The Roughest Equivalent... Statistics

May 7, 2008 – 12:24 am

As the old saying goes, there are three kinds of lies: lies, damn lies, and statistics. To your left is a screen cap of a pharmaceutical ad from Sanofi Pasteur. It says: "Most hospitalizations and nearly all deaths in the US from whooping cough (pertussis) are reported in infants less than 6 months." Then it goes on to say: "Whooping cough is highly contagious and reported cases have been on the rise in recent years." The point of the ad was to make you believe you should get vaccinated against whooping cough right away, because if you don't, you're going to kill your baby.

Scary, huh? Here's what they didn't tell you...

• Most Americans were already vaccinated against whooping cough as children and since mid-2005, they've been adding a whooping cough booster to the periodic tetanus booster most adults get.
• In 1999, after cutting SIDS (Sudden Infant Death Syndrome) deaths by 40% from the "Back to Sleep" campaign and public health programs, there were 2,689 reported infant deaths from SIDS. In 2003, with whooping cough on the rise, there were... brace yourself... 20 infant deaths from whooping cough. Your child was 134 times more likely to die from SIDS than whooping cough.

Have you ever heard a news report that said doing something doubled your risk of dying from something? Did they ever tell you what the average risk was in the first place?

According to the National Center For Health Statistics, in 2005, the death rate in the United States was 825.9 per 100,000 people alive in the U.S. during that year. That means if you lived in the U.S. in 2005, your statistical odds of dying were slightly over 8/10th of 1%. Not even 1 in 100, but 8.259 in 1000. And this was your chances of dying from anything. Every cause of death in the United States combined took out less than 1 in 100 people. That's smoking, cancer, drunk drivers, axe murderers, suicides, and all the winners of the Darwin awards combined.

Cancer accounted for roughly 22.8% of deaths. And of that, leukemia accounts for a bit under 4%. So your odds of dying from Leukemia in 2005 were roughly 7.5 per 100,000 people alive, 1 in 13,333. And your odds of dying from it in 2006... roughly the same. Unless the mortality rate from Leukemia changed significantly nationwide, your odds would be roughly 1 in 13,333 each year you lived.

Now, to be fair, I should provide some context myself. Your real personal odds of death are influenced moment to moment by age, risk factors, etc. If you're an avid skydiver, your real personal odds of dying from a skydiving accident are significantly higher than someone who doesn't go skydiving. When we factor across the entire population, and you're just an ageless, faceless statistical unit of 1 in 304 million, a 4-year-old has the same odds of dying from a heart attack as an 89-year-old. But it's worth noting that in such a case, the 89-year-old has the same odds of chasing their ball into traffic as the 4-year-old.

That said, if there was some news story that said eating baby back ribs tripled your odds of dying from leukemia, you'd see rib lovers in a panic, and all because no one put those numbers in context, which is: your odds of dying from Leukemia went from 1 in medium thousands to 1 in small thousands.

This is not to say that things that will cut your risk of death from certain causes are worthless. Better health generally translates to better quality of life; more dancing and less wheezing, more playing and less griping. If it's really good for you, it won't just show up as some abstract statistic. You'll feel it.

But next time you see some ad or read some report that says you'll triple your chance of X or lower your odds of Y, ask "and that's roughly equivalent to...?" Call advertisers and news outlets on the carpet and demand they put these statistics in context. Or better yet, put them in context for yourself.