True or False: 5,000 People Die Annually in the U.S. from Foodborne Illness
Perhaps even more often than the 1,500 food-mile canard, the claim that 5,000 people in the U.S. die each year from foodborne illnesses has been repeated again and again by the food activist community. The CDC says it. The WHO says it. Wikipedia says it. FOOD, INC. says it. The Center for Foodborne Illness Research and Prevention says it. S.T.O.P. says it. PBS says it. Daily Kos says it. (Foreshadowing what's to come: This post is one of the very few times I've agreed with something written on Daily Kos.) The Produce Safety Project says it. Resources For the Future says it. Consumers Union says it. And I'm only on the second Google SERP.
Sometimes no source is cited for the number. Other times, the CDC is cited as the source without referencing the actual study. Surprisingly, even the CDC is guilty of this infraction.
Where articles using the number fail us, Google does not. If you Google 5,000+deaths+foodborne+illness, your answer appears at the top of the first page. I'm quite confident this paper, Food-Related Illness and Death in the United States, is the root source of the cited number. I've spend hours searching the Internet in general and the CDC's web site in particular and everything I've found points in its direction.
So, is the statement, "5,000 people in the U.S. die each year from foodborne illnesses," true or false? I'd call it true with caveats.
A strictly correct statement would say, "A 1999 study conducted by CDC researchers estimated that foodborne diseases cause approximately 76 million illnesses, 325,000 hospitalizations and 5,000 deaths in the United States each year."
How Good Is The Estimate?
Note that the number is an estimate, which is not to be confused with a statistic. The estimate is based on scant data combined with a number of assumptions that are outlined in the study and which the authors made with varying degrees of confidence. Also, while published in 1999, the study used data sets that began as early as 1982 and that ended as early as 1992. Further, some estimates also rely on results of studies conducted as early as 1948 through 1957.
That said, the CDC's methodology seems to be reasonably sound. I suspect that the estimate is about as good as could be expected given that paucity of available data.
While the CDC does not appear to have updated the study in the past decade, the Associated Press used data released by the CDC regarding foodborne illness in 2008 combined with "the CDC formula" (presumably the methodology used in Food-Related Illness and Death in the United States although the article does not explicitly say so) and current population numbers to estimate that 5,700 people died in the U.S. because of foodborne illness in 2008.
I did a rough cut based on FoodNet data and came up with an estimate of 4,700 deaths in 2008. Per FoodNet, there were 18,499 laboratory-confirmed cases of foodborne infection in a surveillance area with a population of 46 million people in 2008. The same numbers for 1996 were 9,787 and 20.5 million, respectively. This works out to an incidence of 47.7 cases per 100,000 population in 1998 and 40.2 in 2008. Note that these numbers understate the drop in incidence per 100,000 population because certain variations of E. coli were included in the 2008 data but not in the 1998 data. The population of the U.S. was 304 million in 2008 and 270 million in 1998. Assuming that mortality rates for those who became ill didn't change, I used ratios to adjust the CDC's estimate of 5,000 annual deaths (downward) for the change in incidence per person and (upward) for the change in population, arriving at the 4,700 number.
I tend to believe my estimate more than the AP's. In the AP's estimate, deaths grew faster than population, possibly indicating an increase in infection rates. This is inconsistent with the FoodNet data and with the CDC's statement, "there have been significant declines in the incidence of some foodborne infections since surveillance began in 1996," although infections rates (unfortunately) plateaued starting in 2004.
In a larger sense, it doesn't matter who's right because both estimates are 5,000 +/- 15%. Given the roughness of the estimates, it would be fortunate if any of them were within 15% of the actual number.
The bottom line is that the CDC's widely-cited estimate of 5,000 deaths per year appears to be solid and is still valid today.
So, if the number's good, what are its implications?
I have two concerns with the way the CDC's death due to foodborne illness estimate is being used currently. First, the context in which the number is now cited often implies conclusions that are not supported by the original research. Second, some policy recommendations made by those who cite the number to support their case are of dubious merit.
Context of Deception.
