STATS ARTICLES 2010
Choking on hot dog data: Has pediatricians’ group gone wild?
Trevor Butterworth, February 22 , 2010
The American Academy of Pediatrics issues a stern warning on an “under-appreciated” risk. Good. But can it really be calling for the “recall” of food products with “significant and unacceptable choking hazard” based on 29-year old data showing five deaths per year?
For Andy Warhol, the hotdog was the harbinger of American greatness, the great leveler of class:
“…When Queen Elizabeth came here and President Eisenhower bought her a hot dog I’m sure he felt confident that she couldn’t have delivered to Buckingham Palace a better hot dog that that one he bought her for maybe twenty cents at the ballpark. Because there is no better hot dog than a ballpark hot dog. Not for a dollar, not for ten dollars, not for a hundred thousand dollars could she get a better hot dog. She could get one for twenty cents and so could anybody else.”
For the American Academy of Pediatrics, the hot dog is a dangerous child killer. As the Washington Post reports,
“The nation's largest pediatricians group is calling for sweeping changes in the way food is designed and labeled to minimize children's chances for choking. Choking kills more than 100 U.S. children 14 years or younger each year and thousands more - 15,000 in 2001 - are treated in emergency rooms. Food, including candy and gum, is among the leading culprits, along with items like coins and balloons. Of the 141 choking deaths in kids in 2006, 61 were food-related.”
But according to the academy’s policy paper, the reality appears to be the reverse of the Post’s reporting of the story. Latex balloons were the leading cause of choking deaths among children aged 14 or under (29 percent), according to Consumer Product Safety Commission (CPSC) data cited. Hot dogs, the academy claims, accounted for 17 percent of “food-related asphyxiations.” This was based on a different data set compiled between 1979 and 1981, and published in 1984 in the Journal of the American Medical Association (“Childhood Asphyxiation by Food: A National Analysis and Overview” Harris et al).
A close look at this paper shows that the 17 percent was composed of 16 choking deaths from hot dogs and one from a sausage from a total of 41 states over three years. All the children who died were aged three or under. So, to put the risk into perspective, approximately five children died each year in the U.S. from choking on a hotdog - along with, approximately, 3.3 from candy, 3 from peanuts or other nuts, 2.7 from grapes, 2.3 from other meat, 2 from carrots, 1.7 from popcorn, 1. 5 from apples, 1.25 from beans, 1 from macaroni/noodles.
The question is whether data that is 29 years old is still representative of the present risk -
especially as a core aspect of the paper was to suggest strategies to raise awareness about choking and minimize the threat
A 2008 paper, “Fatal and non-fatal food injuries among children (aged 0-14 years),” Altkorn et al, which was published in the International Journal of Pediatric Otorhinolaryngology, may offer some answers.
The researchers analyzed ten years of injury data from 26 pediatric hospitals in the U.S. and Canada from 1989 to 1998. Hot dogs led in choking fatalities, with a rate of 1.6 deaths per year - followed by candy, at 1 death per year, grapes at 0.8 deaths, meat at 0.7 deaths, peanuts at 0.7, carrots at 0.6, cookies at 0.6, apples at 0.5, popcorn at 0.5, and bread at 0.4.
Peanuts accounted for 37.5 non-fatal injuries per year, followed by meat with 9.6, and sunflower seeds at 9.5. Hot dogs accounted for 6.2 injuries per year.
How representative the sample of hospitals is relative to overall national risk is unclear; but it is reasonable to assume, based on this data, that the number of children who die from choking on hot dogs is somewhere between 1.5 and 5 per year in the United States. The risk is overwhelmingly to children aged three and under, based on the natural tendency of infants to put stuff in their mouths. While each death is a terrible tragedy, it is not clear that this is something that this amounts to a “significant and unacceptable choking hazard” to the public warranting recall procedures, product changes and more warnings than already exist.
The problem with the American Academy of Pediatrics paper is that it uses percentages without the context of frequency, so that you end up comparing percentage fatalities for widely different time periods without any indication of the actual number of deaths.
Take the threat of latex balloons, which appear to account for a third of all choking deaths. A closer analysis of this data, as performed by Rimell et al (“Characteristics of Objects that Cause Choking in Children,” JAMA, December 13, 1995), explains that the mortality data for balloons covered 20 years. So that 29 percent cited by the academy breaks down into, roughly, 6.5 deaths per year. (The academy then jumps to a non-peer reviewed study by the activist Public Interest Research Group (PIRG) which puts the death rate from balloons at 4.8 per year, but who knows whether that is credible?). Again, it’s important to know this, but it is equally important to know that this isn’t a huge problem, compared say to other unintentional childhood injuries.
The academy protests the urgent need for a comprehensive strategy, wherein the FDA needs to crack down on the problem of choking through surveillance, education, and enforcement. While choking deaths are preventable, the fatality rate is so low that it is difficult to envision success without drastic intervention. Given the endless diet of scare headlines being propagated through the media, it is interesting to see that, while news organizations such as the Washington Post produced uncritical news reports, reader reaction isn‘t quite as respectful. The New York Times’ Sam Sifton derided the academy with the following tweet: “Grapes and carrots are DANGEROUS. Nanny-state doctors urge choke-warning labels on apples, hotdogs too.” Gawker’s response was “Hot dogs are murdering our children.” Letters responding to ABC News question - should wieners have warning labels - were overwhelmingly critical of the idea.
Sometimes, public health policy by alarmism doesn’t always work.