There is a photo circulating on Facebook that shows the package insert for a flu vaccine that appears to indicate that the vaccine has not been shown to be effective against influenza. Of course, this has gone viral, (sorry for the pun) especially among the anti-vax crowd.
I wanted to do a little investigating to understand the statement on the package insert. The insert says: “FLULAVAL is a vaccine indicated for active immunization against influenza disease caused by influenza virus subtypes A and type B contained in the vaccine… This indication is based on immune response elicited by FLULAVAL, and there have been no controlled trials adequately demonstrating a decrease in influenza disease after vaccination with FLULAVAL”
Its the last statement that has triggered concern: no controlled trials adequately demonstrate a decrease of influenza disease after vaccination? Sounds bad, so lets take a look. The package insert actually has 14 pages. They show the results from a 2005-2006 clinical study involving 7482 people. About half received the vaccine and half received a placebo. Then they followed those individuals to see who got the flu. 23 people receiving the vaccine got a strain of flu against which the vaccine is supposed to protect. 45 receiving the placebo got those same strains of the flu. The statistical analysis shows a vaccine efficacy of about 46%, but the calculation of the confidence interval suggests the efficacy could be as low as 9.8%. Before doing the clinical study, they decided that the lowest limit of the confidence interval had to be above 35% to be considered successful. So it seems that the statement that clinical trials have failed to show efficacy is correct due to the large error range in their data.
Lets consider the other statement, that FLULAVAL is indicated based on it eliciting an immune response. Data from another study is shown in which people were given the vaccine and after 2 weeks, then checked for production of antibodies. In this study the levels of antibody increased to high enough levels in enough individuals that the vaccine met the criteria for success. Furthermore, they did a test called an Immunological Non-Inferiority test. Basically, they wanted to know if FLULAVAL induces an antibody response that is at least as good (ie not inferior to) another vaccine available on the market, FLUZONE. FLULAVAL induced as good a response as FLUZONE. (If you take a look at the FLUZONE package insert, they only report data on antibody responses, and state that no data is available on whether FLUZONE reduces incidence of influenza).
So there appears to be something of a conflict: the clinical trial was not successful but the vaccine appears to induce an appropriate response. Perhaps the measurement of antibodies is not the ideal indicator for predicting protection? This speaks to an important question in vaccine development, which is determining the correlates of protection. That is, what specific part of the immune response is needed for immunity?
Lets also look at other flu vaccines. FLULAVAL is one of seven different flu vaccines available.
Fluarix: Clinical studies show a reduction in influenza disease in vaccinated vs placebo groups.
FluMist: Clinical studies show a reduction in influenza disease in vaccinated vs placebo groups. The data for FluMist are the most impressive, getting as high as 96% efficacy with certain flu strains.
FluVirin: Only shows immunogenicity data, induces antibody response that exceeds the threshold defined for success.
FLUZONE: Only shows immunogenicity data, induces antibody response that exceeds the threshold defined for success.
Afluria: Only shows immunogenicity data, induces antibody response that exceeds the threshold defined for success.
Agriflu: Clinical studies show a reduction in influenza disease in vaccinated vs placebo groups.
Interestingly, it appears that approval of flu vaccines is based on showing that the vaccine can induce a strong antibody response, not showing that the vaccine prevents the disease.
Package inserts don’t communicate the whole story. We also have to consider the total body of evidence, not just one or two tests. There are many other clinical trials demonstrating the efficacy of flu vaccines. Such as this one, this one, this one, and this one.
In the clinical trials described in the package inserts, the severity of disease is not indicated. Did the vaccinated people get less severe disease than the non-vaccinated people? A vaccine that induces sufficient immunity so that it prevents severe disease although you might still get a sniffle, would still be pretty good. There are other outcomes to consider too. Does the vaccine reduce transmission or complications following influenza disease? In Canada, Ontario made efforts to dramatically increase influenza vaccination, with the result of reduced influenza associated mortality and reduced healthcare use. And take this study in which it was found that vaccination of healthcare workers didn’t reduce incidence of flu in those vaccinated but reduced the mortality rate of their patients.
There is an obvious need for a flu vaccine that induces better protection, especially in children and the elderly, and ideally, one that is universal so we dont have to go every year to get a shot and dont have to depend on predictions of what is going to circulate in the future. But the evidence that the flu vaccine is beneficial for individuals and society is pretty strong. Finally, I think this emphasizes the importance of digging deeper to understand the information around us. It is never as simple as it seems and we must avoid reducing information to the simplest single sentence thus removing the underlying complexities.
Disclaimer: I am “not that kind of doctor” so this is not intended to provide any medical advice or recommendations for which vaccine to use.