Lessons From Investigations of Foodborne Disease Outbreaks

KEYWORDS:
• dairy products
• disease outbreaks
• food poisoning
• pasteurization
• salmonella food poisoning

In this issue of THE JOURNAL, Cody and colleagues¹ describe investigations of 2 overlapping outbreaks of salmonellosis that affected residents of northern California; Villar and colleagues² describe a similar epidemic in Washington State. All 3 outbreaks were traced to the consumption of Mexican-style soft cheese (queso fresco) made from unpasteurized (raw) milk. These 2 articles provide vivid examples of the importance of outbreak investigation, as well as the challenges to reducing the incidence of foodborne illness.

The vast majority of enteric infections reported to public health agencies are not linked to recognized common-source outbreaks. Even the proportion that are foodborne is not known with any certainty, although it certainly varies by disease from fairly low (eg, giardiasis) to fairly high (eg, salmonellosis). Most acute gastroenteritis is never specifically diagnosed or reported.

Although involving only a minority of case reports, outbreak investigations contribute disproportionately to the understanding of the transmission of enteric pathogens because, in general, only for outbreak cases can a specific source be confirmed. Without multiple case histories to compare for common exposures, individual "sporadic" cases usually remain of unknown origin.

Outbreak investigations are triggered in many ways (eg, by a citizen's complaint of multiple illnesses following a group meal or by a report from an alert physician who notices "seeing a lot of people with diarrhea recently"). Not all outbreaks are localized, however, and even localized outbreaks may not result in a critical mass at a single health care facility. Many clusters, including those reported by Cody et al and Villar et al, are first detected by public health workers collating routine laboratory and physician reports. Suspicions of a common source may be bolstered by demographic similarities among patients (as in these outbreaks, in which most patients were very young, predominantly Hispanic, and geographically concentrated).

There are many reasons to investigate foodborne outbreaks. Most obviously, the detection of ongoing problems can result in direct action to prevent additional exposures and disease. When Escherichia coli O157:H7 case counts suddenly increased in early 1993, investigators in Washington State quickly pinpointed deficient cooking practices at a fast food chain as the culprit. The restaurants closed, and more than 250,000 hamburger patties from the implicated lot were recalled, preventing hundreds of infections.³

In addition, many investigations stimulate follow-up research. Risk factors identified through outbreak investigations are probably similar or identical to those that explain most sporadic cases, and the relative importance of these risks can be explored with case-control or other analytical studies. Field investigations also often provoke important laboratory studies, such as measuring the proliferation of salmonellae in germinating alfalfa sprouts⁴ or identifying the survival ofE coli O157:H7 in apple cider.⁵

Although the effect is less immediate, good investigations, particularly if published, add to general knowledge about how enteric pathogens are transmitted. Such understanding not only facilitates future investigations but is essential for developing effective long-term control strategies. For example, alfalfa sprouts were rarely considered a vehicle for disease transmission before Scandinavian epidemiologists published a brief note about a 1994 outbreak.⁶ Only a few years (and several outbreaks) later, sprouts are considered one of the "usual suspects," enabling much more rapid identification and control of outbreaks when sprouts are the culprit, and growers are being pushed to develop safer means of producing their products.

While there are always new twists, and new pathogens are identified with distressing regularity, a great deal has been learned about the foodborne transmission of many enteric diseases. Much of it boils down to this unpleasant truth: food is often contaminated with feces—animal or human—that sometimes contain human pathogens. Milk is, meat is, poultry is, fruits and vegetables are. Although better methods can be and are being developed to reduce the risk and degree of microbial contamination, some level of contamination is likely to remain the rule, not the exception. While fasting may be the only way to eliminate all risk of foodborne disease, proper food handling to minimize cross-contamination (and, for bacteria, amplification), and adequate processing whenever possible to remove, inhibit, or kill pathogens (eg, by heating, irradiation, or acidification) can greatly reduce it.

Translating this reality into policies and practices that reduce the incidence of disease is a surprisingly difficult challenge. Public agencies must pursue both educational and regulatory programs, buttressed by a commitment to research and disease surveillance, including outbreak investigation.

Public education is essential. Presentations at scientific meetings and articles in peer-reviewed journals rarely have a wide impact by themselves. In 1975, for example, investigators traced a large, multistate outbreak of salmonellosis to undercooked hamburger.⁷ "The hazards of ingesting raw ground beef are not generally appreciated," the authors noted. Yet, despite several subsequent meat-associated clusters, few paid attention until several children died and hundreds were sickened in a single outbreak 18 years later.³ Inevitably, information about foodborne disease is filtered by a host of public and private institutions, including the mass media, and must compete with other health news and current events. Unfortunately, dramatic outbreak stories seem to be the only reliable way to reach an oversaturated audience. Even then, bitter lessons can be quickly forgotten.