Numerous supporters of the local/organic/sustainable food movement, as well as advocates of increased government regulation of our food system, cite the 5,000 deaths due to foodborne illness per year figure to buttress their claims that America’s food is not safe. While doing so, they often place the number in a context that implies conclusions that are not supported by the original research. In particular, they do not acknowledge that foodborne illness and death can be caused by factors close to the fork, including improper storage and preparation in home kitchens, as opposed to those close to the farm.
For example, the Flash slideshow on S.T.O.P.'s homepage uses the number and, a few slides later, claims:
Foodborne illness [is] the result of emerging pathogens that have become increasingly virulent and antibiotic-resistant [and] the agricultural environment, where farm animal manure contaminated with pathogens causes irrigation water to infect our produce and crops [and] huge waste disposal problems.Yet, there is nothing in the CDC study that points a finger at these factors (or any other factors, for that matter) as the causes of foodborne illness and death.
Similarly, a Consumers Union press release headlined, “Foodborne illness victims mark President Obama’s first 100 days by calling for permanent fix to food safety system,” cited the CDC figure while stating:
Stopping [a foodborne pathogen] before it gets to our grocery stores and homes would save so much money and heartache.Again, nothing in the CDC study points to contamination upstream of the grocery store as being responsible for foodborne illness and death.
The FOOD, INC. web site states:
In Food, Inc. [sic] we meet Barbara Kowalcyk, whose 4 year old son died from E.coli poisoning after eating a hamburger. The Centers for Disease Control and Prevention estimates that 76 million Americans are sickened, 325,000 are hospitalized and 5,000 die each year from foodborne illnesses.But, the CDC study estimates that only 78 of these deaths are due to E. coli. And, as noted in my review of FOOD, INC., the actual source of Kevin Kowalcyk’s infection was never determined.
The 76 million people a year made ill by food figure is similarly abused. A Produce Safety Project fact sheet says:
• The Centers for Disease Control and Prevention (CDC) estimates that foodborne diseases cause approximately 76 million illnesses, 325,000 hospitalizations, and 5,000 deaths in the United States each year.One could easily view these bullets, one after another, and conclude that, adjusting for underreporting, there were 40 x 76 million = 3 billion cases of foodborne illness every year in the U.S. Nowhere does it mention that the CDC's numbers used a multiplier of up 38 in their estimation process, adjusting their estimates to compensate for underreporting.
• For every foodborne illness case that is reported, as many as 40 more illnesses are not reported or lab-confirmed.
These examples show how food safety advocates manipulate the CDC's solid estimates to manufacture a perceived crisis.
Policy Implications.
Obviously, every life lost prematurely is a tragedy. Much of the push for changes in food safety policy comes from people who have lost loved one to foodborne illness. My heart goes out to them.
However, policies have costs. Vast as our resources are in the United States, they are not limitless, as a quick look at our federal budget deficit shows. When examining policy changes, we must ask how much benefit will we get from the change, how much will the change cost and how do the benefits of a specific proposal compare to the benefits we could derive from other policy options with similar costs. Food safety improvement advocates have utterly failed to show that the costs of their proposals are worth the benefits. And don't give me the argument that saving even a single human life is worth it — if I can spend $X to save one life or $X to save 10 lives, the 10 should trump the one.
The bottom line is that, even at 5,000 deaths from foodborne illness per year, the American food system is amazingly safe. This number of deaths doesn't even come close to cracking the top 15 causes of death the U.S. Homicide, the 15th leading cause of death claimed 19,000 lives in 2006, almost four times the number of deaths from foodborne illness. In other words, eating is four times safer than walking down the street. This leads me to believe that, if we're going to invest additional societal resources (note, it's not just government resources) in saving lives, the dollars could be better used fighting other causes of illness and death.
Further, it's not clear that the policies that food safety activists advocate would make us safer, that they would save any lives. Take farm to fork tracing of foodstuffs, for example.
Here's a quick quiz: what's the leading cause of foodborne illness, according to the CDC? Salmonella? E. coli? Listeria? No, no and no. The leading cause of foodborne illness, according to the CDC, is unknown! That's right, folks, unknown.
