The Norwalk Virus: A Guidance Template for Local Environmental Health Professionals.
According to new statistics from the Centers for Disease Control and Prevention, Norwalk-like, calicivirus-borne disease, a form of nonbacterial gastroenteritis, is the most prevalent foodborne illness in the United States, The Norwalk-like viruses and other small, round-structured gastrointestinal viruses are discussed as a group in this paper under the rubric "Norwalk virus." The Hamilton County Health District in Ohio confirmed four foodborne-disease outbreaks associated with the Norwalk virus over a three-year period from April 1997 to September 1999. Two of the outbreaks were traced back to ill food handlers. The purpose of this paper is to share the lessons learned from confirmed Norwalk virus outbreaks and provide insight into the prediction, detection, and prevention of outbreaks.
Norwalk-like, calicivirus-borne disease, a form of nonbacterial gastroenteritis, is the most prevalent foodborne illness in the United States . It is estimated that in the United States, Norwalk-like viruses cause 23,000,000 of the 38,629,641 annual foodborne-illness cases associated with known pathogens . (Annual illnesses caused by a combination of bacterial and parasitic pathogens are estimated at 7,746,250. Rotoviruses, astroviruses, and hepatitis A account for the remaining 7,883,391 annual estimated cases ). The Norwalk-like viruses are virtually ubiquitous in the environment, and future research may reveal them to be interspecies zoonotic agents [2,3]. They are common constituents of human excrement and sewage, and cross-contaminants of food and water. The virulence of the pathogen, gauged by the infective dose, is prodigious: 10 to 100 particles can cause infection in humans .
Viruses are less than 0.1 microns in size. The Norwalk-like viruses and other small, round-structured gastrointestinal viruses will be discussed hereafter as a group, under the rubric "Norwalk virus." The pathogen is an obligate parasite dependent upon live animal-host cells. Because the virus is identified by electron microscopy and genetic amplification techniques, identification of outbreaks historically has been difficult. The purpose of this paper is to share some lessons learned from four confirmed Norwalk virus outbreaks and to provide insight into the prediction, detection, and prevention of outbreaks.
The Norwalk virus was first recognized as the agent of an epidemic in Norwalk, Ohio, in 1968. The laboratory capabilities necessary for detecting viruses generally were not available at the time, and researchers could only speculate that the outbreak was viral in origin because bacteriological and protozoal stool sample tests were negative. When volunteers became ill after being fed extracts of stool specimens from symptomatic patients, indirect evidence of a viral epidemic source was obtained. The Norwalk virus was later observed under an electron microscope in the early 1970s. In the 1990s, molecular cloning techniques allowed for easier virus identification.
Although the Norwalk virus was "discovered" in Ohio, the primary calicivirus reservoir is the sea . Sewage directly discharged from municipal treatment plants or dumped overboard by oyster harvesters, recreational vessels, and oil rigs has resulted in oyster bed contamination . Generic Norwalk virus control literature typically suggests that when foods and beverages other than shellfish are implicated, they usually have been contaminated by ill food handlers. Sewage-contaminated ice and contaminated water wells, however, have resulted in numerous recorded outbreaks. Potential viral cross-contamination pathways in food and water are unlimited.
Viruses are obligate parasites. They require a viable host cell for growth and do not replicate on foods; viruses do not technically "live." By contrast, bacteria generally require some time, a favorable temperature climate, and an appropriate growth medium before amplification adequate for infection can occur. Unlike bacteria, viruses can remain viable for long periods of time on food. The lowness of the doses required for infectivity contributes to the high frequency of disease.
The mode of disease transmission is primarily fecal- and vomit-oral. Some evidence for the vomit-respiratory mode has been collected from cruise ship outbreaks in which shared passenger rooms were the only discovered link between outbreak victims. Secondary transmission between relatives and intimate partners of symptomatic patients is common. The disease vehicles are water, ice, shellfish, food contaminated by an infected food worker, food and beverages contaminated by sewage, aerosol droplets, and contaminated environmental surfaces. Table 1 presents outbreak characteristics associated with the Norwalk virus.
Foods traditionally associated with the Norwalk virus are "ready-to-eat foods" such as salads, frostings, raw shellfish, sandwiches, fruits, deviled eggs, and bakery items. Any food, however, can be contaminated by the virus. Although the virus is heat labile, time/temperature inactivation is certain only at boiling (100[degrees]C) .
Over a three-year period from April 1997 to December 1999, the Hamilton County General Health District (HCGHD), located in the southwest corner of Ohio, implicated the Norwalk virus in seven separate outbreaks of foodborne illness. The outbreaks originated at a daycare facility a country club, a restaurant, and residential sources. The illness is self-limiting but can be severe. No deaths resulted from the seven confirmed outbreaks. Four of the outbreaks, outlined below, are representative of the Hamilton County Norwalk virus experience.
Outbreak I-March 3, 1997
A sandwich shop catered a lunch for an industrial manufacturing facility Seventy-eight of the 173 people who ate lunch meat sandwiches met the case description. On the first day of the investigation, an ice machine was discovered to be inadequately protected from sewage back flow from a sewer line (no "air gap").
