A few weeks ago I was looking at the 2009 FoodNet data and started thinking about the upcoming release of new numbers on the burden of foodborne illness in the United States.  As you might recall in 1999 Mead and colleagues published estimates of the annual burden of foodborne illness in the U.S. These estimates – 76 million ill, 325,000 hospitalizations, and 5,000 deaths – have been cited by government, academics, citizen groups, and even industry as a call for action to improve food safety nationally and internationally. Ten years ago, who anticipated the deluge of “Monday night quarterbacks” today eager to criticize a sincere attempt to characterize a serious public health and individualal health problem in our country? Yes, it is time for revised estimates and I, along with others, eagerly anticipate release of the new estimates rumored to occur sometime this next week.

cdc logo.jpgUndoubtedly, the revised estimates will more accurately reflect truer incidence of foodborne illness. Credit must be given to public health that has strengthened detection, surveillance, and investigation of illnesses and outbreaks at the local, state and federal level. I can only imagine the obstacles that were overcome for implementation of electronic laboratory reporting or establishing NORS, the national outbreak reporting system. Data from both are likely to be used to calculate the forthcoming estimates.

Mead et al. were clear as to the limitations of the data sources used in their seminal 1999 article. The authors described several assumptions that were made such as uncertainty in the degree of under-reporting and the frequency of acute gastroenteritis in the general population. For example, in 1999 few public health laboratories could test for norovirus in food or stool. Mead et al. attributed 30% (23 million) of the estimated 76 million illnesses to norovirus. Thanks to better laboratory detection and understanding of how norovirus spreads, the percentage of norovirus illnesses attributable to food in is likely to drop considerably in the new estimates. A better norovirus estimate ten years after 1999 does not imply that Mead et al. were careless in their methods. Rather, it highlights advancements in basic sciences and public health.

Recently, there has been no shortage of critics willing to dispute the 5,000 deaths Mead et al. attributed to foodborne illness. In June 2010 the Foodborne Disease Active Surveillance Network (FoodNet) published preliminary surveillance data for 2009 showing 74 deaths caused by the 10 foodborne pathogen it tracks. FoodNet sites represent about 15 percent of the US population which extrapolates to about 500 US deaths caused by foodborne pathogens in 2009, considerably lower than the 5,000 deaths estimated in the Mead article. However, the 500 estimated deaths would not include patients who died from a foodborne illness but who were not laboratory confirmed with a pathogen. Furthermore, I know from personal experience that patients die from injuries caused by foodborne illness after public health involvement has ceased. These patients are not counted as food related deaths. Using death certificate data to identify foodborne illness related deaths is problematic. For example, the certificate might list renal failure as the cause of death with no mention of the shigatoxin E. coli infection the decedent had two months prior. This patient also would not be counted in food death data.

The effort to estimate the “true” number of illnesses, hospitalizations, and death associated with foodborne illness requires use of many data sources and statistical adjustments to account for under-reporting and other unknown elements. We need the best data we can get to use to make decisions on how and where to fight foodborne illnesses. 

The new estimates will certainly be scrutinized and used to promote different agendas. Remember, the data is only as good as the surveillance.  We need to continue to support local, state and federal public health.  They need the resources to keep us all healthier.  I also encourage everyone to pause and remember that the estimates represent a compilation of individuals – hundreds and thousands of people I have met personally – my clients. Thankfully, most individuals will experience mild symptoms and think of their illness as a minor inconvenience. For an unfortunate few, foodborne illness will mean physical and mental anguish, financial expenditures, long-term consequences, or death.

Real people; real consequences….

  • keith Warriner

    I always thought the 73m value was a gross over-extimate of foodborne illness. No doubt the new figures will be lower, probably by a factor of 10. Yet, the important point to realize is the potential impact of foodborne illness on the population and industry. In this respect, it is questionable of what value the actual foodborne illness levels are- except for an nice opening line in a paper.

  • I’m going to have to change all my power points.

  • Gabrielle Meunier

    I have to wonder how you can calculate the 76 million number? But it would seem to be true. How often do people get a badly upset stomach from something they ate (or the “flue” symptoms)? Its unfortunate and happens a lot. But how would that ever be officially quantified? So many Doctors (or the patients who’ve been told to bring in a stool sample, don’t) do not do stool samples, so again even official numbers are going to be remarkably lower than actual incidence. Then of course the deaths, as with any illness, what gets you in the end is usually the organ failure which is the “cause” of death. Oh but guess what, the organ failure was cause by the root illness. In the elderly, however, how many “root” illnesses might they have? Difficult to then quantify. The numbers we’ve been using seem reasonable to me.

  • I think the big change will be Norovirus. In 1999 it was assumed that more of it was food borne.

  • Gabrielle Meunier

    I still believe C-diff is food borne! If anyone ever sees anything on this (Clostridium Difficile (C-diff) and relations to foodborne illness) I sure would like a note! Also as it pertains to counting numbers, aren’t the only numbers we have going to be based upon the reportable bugs? What about all the other bugs that are not on the reportable list? Do those bugs get estimated into the statistics?

  • It all is a moving target I’m afraid.

  • John Munsell

    I don’t care if the corrected, updated figures are one half or one tenth of the estimated statistics published in 1999. Regardless of what the new “ESTIMATES’ will reflect, the bottom line remains our resolve to strive for improvements, to leave no stones unturned to (a) detect the source of pathogens, and (b) to require corrective actions AT THE SOURCE. Instead of 76 million ill, if the new statistics only claim 15 million ill, as an example, do we then claim “success”, pat ourselves on the back, and maintain the status quo? The existing USDA/FSIS attitude is that meat companies have invested hundreds of $ millions into interventions (true), and that our food is the safest on the globe (I’ve not seen proof). Even if it’s true, how do we address the sickened 15 million consumers, some of whom die?
    Frankly, food safety takes the back seat to this globe’s focus on unfettered global trade, void of any artificial restrictions which would promote public health and food safety. Ultimate power in this global focus resides within the multinational companies, which lobby (“purchase”) our policy makers’ compliance to Free Trade protocol.
    John Munsell

  • Ralph Weber

    We start with the truth (as close as science can get us to it, anyway) and go from there. As we sharpen the focus on the problem, we hope the solutions become clearer.

  • Perhaps anecdotal, but I think one thing I believe, is that the beef industry in the US is doing a far better job in 2010 than it did in 1993. For that, the clients that I did not have and I are thankful.