How safe is our food? Put another way, how much illness in the United States is caused by foodborne pathogens? It sounds like a simple question. Getting a reasonable answer, however, is far from simple. The basic problem lies in the fact that only a small fraction of foodborne disease cases get reported through official (or unofficial) reporting systems. Calculating the “real” rate of foodborne illness requires development of models that use reported cases as a starting point to estimate underlying disease rates. Given the plethora of pathogens that can be transmitted through foodborne routes, this is a complex, and somewhat daunting, process. It is, however, necessary for assessing the safety of foods and developing strategies for disease prevention.

Morris JG Jr., "How safe is our food?," Emerg Infect Dis. 2011 Jan

The CDC has again undertaken the difficult task of tallying the annual incidence of foodborne disease.  Today, CDC announced two studies on the subject; one for 31 major foodborne pathogens, and another that assesses incidence, hospitalization, and deaths caused by foodborne transmission of "unspecified agents."  Combined, the known pathogens and unspecified agents cause approximately 47.8 million foodborne illnesses, 127,839 hospitalizations, and 3,037 deaths per year. 

These studies follow the 1999 study by Mead et al., which found, using some similar and some different methodologies and classifications as the current studies, that known and unknown foodborne pathogens caused approximately 76 million cases of foodpoisoning annually, 325,000 hospitalizations, and 5,000 deaths. 

The "unspecified agents" study pertains to "agents that cause acute gastroenteritis but that were not included in our estimate of foodborne illness caused by 31 major known pathogens."  More specifically, "unspecified agents" includes (1) illnesses likely caused by one of the major known pathogens, but where there was insufficient data for an agent-specific diagnosis; (2) known agents not yet recognized as causing foodborne illness; (3) microbes, chemicals, or other substances known to be in food but for which pathogenicity is unproven; and (4) agents not yet described. 

The study finds that, of the 47.8 million total annual cases of foodpoisoning in the US, "unspecified agents" account for 38.4 million illnesses, 71,878 hospitalizations, and 1,686 deaths.

The study states that the approximate 33% decrease in the number of illnesses due to nspecified agents since Mead’s 1999 study, which found approximately 62 million illnesses annually, is likely not due to any decrease in the actual rate of illness.  The current study used a 24% lower estimate for "gastroenteritis" than the 1999 Mead study, excluding illnesses where a concurrent cough or sore throat was reported, where less than 3 episodes of diarrhea or loose stools were reported in a day, and where people reporting vomiting had been ill for less than 1 day or whose illness did not result in restricted daily activities (the1999 estimate included all persons with vomiting).