Flu results in "about 250,000 to 500,000 yearly deaths" worldwide, Wikipedia tells us. "The typical estimate is 36,000 [deaths] a year in the United States," reports NBC, citing the Centers for Disease Control. "Somewhere between 4,000 and 8,000 Canadians a year die of influenza and its related complications", according to the Public Health Agency of Canada, adding that "Those numbers are controversial because they are estimates."

"Controversial" is an understatement, and not just in Canada, and not just because the numbers are estimates. The numbers differ wildly from the sober tallies recorded on death certificates -- by law every certificate must show a cause -- and reported by the official agencies that collect and keep vital statistics.

According to the National Vital Statistics System in the U.S., for example, annual flu deaths in 2010 amounted to just 500 per year -- fewer than deaths from ulcers (2,977), hernias (1,832) and pregnancy and childbirth (825), and a far cry from the big killers such as heart disease (597,689) and cancers (574,743). The story is similar in Canada, where unlikely killers likewise dwarf Statistics Canada's count of flu deaths.

Even that 500 figure for the U.S. could be too high, according to analyses in authoritative journals such as the American Journal of Public Health and the British Medical Journal. Only about 15-20 per cent of people who come down with flu-like symptoms have the influenza virus -- the other 80-85 per cent actually caught rhinovirus or other germs that are indistinguishable from the true flu without laboratory tests, which are rarely done. In 2001, a year in which death certificates listed 257 Americans as having died of flu, only 18 were positively identified as true flus. The other 239 were simply assumed to be flus and most likely had few true flus among them.

"U.S. data on influenza deaths are a mess," states a 2005 article in the British Medical Journal entitled "Are U.S. flu death figures more PR than science?" This article takes issue with the 36,000 flu-death figure commonly claimed, and with describing "influenza/pneumonia" as the seventh leading cause of death in the U.S.

"But why are flu and pneumonia bundled together?" the article asks. "Is the relationship so strong or unique to warrant characterizing them as a single cause of death?"

The article's answer is no. Most pneumonia deaths are unrelated to influenza. For example, "stomach acid suppressing drugs are associated with a higher risk of community-acquired pneumonia, but such drugs and pneumonia are not compiled as a single statistic," explained Dr. David Rosenthal, director of Harvard University Health Services. "People don't necessarily die, per se, of the [flu] virus -- the viraemia. What they die of is a secondary pneumonia."

Pneumonia, according to the American Lung Association, has more than 30 different causes, influenza being but one of them. The CDC itself acknowledges the slim relationship, saying "only a small proportion of deaths... only 8.5 per cent of all pneumonia and influenza deaths [are] influenza-related."

Because death certificates belie claims of numerous flu deaths, CDC enlisted computer models to arrive at its 36,000 flu-death estimate. But even here it needed to bend conventional medical terminology to arrive at compelling death numbers.

"Cause-of-death statistics are based solely on the underlying cause of death [internationally defined] as 'the disease or injury which initiated the train of events leading directly to death,'" explains the National Center for Health Statistics. Because the flu was rarely an "underlying cause of death," the CDC created the sound-alike term, "influenza-associated death."

Using this new, loose definition, CDC's computer models could tally people who died of a heart ailment or other causes after having the flu. As William Thompson of the CDC's National Immunization Program admitted, influenza-associated mortality is "a statistical association ... I don't know that we would say that it's the underlying cause of death."

The CDC's decision to play up flu deaths dates back a decade, when it realized the public wasn't following its advice on the flu vaccine. During the 2003 flu season "the manufacturers were telling us that they weren't receiving a lot of orders for vaccine,"Dr. Glen Nowak, associate director for communications at CDC's National Immunization Program, told National Public Radio. "It really did look like we needed to do something to encourage people to get a flu shot."

The CDC's response was its "Seven-Step 'Recipe' for Generating Interest in, and Demand for, Flu (or any other) Vaccination," a slide show Nowak presented at the 2004 National Influenza Vaccine Summit.

Here is the "Recipe that fosters influenza vaccine interest and demand," in the truncated language that appears on his slides: "Medical experts and public health authorities [should] publicly (e.g. via media) state concern and alarm (and predict dire outcomes) - and urge influenza vaccination." This recipe, his slide show indicated, would result in "Significant media interest and attention ... in terms that motivate behavior (e.g. as 'very severe,' 'more severe than last or past years,' 'deadly')." Other emotive recommendations included fostering "the perception that many people are susceptible to a bad case of influenza" and "Visible/tangible examples of the seriousness of the illness (e.g., pictures of children, families of those affected coming forward) and people getting vaccinated (the first to motivate, the latter to reinforce)."

The CDC unabashedly decided to create a mass market for the flu vaccine by enlisting the media into panicking the public. An obedient and unquestioning media obliged by hyping the numbers, and 10 years later it is obliging still.