Food Safety: Familiarity Breeds Fear

In the US, we now keep track of more information about our food supply than we ever did before. Even eggs now have individual tracking codes. Bacteria can get their DNA fingerprints checked. This makes it possible to say that 50 gastroenteritis patients all carry the same strain of Salmonella, and all ate cantaloupes from a single farm in Honduras.

Lately, we hear more and more stories like this. We seem to hear more and more stories about how some food or food product is making people sick, and wonder: Is this a new threat to our food supply? I don't think so. But, then, I don't really know. I went looking for summaries of cases food-related illness, hoping to see some per-capita trends. But it wasn't that simple. Because it is really tough to get a handle on how many people are getting sick. Many of them don't go to the doctor. Many doctors don't take samples to be cultured. How many? Who can tell? Maybe the rate of "underreporting" as it is called is changing. How much? Who can tell? How many of these illnesses are caused by food we buy in the supermarkets? That's hard to tell, too. It is clear, though, that things are changing. Supermarkets, warehousers, shippers, packers, and farmers are all required to keep more comprehensive and elaborate records than they've ever kept before, and the CDC is working hard to get doctors to check more carefully for evidence of foodborne illnesses.

Just exactly how did the FDA link 50 cases of Salmonella in 16 states with cantaloupes from a single farm in Honduras?

"The CDC monitors the frequency of Salmonella ... and assists ... Health Departments".¹ The CDC had to recognize 50 or more Salmonella patients (0.000017% of the US population) as a significant increase in frequency, quite separate from gastroenteritis patients with E. Coli, Giardia, virus or other infections, or no infections at all. For each of those 50 or more patients, there was a physical examination, stool sample collection and culture, and an expensive DNA test to determine how many had the same strain of Salmonella. According to the CDC, the FDA collaborated with the CDC and state health officials to ask a lot of questions about everything the infected people had eaten recently. And compare that to everything a bunch of uninfected people had eaten recently. They would have had to consider all the eggs, milk, poultry, beef, and produce the patients had eaten, as well as any pet turtles or other domesticated or wild animals or animal poop they may have come in contact with. They needed to figure out if the infections came from restaurants or if the foods came from grocery stores. They studied data collected between January 18 and March 5, and concluded that "cantaloupe is the likely source of infections".²

At some point, the FDA started a "traceback investigation" in which they needed to figure out where every bite of cantaloupe had come from, and figure out how it had gotten there, long after the food itself was gone. They needed to figure out if the infections had come from restaurants or supermarkets, from delivery trucks, food warehouses, ships, or trains, a packing house, or a farm. There must have been a team assigned to build and use a database of information from infected and uninfected people, grocers and warehouses and shippers, following all the paths the cantaloupes might have followed. Eventually, they eliminated all other leads, reported that "[p]reliminary results ... indicate that cantaloupes consumed by ill persons were grown in Honduras,² and issued a statement that they all came from one farm.³

I can only guess about all the steps involved. Can you imagine the number of man-hours involved in this investigation? The medical and lab supplies used? Computer power required to build a database of every food the patients ate? Resources used by grocers, truckers, warehouse operators, and other shippers to maintain the records that made it possible to traceback all that food and determine that the only thing those 50 patients had in common was the farm their cantaloupes were grown on?

This kind of investigation simply was not possible even ten years ago. The pieces have only fallen into place recently -- you can see the evidence in the farmer's name on every piece of fruit, and the tracking code on every egg.

Ten or fifteen years ago, this outbreak might not have been recognized. Fifty cases out of 300 million is only 0.000017% of the US population. The CDC might not have been able to connect 2 cases of gastroenteritis in one state to 2 cases of Salmonellosis in another. And no-one could have traced individual pieces of produce back to the farm they grew on in another country. The records simply were not there. But now they are. Since it was established in 1996, FoodNet (a collaborative effort by the CDC, FDA, USDA and state health departments) has expanded to include 15% of the US population in 10 states.⁴

So expect more bad news about produce, simply because the news is available. FoodNet reports make clear that the incidence of laboratory-confirmed infections by commonly-foodborne pathogens has increased in the past couple of years after a few years of decreases. But FoodNet does not make it clear how much of this increase is due to an increase in reporting. Their 1999 discussion of assumptions made in surveillance of foodborne illness makes it clear that they don't have a handle on just how accurate these assumptions are.⁵ But it is clear that as we become more and more familiar with stories of foodborne illnes, we are growing more and more fearful of our food supply.

And I still can't tell: Are things getting worse, or are we just better informed! As long as foodborne illness underreporting rates keep changing, it is going to be difficult to talk about the true rate of change in foodborne illnesses in this country. The CDC publishes weekly tables of the incidence of notifiable diseases in its Morbidity and Mortality Weekly Report. That's a lot of data! I certainly can't make sense of it in an afternoon. I don't have the knowledge and experience to make sense of it in a week or two. I don't even know what field I'd need to study in order to learn how to make sense of it, and what assumptions the experts make.

So I am going to wait for the experts to duke it out, knowing full well that the evidence will seem to point first in one direction and then in another. In the meantime, what am I going to eat?

1. Salmonellosis: What is the government doing, US Centers for Disease Control, as of March 28, 2008
2. Investigation of Outbreak of Infections Caused by Salmonella Litchfield US Centers for Disease Control, March 22, 2008
3. FDA Warns of Salmonella Risk with Cantaloupes from Agropecuaria Montelibano, US Food and Drug Administration, March 22, 2008
4. Preliminary FoodNet Data on the Incidence of Infection with Pathogens Transmitted Commonly Through Food --- 10 States, 2006, FoodNet, April 13, 2007
5. Food-Related Illness and Death in the United States, CDC FoodNet, Emerging Infections Diseases V5No5, September-October 1999