On Friday, June 6, 1997, the Bureau of Epidemiology received a telephone call from a physician in a Salt Lake City clinic, reporting an outbreak of gastrointestinal illness among a group of tourists from the United Kingdom.  Upon interviewing the four patients visiting the clinic, it was found that the common symptoms of the patients were nausea, vomiting, abdominal cramps, and explosive diarrhea.  A subsequent interview with the tour guide revealed that more than half of this group of tourists had been ill.  The index case had onset of symptoms on May 29, 1997, i.e., one day after these tourists arrived in the United States.  Since this group of tourists was scheduled to leave Salt Lake City that afternoon for Pocatello, Idaho, and then to Jackson, Wyoming, Dr. Bao-Ping Zhu, EIS Officer at the Department of Health, followed this tourist group to conduct further investigations of this outbreak.

In this investigation, a case was defined as a person who: (1) was riding in the tour bus between May 29 and June 6; and (2) had vomited at least once during the period, or had diarrhea (i.e., loose stools or bowel movements at least twice during any 24-hour period).  Cases were identified through interviews with all persons riding the tour bus, including the 45 tourists, the tour guide, and the bus driver, using the case definition described above.

Information provided by the tour guide (which was later verified during the interviews of the tourists) showed that, among the 45 tourists, 39 had taken the same airplane from the United Kingdom to Newark, New Jersey, on May 27.  From Newark, these 39 tourists boarded another airplane to Denver, Colorado.  They arrived in Denver late in the afternoon on May 28, 1997.  The remaining six tourists made their own travel arrangements and arrived in Denver on May 28 on separate airlines.

Since three of the six tourists who had made their own travel arrangements and arrived in Denver, Colorado on separate airlines later also developed the illness, it appeared that the pathogenic agent causing this outbreak was not contracted on the airplanes.  Therefore, efforts were focused on potential common food exposures after these tourists had arrived in Denver.  Because the index case had onset of symptoms on the evening of May 29, and the only common meal shared before the onset of the index case was the breakfast buffet served by Hotel A in Denver between 7:00 a.m. and 8:00 a.m. on May 29, the investigation was concentrated on food items served at that breakfast buffet.  In addition, information was collected about the food items served on the airplanes and at a similar breakfast buffet at the same hotel on the morning of May 30.  A questionnaire was constructed based on the information provided by the tour guide and by the four patients visiting the Salt Lake City clinic.  The menus of the breakfast buffet on May 29 and May 30 and of the food items served on the airplanes between the United Kingdom and Denver were obtained from Hotel A and Airline X, respectively, to help the tourists recall their food items.  All persons involved agreed to participate in the investigation, yielding a response rate of 100%.

Among the 47 persons riding the tour bus, 45 were tourists, one was the tour guide, and one was the bus driver.  Twenty-five (53.2%) were men, and 22 (46.8%) were women.  The mean age of these persons was 65.6 years (range=51-83 years).  The mean age for the cases was 66.3 years, and that for the non-cases was 64.5 years.  All persons were white.

Of the 47 persons involved, 28 had either vomited or had at least two episodes of diarrhea during a 24-hour period, thus meeting our case definition. The attack rate was 63.6%.  Among the 28 cases, nausea, vomiting and diarrhea were the most common symptoms, each being reported by more than 70% of the cases (Table 1).  No person had bloody diarrhea.  Low grade fever was reported by 35.7% of the cases.  Seven (25.0%) had visited a health care provider, and two persons (7.1%) were hospitalized.  Several cases developed a second episode of symptoms 1-3 days after the first attack.

The index case had onset on the  evening of May 29.  On the following day there were no cases.  Four cases had onset on May 31, followed by another day without cases.  During the next three days starting on June 2, the number of cases dramatically increased, when there were eight, ten, and four cases, respectively.  No new cases emerged after June 4, and by June 8, all patients had recovered.  Assuming the index case had the same illness as the other cases, the breakfast buffet served by Hotel A on May 29 would be the most logical common exposure.  Therefore, the risk of becoming ill in relation to the consumption of food and beverage items served at that breakfast buffet (Table 2) was assessed.  Three persons were excluded from the analysis because they developed some symptoms but did not meet the case definition.  Of all food and beverage items examined, only the breakfast potato was significantly associated with the risk of developing illness (p=0.04).  However, the association was rather weak, with the relative risk (RR) being only 1.6.

An interesting case was the bus driver.  According to the driver, he only started to have meals together with the tourists on May 31.  Thus he had no opportunity of being exposed to the food or beverage items at the breakfast buffets on either May 29 or May 30.  However, on June 3 he also became ill.  Another relevant piece of information was the seating patterns of the tourists on the bus.  Instead of staying in the same seat every day, the tourists systematically rotated their seats twice a day.  That is, if a person sat in the front row in the morning, he or she would be rotated to the last row in the afternoon, while the persons sitting in the second row in the morning would be rotated to the front row in the afternoon; and so on.  Also worth mentioning was the fact that there was a bathroom on the bus.  According to the tourists, during the first few days of the tour several persons had used the bathroom when they were ill, although they had been discouraged from using the bathroom later on when more and more persons became ill.

