Recently on Food for Thought:
Fisheries Don't Welcome This Whelk (10/9/99)
Cheery, Beer-y Kidneys (9/11/99)
What's Coming Out of Baby's Bottle? (7/24/99)

While all of this makes tallying the incidence of food poisoning quite challenging, it hasn't stopped Uncle Sam from trying. Last month, Paul S. Mead and his colleagues at the federal Centers for Disease Control and Prevention offered up their latest estimate in a 19-page report. Published in the September-October issue of Emerging Infectious Diseases, it concludes that some 76 million U.S. residents develop foodborne illness each year.

That incidence rate would indicate that on average more than one in four people eat sickening food each year. The data also indicate that an estimated 325,000 require hospitalization-and almost 5,200 die-because of foodborne illness.

Where did Mead's team come up with these numbers?

They extracted confirmed cases of food poisoning from nine data bases, such as the Foodborne Diseases Active Surveillance Network (FoodNet) and the National Hospital Discharge Survey. In addition, they read studies that described investigations into particular outbreaks and the degree to which poisoning events appear to have been underreported. Then, they multiplied the number of known cases by the likely underreporting figure, taking into account the different types of disease-causing agents, and summed the totals.

For instance, they cite unpublished data indicating that about 38 times as many cases of Salmonella poisoning occur as are reported. Because the bacterium responsible for this illness causes nonbloody diarrhea, Mead's team multiplied the number of cases of Salmonella poisoning and other nonbloody diarrheal incidents by 38. Because the underreporting rate for Escherichia coli O157:H7—which causes a bloody stool—is only about half as large, the epidemiologists upped the known incidence of bloody diarrheal disease 20-fold.

That gave them the gross, upper estimate of incidence for diseases caused by these germs. However, because these germs can be transmitted by means other than food—such as water contamination—they had to scale down their tally, in some cases by around two-thirds.

What they know...

Among all illnesses linked to food, the scientists estimated that 67 percent trace to contamination with viruses such as Rotavirus, Norwalk-like viruses, or Hepatitis A; 30 percent are caused by bacteria, such as Salmonella, E. coli, and Campylobacter; and less than 3 percent are caused by parasites such as Cryptosporidium or Trichinella.

As their data show (excerpted in table, below), the most common causes of food poisoning—viruses—are least likely to lead to fatalities. Even among the other classes of disease-causing agents, only a few stand out as being particularly deadly. Toxoplasmosis, for instance, caused by a parasite most commonly associated with sheep and cat feces, was linked to 20 percent of food-poisoning deaths, while accounting for less than 1 percent of all foodborne illness.

And Listeria—a bacterium that can multiply prolifically even in refrigerated foods (SN: 2/7/98, p. 89)—caused nearly 30 percent of food-poisoning deaths, while hardly registering as a major source of illness. Indeed, these new data indicate that nearly every Listeria victim requires hospitalization, and one in five of Listeria poisonings proves fatal.

Though most people know and fear botulism, only about 60 people in the United States contract this disease annually. That's just 2.5 percent as many people as become sickened by Listeria, and the fatality rate is only about one-third as high as Listeria's. Only Vibrio vulnificus, a bacterium usually transmitted by uncooked shellfish from polluted coastal waters, is deadlier than Listeria. This Vibrio kills almost 40 percent of its victims.

...and don't know

Unfortunately, Mead's team points out, those statistics represent only the tip of the iceberg. The majority of food poisonings that are characterized by acute gastrointestinal symptoms—62 million cases, or 81 percent of all foodborne disease—cannot be attributed to known agents.

This isn't surprising, the CDC scientists point out, since many pathogens of greatest concern today—notably Campylobacter jejuni, E. coli O157:H7, Listeria monocytogenes, and Cyclospora cayetanesis—"were not recognized as causes of foodborne illness just 20 years ago."

The new CDC calculations suggest that more than a quarter-million hospitalizations for acute gastroenteritis stem from food poisoning by unknown agents. Similarly, 3,360 deaths—or 65 percent of those attributable to food poisonings—trace to unknown agents.

Mead and his colleagues concede that their estimates are based on considerable extrapolation and inference. However, they note that these numbers are also grounded in more data than most earlier estimates.

Surprisingly, CDC's new overall total estimate of annual food poisonings falls within just 7 percent of the 81 million cases per year calculated by the U.S. General Accounting Office.

Interestingly, last year a reporter with the Post-Crescent in Appleton, Wisc., set about trying to track down the basis of many widely circulated—and largely unattributed—numbers quantifying food-poisoning in the United States. Dan Wilson noted that most news stories treated whatever number they cited as "one of those accepted truths that require no attribution, like 'squirrels have bushy tails.'"

In the May/June 1998 Columbia Journalism Review, he described his trek to verify those numbers and establish their source. He ended up frustrated as he learned that most of the cited food-poisoning stats were based on reports whose tallies were quite not firm.

CDC is similarly frustrated by the imprecise data it has to work with.

Things are improving, however. For instance, FoodNet, one of the data bases on which CDC relied, has recently started collecting data on cases of vomiting not associated with diarrhea. That could capture many unreported episodes of acute, short-duration poisonings.

Several other countries are already doing a much better job of catching cases of food poisoning, notes Elizabeth Scott, a consulting Boston-area microbiologist focusing on foodborne pathogens. In Britain and Holland, for instance, physicians must report all cases of gastroenteritis. Particularly where these reports turn up sporadic cases, she says, one begins to suspect food poisoning—especially in the home.

In many ways, she says, the big surprise is that there isn't more food poisoning. Her studies and those by others have shown that people don't tend to practice good kitchen hygiene. She notes, "People consider it common sense to use detergent and hot water to wash cutting boards and sponges." Not so. "The detergent just breaks up [colonies of] the bacteria and spreads them around. It doesn't kill them," she told Science News Online.

This means that using damp sponges that have been hanging around the sink "and cleaned with nothing more than detergent" risks seeding counters and kitchenware with millions of potentially sickening bacteria and viruses. Completely drying sponges and counters or cleansing them with chlorine bleach is effective in killing microbes.

Indeed, a 1998 survey by the Food and Drug Administration found that though increasing numbers of Americans are becoming aware of food-safety issues, they continue to practice "risky behaviors." To quantify this, FDA has begun videotaping 150 Utah residents as they cook at home. The goal is to identify where people might be making mistakes—compromising safety without realizing it. The findings are slated to be synthesized and published early next year.

(Mead et al./Emerging Infectious Diseases)