Header graphic for print

Food Safety News

Breaking news for everyone's consumption

Letter From The Editor: The Numbers

I’ve always been a numbers guy.  I am not talking about mathematics, but just plain numbers.

   

I do still wake up at night occasionally thinking I have a college algebra or calculus test in the morning.  But that’s math.

At some point I took an aptitude test and learned that I should be a certified public accountant, a profession I’ve come to respect, but at the time I could think of nothing more boring. 

Those tests did point me at arithmetic.  Just give me the numbers, please.

No surprise I am a baseball fan and, yes, on occasion I’ve placed a wager knowing the odds.  I know the OBP and the ERA, and the over and under.  Point me to the sports book.

So let me tell you how I view the new numbers on foodborne illness in America from the Centers for Disease Control and Prevention (CDC).   

The new projections and those old estimates CDC made in 1999 are kind of like Powerball and Mega Millions.

No, I do not suggest buying lottery tickets.  Back when the mob ran the numbers you had a decent chance of winning occasionally, but not since the government took over the business.

Powerball, which is played in 42 states and the District of Columbia, gives you a 1 in 62 chance of turning your $1 bet into the minimum $3 win.  Mega Millions, played in 41 states, gives you a 1 in 75 shot of turning a $1 bet into a $2 win.

People in many states can play both games.  Those who can play both probably do not pick one or the other based on the differences in those odds.  Instead, they probably decide which game to play based on habit or the size of the current jackpot.

If there were a game, however, in the form of a lottery or a casino table where you had a 1 in 4 chance of winning and it did not pay out in foodborne illnesses, people like me would be lining up to get in.

Now let’s pretend we’d been playing in this casino with 1 in 4 odds since 1999, and the word comes down it is changing to 1 in 6.   What would we do?  My bet is we would not change casinos, nor even lift our heads up from our tables.

We would continue to play because, just like the difference between Powerball and Mega Millions, there would not be much distinction in those changed odds.

The fact that CDC’s new best estimate for foodborne illnesses–48 million illnesses, 128,000 hospitalizations, and 3,000 deaths–is less than its 1999 estimate of 76 million illnesses, 325,000 hospitalizations, and 5,000 deaths is not much of a game changer.

The day CDC steps up and says the chance of you getting a foodborne illness is no more likely than your winning $3 in Powerball, now that would be a game changer.

Right now, winning a week of explosive diarrhea or worse is still way too easy.  It will happen to you long before you win $3.  It’s all in the numbers.

© Food Safety News
  • The new numbers can become a game changer ONLY IF the news media and talking head report them accurately. Unfortunately, they usually report them inaccurately. Even worse, they usually quote important players in the food safety arena (e.g., Sens. Reid, Durbin and Harkin) when they are inaccurately quoting them without pointing out that the estimates are being misquoted.
    Already, a writer for Change.org (Danny Jensen) has written, “Nearly one in six people — or roughly 48 million individuals — in the United States are sickened by the food they eat each year and more than 3,000 die, according to two new studies from the U.S. Centers for Disease Control and Prevention (CDC). ” (http://food.change.org/blog/view/rotten_odds_food_poisoning_sickens_1_in_6_americans_yearly#reply_form451819). Much to his credit, Mr. Jensen acknowledged the error when I pointed it out to him. What matters most is whether or not he will quote them accurately next time.
    Additionally, it is hard for the new numbers to change the game when people don’t understand that these “estimates” are, at best, “educated guesses.” Due to the 9.5% increase in population between 1997 and 2006 (the actual points in time of the 1999 and 2010 estimates), the “estimated” incidence rate of illnesses dropped 42%, hospitalizations dropped 67 % and deaths dropped 47%. YET, the researchers emphasize this doesn’t show any improvement.
    What the difference in the estimates between 1999 and 2010 does show is how little faith should be placed in them. They have a very low credibility.
    In fact, their credibility is so low, that Marion Nestle wrote on her blog, Food Politics, “But the 2010 estimates, like the 1999 estimates, are still guesses—just better ones based on methods that were not available in 1999” (http://www.foodpolitics.com/2010/12/cdc-halves-foodborne-illness-count-but-why-now/).
    Guesses. That’s why, during all those years I was teasing meaning out of numbers, we called estimates like these, “guesstimates.”
    Finally, the last 2 paragraphs show a fundamental misunderstanding of the actual risk of contracting a foodborne illness in the US. The chance of $1 becoming $3 in Powerball is already many times higher the chance of the average American having “a week of explosive diarrhea or worse” due to a foodborne illness. And CDC’s FoodNet numbers clearly show that the chances of contracting a foodborne illness are continuing to decline.
    Chance is calculated based upon the number of opportunities. In a given day, how many different ingredients are in the food a person eats? Dozens? And when was the last time anyone at Food Safety News had a foodborne illness? If a person only had 10 ingredients in his/her food each day, that would be 3650 opportunities to contract a foodborne illness each year. Yet even the CDC’s clearly overstated numbers of illnesses give an incidence rate of 48,000,000 illnesses/299,000,000 people/year or .1605/year. In that same year, with 3650 opportunities in Powerball, the likelihood of winning $3 is essentially 100%. The chance of not winning $3 is infinitesimally small.

