TORONTO -- This year's flu vaccine provided little protection to people 65 and older who got a shot, new data released by the U.S. Centers for Disease Control reveal.

When it came to the influenza A strain H3N2 -- the main virus circulating in North America this winter -- the study found the vaccine cut an older adult's risk of getting sick enough to need to see a doctor by a mere nine per cent.

"Clearly it's much lower than we had hoped for," Dr. Joe Bresee, a flu expert with the CDC, said in an interview.

When all three strains in the vaccine were taken together, the study suggested the shot was about 27 per cent effective in preventing seniors from becoming ill enough to need to see a doctor. For all age groups taken together, the effectiveness estimate was 56 per cent.

Bresee said seniors need to know that even if they got a flu shot this year, they could still get sick and should see a doctor about taking antiviral drugs if they do. H3N2 viruses hit the elderly hard and there has been a big spike in hospitalizations among the elderly this winter.

"We wanted to make sure that people knew that -- especially if you're elderly to know that -- because we still want to communicate the fact that if you do get sick, get antivirals. And don't assume just because you got the vaccine that that flu-like illness you've got can't be flu," Bresee said.

While the results are disturbing, they aren't entirely surprising. In recent years, a number of studies have shown that flu vaccine doesn't work as well as had been previously believed, and provides particularly poor protection in older adults, who are among those hardest hit by flu.

"Clearly the influenza vaccine world is in a state of rather significant transition -- one from where we overstated the effectiveness of the vaccine and now one where we're confronted head on with the scientific data which says that the impact is at best moderate for most age groups and in some age groups significantly below that which had been previously stated," said Dr. Michael Osterholm, director of the Center for Infectious Diseases Research and Policy at the University of Minnesota.

In the past year or so Osterholm and colleagues have published a scientific analysis and a major report on flu vaccine, concluding that new and better flu vaccines are needed because the existing products aren't getting the job done.

This latest CDC report, for instance, showed that young adults -- those between ages 17 and 49 -- who got flu vaccine this year saw their risk of H3N2 infection cut by only 46 per cent. This is in the age group in which vaccines typically work well and in a year in which the H3N2 virus in the shot is closely matched to the ones making people sick.

"I believe that 46 per cent protection is better than zero. But it is far short of what we want and what we ultimately believe we can have with better vaccines," Osterholm said.

It is important to note that the study measured the ability of the vaccine to protect against what the flu research world calls "medically attended flu." That means the study asked: Were people who were vaccinated this year less likely to get sick enough with the flu to need to see a doctor? And if they were less likely, how big was the protective effect?

This type of study may actually overestimate the effectiveness of the vaccine, some experts say. For one thing, it doesn't capture a picture of the percentage of people who got a shot and got the flu but did not go to a doctor. Most people who get the flu simply ride out the illness in bed.

"People who are infected but do not come to medical detection -- that still would be a vaccine failure, but we're not counting them here," Osterholm said.

As well, it doesn't account for people who become so sick they go directly to hospital, which would not be uncommon in elderly people with influenza.

In fact, assessing vaccine effectiveness in seniors can be tough for that reason, said Dr. Danuta Skowronski, a flu expert with the British Columbia Centre for Disease Control who leads an annual Canadian effort to assess the protection offered by the flu shot here.

This year an interim estimate generated by Skowronski's surveillance network showed the vaccine's H3N2 component cut the risk of infection in Canada by 45 per cent. The CDC study put the overall figure in the U.S. for H3N2 at 47 per cent, and a European study came in at 42 per cent.

Skowronski said such studies have repeatedly shown sub-optimal effectiveness estimates and she suggested it is time to try to figure out what is behind the lack of impact.

While a number of scientific teams are trying to devise a universal flu vaccine -- one that wouldn't need to be changed every year -- such a breakthrough is probably years away. In the meantime, Skowronski said, finding ways to use existing vaccines to better effect makes sense.

"Maybe for next season there's not much we can do," she said.

"But for the subsequent seasons if we focus now and get some studies going, we might be able to investigate why the VE (vaccine effectiveness), even in healthy young people, is lower than the conventional expectation for vaccines given to the population."