According to Food-Related Illness and Death in the United States:
Known pathogens account for an estimated 14 million illnesses, 60,000 hospitalizations, and 1,800 deaths. ...Could someone please tell me what good farm to fork tracing is IF WE WON'T EVEN KNOW WHAT WE'RE TRYING TO TRACE??? And, in cases where the cause is unknown, it may turn out the food is not the culprit!
[U]nknown agents account for the remaining 62 million illnesses, 265,000 hospitalizations, and 3,200 deaths.
Further, nothing in the CDC numbers tell us if the pathogens contaminated the food in the early phases of the chain or final phases. If it turns out that poor storage and/or preparation near the fork causes most foodborne illness, a system to trace back to the farm is a waste.
Food safety is an important issue. Trust me, I don't want the food I eat to make me sick. But, it's time to stop the sensationalism and take a good, hard look at the true benefits and costs of additional government regulation of our food system.
42,000 pounds of beef tainted with e.coli 0157:H7 was recalled since you wrote this article - a Class I recall no less, the USDA FSIS' highest level.
I dont know if you are aware of this but the FDA/USDA have no authority to recall food either. The beef recall above was at the company's discretion, just as the Nestle cookie dough was the week prior.
I'm not sure if the 5000 number is right but if that number of deaths from food could be 4000, 1000 even 500, shouldn't we pursue that goal?
You should check out the review we did on the latest food safety bill - (http://nutritionwonderland.com/2009/06/food-safety-enhancement-act-hr-2749/). There is will almost no cost to the government to implement the safety measure and the bill has been endorsed by many small farmer advocacy groups.
I understand your feelings towards compromised motives - and I share them - but enacting food safety legislation seems like a no brainer.
Reply to this
As you note in your review of the latest food safety bill, "poisoning your customers is bad business." However, you don't note the implication of this -- that, even in absence of government authority to force food recalls, producers have a strong motivation to 1) avoid contamination in the first place and 2) pull tainted products off of the shelf as quickly as possible. These motivations are enforced by the market. Note for example, the that recent pistachio recall was trigger not by government inspection but by routine quality testing at Kraft. (See http://www.foodnavigator-usa.com/Financial-Industry/Snack-supplier-repackaged-recalled-pistachios-warns-FDA until I can get embedded links working in this comment.)
So, yes, I am aware that FDA/USDA have no authority to recall food and yes, I support those agencies having mandatory recall authority for cases of confirmed contamination provided that an appeal process is available to producers who believe that they have been unjustly told to recall a product. But let's also recognize that the current system works rather well.
Should the government seek to decrease the number deaths due to foodborne illness even if, say, only 500 people a year die from them? Heck, no!
Again, we as a society have limited resources. Why devote those resources to saving fewer lives at high marginal cost per life when the same amount of money, time and effort could save more lives at lower marginal cost per life if devoted to more prevalent causes of death? Net, you're not saving lives, you're taking them.
And the cost isn't just money and money isn't just the government's. It's also things like consumer choice and quality of life. For example, if we burden farmers, food processors and food distributors with restrictive and costly new regulatory requirements, how many artisan food businesses will close? How many traditional techniques will be lost, as has already happened in Europe (see http://www.newrules.org/agriculture/article/setting-slow-table)? How many prospective food entrepreneurs will not found companies to bring new food products to market?
While advocacy groups for small farmers may support the current legislation, I suspect that they represent largely farmers who sell mainly through farmers' markets and are exempt from the bill's traceability requirements. Sorry, but farmers' markets aren't going to feed the nation. And the more exemptions we make, the less effective is traceability and the more the market tilts toward those who have effective lobbying efforts as opposed to those who actually meet the market's needs.
So, no, I don't think that enacting new food safety legislation is a no-brainer. Far from it.
Reply to this
I would like to know which "small farm advocates" have endorsed this bill?? Names of the organizations please?? I do not know any small farm from east coats to west coast that believes this bill will do anything but make our food supply LESS safe.
