victoria" B strain in the vaccine, was taking off.

About 16 percent of all flu samples this winter are influenza B, and of them, 93 percent are Yamagata.

Together, the Brisbane and Yamagata strains are accounting for 48 percent of all flu samples this winter — and neither is in the vaccine.