Some messages demand a narrower focus. Food safety guidelines for immunocompromised persons are an example. Following the salmonellosis outbreak in Washington,² local public health officials worked with consumer education specialists and community members to develop innovative educational programs, targeted at the Hispanic community, that emphasized safe ways to produce queso fresco. Significantly, efforts were directed not only at consumers, but at food producers; results after 1 year were encouraging.

Recommendations do not always reach their target audiences. Despite a steady stream of salmonellosis and E coli O157 outbreaks linked to alfalfa sprout consumption in recent years,^{6, 8-9} many consumers remain unaware of the problem. The US Food and Drug Administration recently advised elderly and immunocompromised persons not to eat alfalfa sprouts.¹⁰ But has anything been done to get that message to the hospital dieticians, nursing home cooks, and others who prepare food for the individuals at highest risk?

However, not everyone who hears the message is convinced. Even after almost 100 years of effort, medical and public health experts have been unable to eliminate raw milk consumption. Raw milk has been and continues to be a staple in the epidemiological literature, linked to a long list of diseases, including campylobacteriosis, salmonellosis, tuberculosis, brucellosis, hemorrhagic escherichiosis, Brainerd diarrhea, Q fever, listeriosis, yersiniosis, and toxoplasmosis, to name a few.^11-13 The outbreaks described by Cody et al and Villar et al add yet another agent to the list: DT104 phage group Salmonella enterica serotype Typhimurium. DT104 salmonellae have been an object of increasing alarm as they have spread rapidly through Britain and the United States over the past 10 to 15 years.¹⁴ Most DT104 organisms are resistant to a broad range of antibiotics, complicating treatment of patients with invasive disease, and they may be more likely to cause severe disease than most salmonellae.¹⁵

There is no mystery about why raw milk is a common vehicle for salmonellosis and other enteric infections; after all, dairy milk is essentially a suspension of fecal and other microorganisms in a nutrient broth. Without pasteurization or other processing to kill pathogens, consumption of raw milk is a high-risk behavior. Although aging and drying renders some cheeses made from raw milk safe, fresh cheeses made from raw or imperfectly pasteurized milk—including Mexican-style soft cheese—are likewise well-documented hazards.^16-20 Raw milk and raw-milk products will continue to cause morbidity and mortality until people stop consuming them.

Regulation can be a valuable complement to food safety education. Science-based regulations protect public health by establishing a level playing field for industry, discouraging ignorant or unscrupulous operators from cutting corners to increase profits. New regulations are helping the meat industry make significant changes in the way animals are slaughtered, butchered, and processed; the result is a safer product. At least 22 states restrict or prohibit the sale of raw milk, and in those states milk-associated disease outbreaks are much less common than elsewhere.¹³

The effectiveness of even well-crafted regulations can be undercut by inadequate enforcement or a lack of public support. The DT104 outbreaks in California¹ and Washington² could have been prevented had existing laws that prohibit bulk sales of raw milk to unlicensed processors been observed. There is no indication in these reports that the dairy operators who flouted these laws were held accountable. The cheese was made by unlicensed operators who flagrantly disregarded basic standards of hygiene. California authorities estimated that, far from being the exception, similar operations supply more than half of the huge market for Mexican-style soft cheese in that state. Some consumers may prefer cheese from these sources because it is cheaper or because they think it tastes better; they may not appreciate the risks to themselves and their families.

However fitfully, great progress has been made in reducing the overall burden of foodborne disease. Maintaining and extending these advances requires a public health infrastructure that is flexible enough to meet new challenges and respond to old ones that will not go away. Most investigations begin as locally recognized clusters that are investigated locally. However, even small outbreaks can have wide-ranging implications, and local inquiries sometimes take unpredictable turns, as when product tracebacks connect independent investigations²¹ or pathogen subtyping facilitates the discovery of hidden epidemics.⁹

Food safety is everyone's concern, and everyone has a part to play. Routine surveillance works best when clinicians and laboratorians make specific diagnoses and report cases promptly. Good investigations require trained and dedicated public health workers and adequate support resources. Food safety experts must expand efforts to work with producers, processors, and consumers all along the "farm-to-table" continuum to exploit the knowledge gained about the transmission of foodborne diseases—knowledge sometimes gained at tragic cost.