Also, food companies do NOT need a mandatory recall. It is bad for business to make people sick. Any business will voluntarily recall and have for decades.
Perhaps you believe mandatory recall would have made us safer when the FDA thought tomatoes were making us all sick last year. Oh, wait...they made a mistake. It wasn't the tomatoes after all, was it. Would have put a lot of tomato farmers out of business though, huh? Actually, it did anyway.
Do not push for a bill you have little knowledge about. This bill is junk used to make simple minded Americans believe that politicians want to "do good for the neighborhood".
Reply to this
One more thing regarding..."There is well almost no cost to the government to implement the safety measure..."
Fact check: Yes, your are right. Little upfront cost to the government to implement. Seriously high cost to the consumers of the US and to the processors of the US.
I am a small livestock processor. I process livestock under the USDA mark of inspection for consumers who want to fill their freezer with local meats from a farmer and for the small family farmer who wants to offer his/her farm raised meats to the public.
If this bill is enacted, my business will close it's doors and many more like me across the US. We will be charged for USDA services. This is NOT a fee a small butcher can absorb.
So, instead of making the food supply safer...you ran the small business and small farmers OUT of business and put the entire food supply of our country into the hands of the very large FEW corporations.
Congrats!
Reply to this
Grumpy-
Very very few small farms would close because of this legislation. If you sell only direct to consumer, you are exempt from this. If you are deemed a 'food facility' you get inspected every 4-5 years at a cost of $500. These are lenient terms to say the least, not to mention that I have to think the FDA is losing money on each of these inspections.
Curiosity:
The small farmer advocacy organizations who have endorsed this bill are:
-Consumers Union (https://secure.consumersunion.org/site/Advocacy?cmd=display&page=UserAction&id=2135)
-Organic Consumers Association (http://www.organicconsumers.org/articles/article_18368.cfm)
- Maine Organic Farmers and Gardeners Association (MOFGA) - (tentatively: http://mofga.org/Programs/PublicPolicyInitiatives/MOFGAPositionStatements/FoodSafety/tabid/1102/Default.aspx?PageContentID=399)
Thats just a few - there are more. It has received broad support from a wide array of consumer and farming advocacy groups.
Also Curiosity, as a livestock producer you should know you are regulated by the USDA. HR 2749 does not even touch the USDA FSIS service- this bill only deals with the FDA. So, for you, this entire debate is academic because nothing will change on your farm.
I would also like to add that the size of the initial recall I posted on friday has now been increased by 10X to 400,000 lbs of beef: http://www.marlerblog.com/2009/06/articles/case-news/are-jbs-swift-e-coli-recalls-linked-to-illnesses-in-arizona-california-colorado-florida-illinois-michigan-minnesota-nebraska-oregon-south-carolina-tennessee-utah-and-wisconsin-canada/index.html.
Reply to this
John, the costs to farmers go beyond inspection fees. Of course, farmers SHOULD bear the costs associated with producing goods that are safe to consume, regardless of whether there are governments inspections. However, depending on how the regulations implementing laws are written, new legislation can impose huge costs in terms of things such as record keeping. Ask any start-up trying to go public how much implementing Sarbanes-Oxley costs them. The usual answer is $1,000,000 per year. That's money they'd rather spend on innovation instead of record keeping. Because there are often scale economies associated with systems requried to implement government regulation, I fear that additional rules tilt the playing field toward the big guys, endangering small players and, perhaps more importantly, deterring innovative new players from entering the market.
Thank you for the links to the small farmer advocacy orgs. It always helps to have some examples.
Reply to this
Really interesting analysis. Thanks, Grump. One of your best.
But shouldn't we do something, now, anything, about the "walking down the street" crisis? It's deadly.
Reply to this
I think I'll attack left turn problem first. Did you know that 2,800 people a year die from accidents while turning left? (http://www.nhtsa.dot.gov/people/ncsa/809-100.pdf) If the government simply banned left turns, we could prevent a couple of thousand deaths each year!
Reply to this