"On average, the flu kills 36,000 people each year in the U.S."
This statement, from the Centers for Disease Control and Prevention's Web site, uses a common strategy to highlight -- really exaggerate -- risk. The message begins with an attention-grabbing large number, but it provides no information to put the number into context.
To understand this number, readers first need to know "out of how many?": that is, the size of the population at risk. The number of people who could die of flu-related illness is the entire U.S. population. In 2002, the 36,000 flu-related deaths occurred among approximately 288 million people. (The U.S. population today is estimated at about 297 million, but this article and the table below use data from 2002 -- the most recent national death data available.)
By highlighting the numerator (the number of deaths) without mentioning the denominator (the total number at risk), the government focuses attention on a large number instead of a small proportion. Explicitly providing this denominator would probably change how readers perceive their risk of dying from a flu-related illness. A clearer expression of the reader's risk might say: This year 0.01 percent of Americans (one in 10,000) will die of a flu-related illness. For most people, the message feels even less threatening when stated as follows: "Out of 10,000 Americans, 9,999 will not die of a flu-related illness this year."
Since 90 percent of flu-related deaths occur among people 65 and older, it is helpful to think about the risks according to age. For people younger than 65, the chance of a flu-related death is about 1 in 100,000. For people 65 and older, the chance is about 1 in 1,000.
Second, once you know the size of risk, it is important to have information to put the risk in context. One way to do this is to compare the risk of flu-related death to death from other common diseases -- like heart disease. For the U.S. population, the risk of dying from heart disease in a given year is 0.24 percent -- more than 20 times more common than death from flu. Another way to put the 36,000 in context is to compare it to the 2.4 million deaths that occur annually in the United States -- so flu-related deaths make up only 1.5 percent of all deaths that occur each year.
The chance of flu-related death this year -- like any statistic -- can be stated in a variety of ways. We think the most useful way is to present the annual risk (that is, divide the numerator by the denominator) as shown in the table below.
-- Steven Woloshin, Lisa M. Schwartz, H. Gilbert Welch