A shot in the dark: flu vaccinations do save lives, but they are nowhere near as effective as generally claimed, says health analyst Robert Taylor.(Co

THIS month, millions of elderly people in the US, Europe and elsewhere are lining up for their annual flu shots. This is widely seen as a hugely effective public health intervention that halves the risk of dying over the winter for those aged 65 or above. For every 200 vaccinations, one life is saved.

Yet there is now overwhelming evidence that this claim is simply too good to be true, and that we should be looking at additional ways to tackle flu.

According to the US Centers for Disease Control and Prevention (CDC), flu kills around 36,000 people every winter in the US, of whom around 30,000 are 65 or over. That's around 5 per cent of the 650,000 annual winter deaths in that age group. Flu itself is rarely recorded as a cause of death: instead, it makes elderly people more likely to die from other causes such as bacterial pneumonia, heart disease or stroke.

Most rich countries aim to cut this toll by vaccinating those at highest risk--but how well does this actually work? The best way to check would be to carry out trials that compare those who are vaccinated against those who aren't, with participants randomly allocated to each group. But as flu shots are known to be beneficial, it would be unethical to withhold vaccination from some people. So researchers compare those who choose to be vaccinated with those who don't, then use statistical methods to control for any other differences between the two groups. One large meta-analysis of such studies concluded that those who get flu shots are half as likely to die over the winter as their unvaccinated peers (Vaccine, vol 20, p 1831). Several studies have come to a similar conclusion, and this has become the consensus view.

That sounds wonderful--perhaps a little too wonderful. In 2005, Lone Simonsen of George Washington University in Washington DC and her colleagues, including myself, showed that the number of flu deaths among elderly people in the US has remained at around 5 per cent of winter deaths in that group, even though vaccination coverage has skyrocketed--from around 15 per cent receiving the shot in 1980 to around 70 per cent today (Archives of Internal Medicine, vol 165, p 265). How could flu vaccination be preventing half of all winter deaths, when flu accounted for only 5 per cent of those deaths back in 1980, when most people were not vaccinated?

[ILLUSTRATION OMITTED]

Then in 2006, epidemiologist Lisa Jackson and her colleagues at the Center for Health Statistics in Seattle, Washington, analysed a Seattle medical database using the same statistical methods as the previous studies. This showed that the biggest supposed benefit from the flu shot occurred in the months before the flu season even started (International Journal of Epidemiology, vol 35, p 337).

Jackson argued that the studies failed to account for extremely ill and frail elderly people, who are more likely to die but less likely to be vaccinated, making vaccination appear more beneficial than it actually is.

But the debate wasn't over. Last year, Kristin Nichol and her colleagues from the University of Minnesota in Minneapolis published the biggest study yet, which used slightly different statistical methods and involved records from tens of thousands of patients in three cities over 10 years (The New England Journal of Medicine, vol 357, p 1373). It came to the same incredible conclusion--that vaccination was preventing about half of all winter deaths. Researchers including Simonsen, Jackson and myself analysed Nichol's methods, and we believe this finding is subject to the sort of bias previously identified by Jackson (The Lancet Infectious Diseases, vol 7, p 658).

Last week, Simonsen and Nichol discussed the issue at the Interscience Conference on Antimicrobial Agents and Chemotherapy in Washington DC. Nichol acknowledged that there might still be some bias in her latest survey, but pointed out that because flu deaths are estimated indirectly--by counting extra deaths beyond those expected in any particular winter--researchers may be underestimating the number of people that flu kills.

No one suggests that seniors should not be vaccinated: there is no doubt that flu shots reduce their risk of dying because of flu. But health professionals must accept that flu shots do not have the impossible benefits claimed, and start working on more effective prevention programmes. Manufacturers should step up efforts to develop vaccines that are more effective in the elderly, and public health authorities should assess whether it would be better to protect elderly citizens indirectly by vaccinating groups such as schoolchildren or carers, to prevent transmission of the virus.

* Finally, we need to develop better statistical methods for measuring the effectiveness of the flu vaccine. This has implications that go beyond flu, as similar methods are used in other areas of medicine where randomised controlled trials are ruled out--for example to show that people with pneumonia who are given cholesterol-lowering statins are much less likely to die.

Robert Taylor is a global health policy and medical informatics consultant at Sage Analytica in Bethesda, Maryland

Source Citation:Taylor, Robert. "A shot in the dark: flu vaccinations do save lives, but they are nowhere near as effective as generally claimed, says health analyst Robert Taylor.(Comment and analysis)." New Scientist 200.2681 (Nov 8, 2008): 18(1). Academic OneFile. Gale. BROWARD COUNTY LIBRARY. 30 Sept. 2009
.

Gale Document Number:A189289231

Disclaimer:This information is not a tool for self-diagnosis or a substitute for professional care.