Influenza vaccine

There’s been a fair bit of discussion online about the new study in the British Medical Journal1  throwing doubt on Tamiflu’s effectiveness against influenza.  (If you haven’t already seen this, see the Avian Flu Diary for an excellent summary of the situation, and an update here.2 Also see the CDC’s recommendations for antivirals here.)

There’s a new paper3 that takes a similar skeptical look at the effectiveness of influenza vaccines in the elderly (I emphasize, the question is how well the vaccines work in the elderly; as far as I know there’s general agreement that the flu vaccines work reasonably well in younger people and children).  It’s long been agreed that flu vaccines don’t work as well in elderly people as they do in the young.  This is a real concern, because it’s also generally agreed that seasonal influenza takes its greatest toll on people over 65 years or so. 4

However, while it’s been agreed that flu vaccines don’t work as well in the elderly, this new paper suggests the situation is even worse than we thought; and in fact the vaccine may hardly work at all (in the elderly!).  The authors argue that the apparent benefits of vaccination, as previously detected, aren’t real; they’re the result of bias in vaccination.  That is, it’s not the people who are vaccinated don’t die of flu; rather, people who aren’t going to die of flu, are the ones who get vaccinated. Correlation is not causation:

Because persons who are most likely to die are less likely to receive the vaccine, vaccination appears to be associated with a much lower chance of dying; thus, the “effectiveness” of the vaccine is in great part due to the selection of healthier individuals for vaccination, rather than due to true effectiveness of the vaccine.3

They offer a number of possible reasons for this bias — perhaps the most frail people don’t get to vaccine clinics; perhaps doctors try harder to have their most energetic patients vaccinated; perhaps those people on the verge of death don’t want further intervention.  I don’t have the statistical chops to critique their analysis in detail, but from what I do follow it seems well supported.  I’d like to see more work on this, because it obviously has huge, huge implications for vaccination policy and vaccination research.

We hope this knowledge will stimulate research into better vaccines for elderly patients (perhaps by use of higher doses or adjuvants) and will lend more weight to the importance of vaccinating schoolchildren to prevent disease in the rest of the population.3

Even if this suggestion eventually doesn’t hold up, it’s a really important message: Always be skeptical.

By the way, I and my family got the H1N1 vaccine last night (even though I read this article last week some time), so I’m walking the walk as well as talking the talk.


  1. Jefferson, T., Jones, M., Doshi, P., & Del Mar, C. (2009). Neuraminidase inhibitors for preventing and treating influenza in healthy adults: systematic review and meta-analysis BMJ, 339 (dec07 2) DOI: 10.1136/bmj.b5106
    Also see the editorial:
    Godlee, F., & Clarke, M. (2009). Why don’t we have all the evidence on oseltamivir? BMJ, 339 (dec08 3) DOI: 10.1136/bmj.b5351[]
  2. And if you’re not reading AFD already, why aren’t you?[]
  3. Baxter, R., Lee, J., & Fireman, B. (2009). Evidence of Bias in Studies of Influenza Vaccine Effectiveness in Elderly Patients The Journal of Infectious Diseases DOI: 10.1086/649568[][][]
  4. The biggest difference between this H1N1 pandemic influenza this year and seasonal flu is in the age groups at risk.  Normally, seasonal influenza causes death almost entirely in the older people.  This year, older people are relatively, though not absolutely, not being killed by influenza, while younger people are.  So while it’s true that the mortality rate due to the new swine-origin influenza virus isn’t really higher than usual, and may even be lower, that mortality is clustered in an unfamiliar group — young people, and children, who normally are at essentially no risk of flu-caused death.[]