  • The new numbers can become a game changer ONLY IF the news media and talking head report them accurately. Unfortunately, they usually report them inaccurately. Even worse, they usually quote important players in the food safety arena (e.g., Sens. Reid, Durbin and Harkin) when they are inaccurately quoting them without pointing out that the estimates are being misquoted.
    Already, a writer for Change.org (Danny Jensen) has written, “Nearly one in six people — or roughly 48 million individuals — in the United States are sickened by the food they eat each year and more than 3,000 die, according to two new studies from the U.S. Centers for Disease Control and Prevention (CDC). ” (http://food.change.org/blog/view/rotten_odds_food_poisoning_sickens_1_in_6_americans_yearly#reply_form451819). Much to his credit, Mr. Jensen acknowledged the error when I pointed it out to him. What matters most is whether or not he will quote them accurately next time.
    Additionally, it is hard for the new numbers to change the game when people don’t understand that these “estimates” are, at best, “educated guesses.” Due to the 9.5% increase in population between 1997 and 2006 (the actual points in time of the 1999 and 2010 estimates), the “estimated” incidence rate of illnesses dropped 42%, hospitalizations dropped 67 % and deaths dropped 47%. YET, the researchers emphasize this doesn’t show any improvement.
    What the difference in the estimates between 1999 and 2010 does show is how little faith should be placed in them. They have a very low credibility.
    In fact, their credibility is so low, that Marion Nestle wrote on her blog, Food Politics, “But the 2010 estimates, like the 1999 estimates, are still guesses—just better ones based on methods that were not available in 1999” (http://www.foodpolitics.com/2010/12/cdc-halves-foodborne-illness-count-but-why-now/).
    Guesses. That’s why, during all those years I was teasing meaning out of numbers, we called estimates like these, “guesstimates.”
    Finally, the last 2 paragraphs show a fundamental misunderstanding of the actual risk of contracting a foodborne illness in the US. The chance of $1 becoming $3 in Powerball is already many times higher the chance of the average American having “a week of explosive diarrhea or worse” due to a foodborne illness. And CDC’s FoodNet numbers clearly show that the chances of contracting a foodborne illness are continuing to decline.
    Chance is calculated based upon the number of opportunities. In a given day, how many different ingredients are in the food a person eats? Dozens? And when was the last time anyone at Food Safety News had a foodborne illness? If a person only had 10 ingredients in his/her food each day, that would be 3650 opportunities to contract a foodborne illness each year. Yet even the CDC’s clearly overstated numbers of illnesses give an incidence rate of 48,000,000 illnesses/299,000,000 people/year or .1605/year. In that same year, with 3650 opportunities in Powerball, the likelihood of winning $3 is essentially 100%. The chance of not winning $3 is infinitesimally small.