The investigator's impression of the tour bus's interior was that both the floor and the seats were remarkably clean.  The bathroom was not inspected; however, according to both the tourists and the bus driver, the bathroom had been cleaned frequently throughout the tour.

In addition to interviewing the tourists, the food services management of Hotel A was contacted on June 9.  According to the hotel's management, no food handlers preparing the breakfast buffet on May 29 had been ill prior to this outbreak.  Also, several employees of the hotel also ate at the breakfast buffet on May 29, but none became ill afterwards.

Three cases submitted stool samples to the State Public Health Laboratory, where the samples were tested for Salmonella, Shigella, E. coli 0157:H7, and Campylobacter spp.  None of these pathogenic bacteria were found.

What pathogen has caused this outbreak?  Among the four types of pathogens that can cause gastrointestinal illnesses (i.e., bacteria, viruses, toxins and parasites), it appears that toxins and parasites can be readily excluded as the cause of this outbreak, because intoxication (both chemical and biological) usually has an acute and violent manifestation and a short (<8 hours) incubation period, while parasite infections (such as giardiasis) is usually characterized by chronic diarrhea lasting several weeks, and patients rarely have vomiting.  From the distribution of cases' symptoms, it appears that this outbreak was also unlikely to have been due to bacterial pathogens.  In most bacterial gastrointestinal infections, fever and abdominal cramps are the most common symptoms.  In addition, in many bacterial gastroenteric infections (such as those caused by E. coli, Shigella, and Campylobacter), patients frequently have bloody stool.  In this outbreak, however, only 35.7% of the cases had self-perceived low grade fever; just 50% of the cases had abdominal cramps; and none of the patients had bloody stool.

On the other hand, the following characteristics regarding the distribution of symptoms among cases appears to be in accordance with the hypothesis that the outbreak was an epidemic viral gastroenteropathy caused by Norwalk virus or a similar agent: that nausea, vomiting, and watery diarrhea each affected more than 70% of the cases; that fever, abdominal cramps, and myalgia were relatively infrequent; and that no cases had bloody diarrhea.  Unfortunately, the stool samples submitted by the three cases were not properly preserved for virus testing, nor were there acute and convalescent sera collected from any of the cases, making it impossible to verify this hypothesis.

Regarding the mode of transmission, although the breakfast potato  was significantly associated with the risk of developing the gastrointestinal illness, this outbreak is unlikely to have been due to a common food exposure for the following three reasons.  First, only a weak (albeit statistically significant) association (RR=1.6) was observed between exposure to the implicated food item (i.e., breakfast potato) and risk of illness, whereas generally food-borne outbreaks are characterized by large relative risks.  Second, 47.6% of the persons who did not eat breakfast potatoes also developed the illness, including the bus driver, who had no chance to have been exposed to the breakfast items on either day.  Third, since breakfast potatoes are usually fully cooked, it was unlikely for this food item to be the carrier of pathogenic agents, unless it was contaminated by a food handler who was shedding the pathogens immediately before it was served.  According to the manager of food services at Hotel A, no employees preparing the breakfast buffet had been sick; also, several employees had eaten at the breakfast buffet on May 29, but none had been sick afterwards.

On the other hand, neither person-to-person nor airborne exposure can be ruled out as the modes of transmission in this outbreak.  All the tourists stayed in close proximity and they systematically rotated their seats twice a day during the entire tour.  Both of these factors had the potential to facilitate person-to-person transmission.  On the other hand, airborne transmission could also have happened, given that the tour bus was a confined environment, the tourists stayed in the bus for a prolonged period of time, and several of the cases had used the bathroom on the bus when they were vomiting or having diarrhea.  Previous investigations of Norwalk disease outbreaks have suggested that airborne or contact transmission of the Norwalk agent was possible in a confined environment, such as in a hospital.¹

The incubation period of Norwalk agent disease is usually 24 to 48 hours.  It is possible that the index case whose onset date was May 29 may have transmitted her illness to the four cases whose onset date was May 31.  These four cases in turn may have transmitted their illness to all the other cases.

In summary, this outbreak appears to have been an epidemic viral gastroenteropathy caused by Norwalk agent.  Epidemiologic evidence suggests that person-to-person or airborne were the most likely modes of transmission for this outbreak.

We recommend that in the future, preventive measures be adopted promptly after the emergence of the first case.  These measures could include educating passengers about proper hand-washing after using the bathroom and advising symptomatic passengers to limit close contacts with others.  In addition, in the setting of a possible outbreak, local health departments should be notified immediately.