CDC's Investigation of Gastroenteritis Aboard Spirit of Glacier Bay May 28 - June 1, 2000
CDC's Investigation of Gastroenteritis Aboard Spirit of Glacier Bay May 28 - June 1, 2000
- Recommendations and Follow-up Activities
On Thursday, June 1, 2000, a cruise ship passenger and a representative from the excursion company Alaska Sightseeing, a division of Cruise West, contacted the Section of Epidemiology, Alaska State Department of Health and Social Services, to report an outbreak of gastroenteritis among passengers and crew members aboard the cruise ship, Spirit of Glacier Bay. The Spirit of Glacier Bay sailed round-trip from Whittier, Alaska, for 4 nights, with stops in Cordova and Valdez (Table 1). Alaska Sightseeing staff reported that 18 (58%) of 31 passengers and nine (53%) of 17 crew members had reported illness to the captain and a senior customer service representative. Illnesses were characterized primarily by vomiting, diarrhea, nausea, and abdominal cramps. Because of the small size of the vessel (passenger capacity, 52 persons) a staff physician did not routinely sail with the ship.
Epidemiologists from the Alaska State Section of Epidemiology notified the Vessel Sanitation Program (VSP) at the Centers for Disease Control and Prevention (CDC) of the outbreak. Because the ship’s itinerary was limited to the domestic United States with no international port of call during the calendar year, it was subject to regulation and inspection by the Food and Drug Administration (FDA). FDA officials from the Division of Field Programs were notified by VSP of the outbreak.
Based on the report of greater than 3% of passengers and crew with gastrointestinal illness, FDA requested assistance from VSP with an epidemiologic and environmental investigation aboard the vessel. CDC issued an EPI-AID, designating the Epidemic Intelligence Service Officer (EISO) stationed at the Alaska State Section of Epidemiology as the principal investigator, under the supervision of VSP.
The objectives of the investigation were 1) to identify the etiologic agent associated with the outbreak 2) to determine the source(s) of spread of illness among passengers and crew members and 3) to formulate intervention strategies to control the outbreak and prevent a recurrence.
A case of gastroenteritis was defined as vomiting or diarrhea (three or more episodes of loose stools in a 24-hour period) in a passenger or crew member aboard Spirit of Glacier Bay with onset any time during the cruise from May 28 to June 1, 2000. Well persons were passengers or crew members who reported no symptoms of gastroenteritis during the study period. Ill passengers or crew members whose illness did not meet the case definition were classified as having indeterminate illness and were excluded from analyses.
Case-patients among the passengers and crew were interviewed in person at their hotels on June 1, the day of disembarkation, by the principal investigator and a nurse epidemiologist from the Alaska State Section of Epidemiology. The purpose of these interviews was to review exposure histories and identify individual risk factors for contracting gastroenteritis before boarding or while aboard the vessel.
Food Menus and Passenger Activities
Management officials of Alaska Sightseeing provided the principal investigator with passenger food menus from the cruise. A senior customer service representative reviewed the menus to ensure accuracy. In addition, the principal investigator interviewed the chef in person to review food handling practices and food preparation procedures. The ship did not prepare formal menus for crew meals. The principal investigator reviewed preparation of the crew meals with the chef and galley preparation worker. Crew food histories were based on individual retrospective crew member reports. Passenger activities on the ship and during shore excursions were reviewed with a senior customer service representative.
A standardized questionnaire based on passenger food menus, consumption of water and ice, shore excursions, and the likely period of exposure to the suspected causal pathogens as determined by the epidemic curve (Figure 1) was administered to all passengers for whom contact information was available. The survey was administered by telephone, e-mail, facsimile, and mail from June 7 through June 26. The questionnaire was used to characterize demographic information, symptomatology, and risk factors for illness.
The passenger questionnaire was modified to create the crew member questionnaire. Specific food items were not listed on the crew questionnaire. Instead, crew members were asked to list all food and drinks they had consumed while aboard the cruise ship. Crew members were given these questionnaires on June 9 to complete and return.
