During April and May 2000, epidemiologists, laboratory staff, and environmentalists from the Tennessee Department of Health presented a series of lectures to public health personnel throughout the state to review the process of investigating foodborne illness outbreaks. Participants included epidemiologists, public health nurses, laboratory staff, and environmentalists.

Before each session, participants were asked the following question: "What are the three most common pathogens causing foodborne illness in the United States?" Verbal instructions included clarification that the question referred to which pathogens were numerically the most frequent causes of illness. Participants ranked their top three answers in writing and submitted them to the course director. Responses were anonymous, although the job category of each respondent was collected. Data were entered and analyzed by using EpiInfo software (7).

Of 553 attendees, 388 (70%) participants responded to the survey. Respondents included 128 environmentalists, 233 public health nurses, 11 health department physicians, 4 laboratorians, and 12 persons in other positions in the health department. The proportion of participants was representative of the proportion of responders to foodborne illness within the health department.

Ninety percent of persons listed Salmonella among the top three most common causes of foodborne illness in the United States; 56% listed Escherichia coli; 36% cited Staphylococcus; and 32% Shigella (Table). Other commonly cited causes of foodborne illness included Campylobacter, Listeria, Hepatitis A virus, and Clostridium.

Only 5% of respondents listed Norwalk-like virus (NLV) among the three most common causes of foodborne illness, and an additional 4% noted "viruses" more generically. Only 4% of respondents listed NLV or viruses as the most common source of foodborne illness.

Results did not vary significantly by job category of the respondents. Public health nurses, environmentalists, and physicians, for example, all listed Salmonella, E. coli, Staphylococcus, and Shigella as the most common causes of foodborne illness. Persons from all job categories were represented among the 9% of respondents who listed viruses or NLV among the top three causes. No job category was statistically more likely to identify viruses as a common etiology, and in no group was NLV among the five most commonly listed pathogens.

The four pathogens most commonly believed by the survey respondents to be among the major causes of foodborne illness (Salmonella, E. coli, Staphylococcus, and Shigella) are actually estimated to account for <13% cumulatively of all foodborne disease in the United States (Table). Recent estimates suggest that the most common causes of foodborne illness in the United States, in decreasing order of frequency, are NLV, Campylobacter, Salmonella, Clostridium perfringens, and Giardia (1). Only 5% of respondents listed NLV among the three most common causes of foodborne illness; this agent is estimated to cause 67% of all foodborne disease in the United States (1). In contrast, Listeria, which causes <0.1% of foodborne illness in the country, was believed by 15% of respondents to be among the three most common causes of disease.

The response to a suspected foodborne illness may differ, depending on the likely etiology. Basic methods of case-finding, hypothesis-generating, and investigating exposure histories do not necessarily require knowledge of the frequency of possible pathogens. Other issues, such as stool collection and testing techniques, treatment and follow-up, and preventive recommendations may differ greatly depending on a particular pathogen. If the personnel commonly responsible for recognizing, reporting, and intervening in foodborne illness are unfamiliar with common pathogenic causes of such illnesses, the appropriateness of their responses may be compromised.

Reasons for the discrepancy between perceived and actual etiologies of foodborne illness are unknown. While estimates that two-thirds of foodborne illnesses are caused by caliciviruses may be debated, perceptions of study respondents reflect neither national estimates nor recent experience in Tennessee. Some pathogens incorrectly believed to be common causes of foodborne illness, such as E. coli and Listeria, cause relatively severe disease, which often generates substantial media attention. Highly publicized outbreaks and severe cases may disproportionately affect perception of a pathogen's incidence. Such factors might be expected to influence public perception more than that of health-care workers, although this study suggests otherwise. There is no evidence that the public health personnel we surveyed have a substantially different understanding than health-care workers elsewhere. Studies on factors that affect both academic knowledge of the causes of foodborne illness, as well as factors such as severity and risk (which likely strongly influence perception of their relative importance), would be of value.

While it is true that the etiology is not identified in a large proportion of foodborne illnesses, lack of knowledge on the part of public health personnel is only one barrier to improving this situation. Lack of resources, competing priorities, the health-seeking behaviors of ill persons, and the activities of clinical and laboratory providers all have important effects on the response to suspected foodborne illness. Despite that, this study suggests that public health personnel on the front lines in responding to foodborne illness have incorrect perceptions of its causes. If this substantial public health threat is to be effectively addressed, appropriately educating the persons relied upon to address the problem is necessary.