A 1995 outbreak involving 500 people at a convention in Columbus, Ohio, had been determined to be caused by sewage-contaminated ice obtained from an ice machine directly connected to a sewer. A "grab bag" of pathogens was cultured from four patients--including Giardia lamblia, Blastocystis hominis, and Clostridium difficile--confounding the course of the Columbus investigation. Later it was theorized (without laboratory confirmation) that the Norwalk virus was the major cause of illness.
After the March 1997 outbreak, however, analyses of the ice-machine condensate line revealed no pathogens. The catered sandwiches were determined to be the source of illness.
A comparison of DNA sequences through polymerase chain reaction (PCR) genetic-material amplification techniques, performed by experts from the Centers for Disease Control and Prevention (CDC), directly linked the stool samples of two food handlers to the stool samples collected from outbreak victims . One of the two food handlers admitted to gastrointestinal symptoms but insisted that the symptoms had cleared up completely four days before preparation of the sandwiches, Outbreak investigation activities included the determination that a food handler had shed the virus 10 days after symptom cessation or that an asymptomatic food handler had transmitted disease . It was not possible to determine the point at which the food handlers became infected with the virus. These findings complicate the use of sanitary controls that exclude sick food workers.
Outbreak II--August 18, 1997
Sixty-three football players from a local high school were reported ill with gastrointestinal symptoms. Two food service operations and several parents had provided lunches and dinners to the athletes during rigorous training sessions in the summer. An immediate environmental investigation of the practice area revealed a complex array of plumbing cross-connections, including the playing-field irrigation system and drinking-fountain networks. Locker room urinals were not equipped with vacuum breakers, and hose bibs were discovered submerged in muddy water. Water samples were collected, and a local laboratory performed a 24-hour acute toxicity screen with live Ceriodaphnia dubia minnows. All of the minnows died. Terse conversations with the laboratory showed that the water had not been dechlorinated prior to the test. Oocyst and parasite (O&P) tests also were performed. The plumbing deficiencies were corrected immediately, and the O&P tests would later show negative test results.
Nine of 27 stool samples collected for analysis tested positive for the Norwalk virus by PCR analysis. Food preparation employees tested negative for the Norwalk virus. Sandwiches obtained from food service operations had, however, been cut and further prepared by parents of the athletes. Foods prepared by parents in residential kitchens also were suspect. Stool samples were not obtained from uncooperative parents, and representative food samples were not available for analysis. Staggered incubation times suggested person-to-person transmission among athletes and family members. The source of the outbreak was determined to be residential; contaminated foods likely transmitted the disease from an ill parent to members of the team.
Outbreak III--June 10, 1998
At a licensed daycare facility, 27 people, out of a total of 92 students and staff, were reported ill. Twenty-one students, three to 12 years of age, fit the case description. Six adults, including parents of students and non-food service employees, were determined to be secondary cases. All stool samples tested negative for a broad spectrum of bacteria. Six of 27 stool samples tested positive for the Norwalk virus. Four students tested positive for rotoviruses, and one student tested positive for astroviruses, which indicates the presence of concurrent infections. Representative food samples were not available for viral-testing processes. All students and adults positive for the Norwalk virus either had consumed chicken nuggets at the school or had come into close contact with an ill student who had eaten the chicken. Although the primary food handler tested negative for the Norwalk virus, a part-time food handler was determined to be an asymptomatic positive carrier. Clearly, a single infected student or an infected food worker can cause a cascading outbreak in close quarters such as a daycare facility
Outbreak IV--September 9, 1999
An anonymous call alerted the Environmental Health Division that 78 people had become ill from a meal served at an elite country club two weeks before the call. An immediate investigation revealed that the country club management had allowed a sick chef to continue to prepare ready-to-eat foods, including salads and lunch-meat sandwiches that were determined to be the source of the outbreak. Six of 10 stool samples collected were positive for the Norwalk virus. Because the ill chef was well past the point of shedding pathogens, a stool sample was not collected. Three points were stressed with the country club board members:
1. the necessity of implementing a policy of sick worker exclusion,
2. the necessity of training for all employees (provided by HCGHD), and
3. the necessity of immediately contacting HCGHD should employees or customers report illness at any time in the future.
Predictive pathogen identification can begin immediately as information is collected during the initial phases of a foodborne illness investigation. Norwalk and other viral epidemics are usually identified by theoretical and analytical exclusion of possible bacterial and protozoal sources. By interviewing several people for a basic case-control study (which simply compares people with disease and people without disease) the sanitarian should immediately be able to determine common disease onset times and symptomology. Epidemiological basics include the determination of common foods, drinks, disease onset times (incubation), symptomology, and duration of illness among outbreak victims. Some caution should be exercised during information collection and prediction processes. Secondary transmission cases, which may appear during the initial phases of a foodborne-illness investigation, can confound the prediction process. Victims with aberrant symptoms, onset times, and duration times should be grouped separately from "classic" cases for later information assimilation.
