Government officials typically say the flu shot is about 50-60 percent effective in preventing influenza. We’ve used that number ourselves, as recently as this month in the usual annual promo for flu vaccines. But health care researchers who count actual patients say it’s far, far less effective, more in the range of 1-3 percent.
What’s the difference between the competing numbers? And why is the difference so big? The source of the 50-60% effective number is actually shrouded in mystery; the Centers for Disease Control and Prevention, which uses the number a lot, has no specific source. The best guess is it’s a “relative risk reduction,” which means that if your risk was, say, one in 100 before vaccination, and then was one in 200 after, that’s a drop of 50% if you divide those two numbers (100/200). But in terms of what doctors call “absolute” risk, it’s a drop of only half a percent: from 1% (1 in 100) to 0.5% (1 in 200).
The point is you need to know actual numbers of human beings who are saved from a miserable bout of flu by the vaccine, not just some relative percentage comparison which sounds more impressive than it really may be.
And for that, we like something called the Number Needed to Treat. An NNT of 100 means you have to vaccinate 100 people to prevent a single case of flu. An NNT of 33 means 33 people stand in line for the flu shot, and one of them avoids the flu.
And those are probably the realistic Number Needed to Treat figures for flu shots: between 33 in a year with a very effective vaccine that matches that year’s brew of viruses quite well, to 100 in a year with less effective vaccine, to prevent a single case. The numbers come from reviewer Alan Cassels, writing on the Health News Review blog. He got them from a comprehensive review published in 2014 by the prestigious Cochrane group, which crunched all the big studies and came up with this report. The Cochrane conclusion:
Influenza vaccines have a very modest effect in reducing influenza symptoms and working days lost in the general population, including pregnant women. No evidence of association between influenza vaccination and serious adverse events was found in the comparative studies considered in the review.
We did a whole issue of our newsletter a few years ago on Number Needed to Treat. We called it then “the one medical statistic you need to know.” Still true.