So Does The Flu Vaccine Work Or Not?

A paper that came out Wednesday on influenza vaccine efficacy has generated a new round of speculation about what is probably the hardest sell in the vaccine business. There’s a lot to complain about with our current flu vaccines: everyone needs a new shot every year, vaccine makers don’t always guess right about which strains of flu will be circulating that season, and plenty of people can tell stories about how they got the shot and still got sick.

Now, in the middle of flu vaccine season, we get a new peer-reviewed report – and of course an accompanying press release – that seems to bring more bad news:

“Evidence for consistent high-level protection is elusive for the present generation of vaccines, especially in individuals at risk of medical complications or those aged 65 years or older. The ongoing health burden caused by seasonal influenza and the potential global effect of a severe pandemic suggests an urgent need for a new generation of more highly effective and cross-protective vaccines that can be manufactured rapidly”, explains Michael Osterholm from the University of Minnesota, USA, lead author of the study.

The news coverage was a mixed bag, ranging from predictable rantings from the antivaccination nuts (no links – you can find them yourself) to balanced, nuanced explanations such as this. Most of it was somewhere in between, no doubt confusing plenty of regular folks.

While Mark Crislip provided a typically excellent and thorough overview of flu vaccine efficacy back in ’09, I’m just going to highlight a few important features of the new paper from Osterholm’s group.

First, there’s not really any news here. Yes, Osterholm and his colleagues did yeoman’s work mining the literature and compiling their data, but they could not escape the fundamental limitations of all meta-analyses. In a meta-analysis, researchers look at existing publications, pick the ones that meet a particular (and in this case extremely strict) set of criteria, and compile the results into a new paper. There’s no new experimentation involved.

Furthermore, the conclusion of this particular meta-analysis should surprise exactly nobody. We’ve known for a long time that flu vaccines are imperfect, and while Osterholm has now put specific numbers on that imperfection for particular age groups, those numbers are neither definitive nor shocking.

Osterholm’s latest results weren’t exactly secret, either. He presented them at the National Influenza Vaccine Summit in May, and the conference report I wrote for them went on the NIVS web site this summer. It may not be in the top of everyone’s news feed, but anyone who’s really tracked this issue closely already knew these results were coming.

These findings don’t alter the main conclusion of decades of public health advice, either. Flu vaccines aren’t 100% effective, but given their outstanding safety record, and the very real risks involved in catching the flu, they’re a whole lot better than nothing.

Finally, while the study’s headline conclusion was that vaccine efficacy averages only around 59% in healthy adults, the team also found that the H1N1 pandemic flu vaccine was a bit above average (69% effective), and discovered even better results for the live attenuated flu vaccine (LAIV, also known as FluMist) in one of the groups at highest risk of severe flu infection:

By contrast, LAIV showed significant protection against infection in young children, preventing influenza in 83% of children aged 7 years or younger. However, the Advisory Committee on Immunization Practices (ACIP) does not currently recommend LAIV over TIV in these children.

Besides its apparently higher efficacy, FluMist has another huge advantage: it’s inhaled rather than injected. My daughter used to scream her head off each Fall before, during, and after her flu shot. Now she can barely stop giggling through the procedure. With the new meta-analysis showing that this snorted vaccine is probably more effective for her than the shot, I can feel good about it as both a virologist and a father.