Sample collection activities also should account for the possibility of multiple pathogens, especially in the case of sewage cross-contamination. A sewage point source may produce a "grab bag" of microbial and viral pathogens that will further confound the prediction and detection of a viral outbreak. When a food, water, or stool sample tests positive for a bacterial pathogen, even the most careful researcher may be induced to take an incorrect prediction pathway that may preclude analytical virus detection.
Once the information gathered from the investigation is sufficient to make a pathogen prediction, representative food samples must be collected immediately--it is important to cull these samples while they are still available. Every effort must be made to locate, quarantine, correctly package, and label suspect food items for analysis. Discarded food samples, even those retrieved from trash cans, dumpsters, and the local landfill, should be immediately analyzed for the presence of bacteria. Food and stool samples that test negative for bacteria will further support viral predictions and will bolster staff confidence about future predictive and theoretical estimation. CDC will work with the local state board of health to determine if viral analysis, which is still relatively experimental, is possible for food samples. The CDC Division of Viral and Rickettsial Diseases can be reached directly at (404) 639-4829.
If representative food and beverage samples cannot be obtained, the investigation should move swiftly into stool collection processes, focusing on the collection of stool samples from outbreak victims and ill food workers. Time is critical, because most symptomatic victims typically stop shedding the virus after seven days. Although ill food workers have been shown to shed the virus for 10 days after cessation of symptoms, stool samples will not reveal viral contamination after that period of time. Inspectors should focus on the health of food workers by interviewing each worker who prepared foods consumed within the disease incubation "window." Large volumes of stool samples should be expediently collected as appropriate. Health jurisdictions should be equipped with Cary-Blair rectal-swab and bulk stool specimen collection units for use as needed. For a complete description of stool collection procedure, see Benenson's Control of Communicable Diseases Manual. 
The inspection of implicated facilities should focus on opportunities for the cross-contamination of water and ice supplies, sewage backups, loss of water pressure, ice machines with condensate drain lines directly connected to sewer lines, and gross sanitation negligence. Company handwashing policy and procedure should be examined. The number, location, and functionality of employee handwashing sinks must be determined. Inspectors should assess employee handwashing practices through observation and employee interviews. The possibility that foods were received in a contaminated state also should be evaluated. For guidance when linked cases of suspected foodborne illnesses are received, communications should be established immediately among local and state departments of health and agriculture.
Viral outbreaks present interesting and challenging dilemmas for public health officials. The combination of a low infective dose and the microscopic size of the virus particle constitutes an unforgiving context for epidemic prevention. History has proven that even relatively "clean," compliant facilities staffed by conscientious employees can be stricken with the Norwalk virus. Symptomatic food workers, exhibiting vomiting, diarrhea, or nasal discharges, should be excluded from food preparation processes. The additional safeguard of equipping such workers with sanitary gloves may or may not provide a greater safety margin. If the gloves are not discarded frequently, they may foster a false sense of hygienic security in the absence of good handwashing practices. The mechanics of fecal-oral disease transmission must be understood by food workers--the informed food worker will become a responsible food worker. The only real prevention advice for the control of any foodborne epidemic is the timeworn sanitarian's credo: Ensure sound food protection techniques, good employee hygiene, and facility sanitation through the education of food service operators.
(1.) Mead, S.P., J.S. Bresee, V Dietz, P.M. Griffin, L.F. McCaig, C. Shapiro, L. Slutsker, R.V. Tauxe (1999), "Food-Related Illness and Death in the United States," Emerging Infectious Diseases, 5(5):1-37.
(2.) Centers for Disease Control and Prevention (1999), "Norwalk-like Viral Gastroenteritis in U.S. Army Trainees--Texas, 1998," Morbidity and Mortality Weekly Report, 48(11):2.
(3.) Smith, A.W., N. Cherry, D.O. Matson, J.H. Mead, and D.E. Skilling (1998), "Calicivirus Emergence from Ocean Reservoirs: Zoonotic and Interspecies Movements," Emerging Infectious Diseases, 4(1):1,7.
(4.) Parashar, U.D., T. Ando, J.S. Bresee, L. Dow, C. Eddy, R.L. Fankhauser, R.I. Glass, C.D. Humphrey, T. Ingram, J. Miller, S.S. Monroe, J.S. Noel, K.S. Williams (1998), "An Outbreak of Viral Gastroenteritis Associated with the Consumption of Sandwiches: Implications for the Control of Transmission by Food Handlers," Epidemiology and Infection, 121:615-621.
(5.) Centers for Disease Control and Prevention (1997), "Viral Gastroenteritis Associated with Eating Oysters--Louisiana, December 1996-January 1997," Morbidity and Mortality Weekly Report, 46(47):1109-1112.
(6.) Centers for Disease Control and Prevention (1990), "Viral Agents of Gastroenteritis: Public Health Importance and Outbreak Management," Morbidity and Mortality Weekly Report, 39(RR-5):9-10.
(7.) Benenson, A.B., ed. (1918, 1995), Control of Communicable Diseases Manual, 16th ed., Rev., Washington, D.C.: American Public Health Association.