Data from questionnaires were entered and analyzed by the principal investigator using Epi Info version 6.02 software. Baseline characteristics of ill and well passengers and crew (e.g., symptoms; water, food, and ice consumption) were described by frequencies and percentiles for categorical variables (e.g., number of persons with diarrhea), and ranges of values for continuous variables (e.g., age). Exposures were coded as dichotomous variables. In univariate analyses, relative risks with 95% confidence intervals (CIs) were used to test for differences in illness risk between ill and well persons. Significance was assessed at a p value of less than 0.05.
On June 2, the principal investigator and the nurse epidemiologist collected vomitus and stool specimens from ill passengers and crew members. Specimens were sent to the Alaska State Public Health Laboratory and were forwarded to the Division of Viral and Rickettsial Diseases Laboratory, CDC, to test for Norwalk-like virus (NLV) using polymerase chain reaction (PCR). Tests to isolate bacterial pathogens from specimens were not performed.
On June 4, two environmental health officers from the Alaska State Department of Environmental Conservation (ADEC) and two environmental health officers from VSP conducted a full inspection of the ship. The inspection included a review of housekeeping procedures (e.g., infection-control measures, and cabin and public space cleaning practices); a review of food storage, preparation and handling practices in the galley; and an examination of the potable water system, including water sampling.
Questionnaires were collected from 30 (97%) of 31 passengers. One passenger reported by a cruise customer service representative as ill did not respond to a mailed questionnaire or follow-up telephone calls. Of the 30 passenger respondents, 20 (67%) reported illness that met the case definition. Four had indeterminate illness and were excluded from further analyses.
Vomiting was reported by 18 (90%) of 20 case-passengers, diarrhea by 16 (80%), nausea by 12 (60%), chills by 11 (55%), abdominal cramps by 10 (50%), fever by eight (40%), muscle aches by eight (40%), and headache by three (15%) (Table 2). No case-passengers reported experiencing bloody diarrhea or a sore throat. Dates of illness onset ranged from May 30 through June 1. The median time of onset of symptoms was between 8 p.m. and midnight on May 31 (Figure 1). Three case-passengers sought medical attention after disembarking the ship; none were hospitalized. The median duration of illness reported was 36 hours (range 18-96 hours).
The ship usually operated with 17 crew members but one member disembarked from the ship with an injury on May 30 before the cruise terminated. Six crew members were interviewed in Anchorage on June 1 by the principal investigator and nurse epidemiologist. Six crew members completed surveys by telephone interview on June 2. Additionally, illness information was obtained from the remaining four crew members by the ship’s senior customer service representative and the captain who had maintained a log of ill crew members and passengers.
Of the 16 crew members aboard the vessel for the duration of the cruise, nine (56%) had illness that met the case definition. Diarrhea was reported by seven (78%) of nine case-crew members, nausea by seven (78%), and vomiting by seven (78%) (Table 2). Abdominal cramps were reported by seven (86%) of eight case-crew members, chills by three (60%) of five, fever by three (50%) of six, and headache by four (50%) of eight. None of the four case-crew members with complete symptom information reported bloody diarrhea, sore throat, or muscle aches. Onset dates of illness ranged from May 29 through June 1. The median onset time was between noon and 6 p.m. on May 31. Two case-crew members sought medical attention after disembarking from the ship; none were hospitalized. The median duration of illness reported was 36 hours (two case-crew members).
Because many crew members were unable to reliably recall and select, on a questionnaire, food that they had eaten during the cruise, crew were excluded from the cohort study.
Among passengers, illness was not associated with consumption of tap water or ice or with shore excursions (Table 3). Chicken, served at the barbecue lunch on May 29 was the only food associated with an increased risk for illness. Passengers who ate chicken were 1.86 times as likely to become ill as persons who did not eat chicken (p < 0.015).
No coliform bacteria were detected in any of the 11 water samples collected from the ship’s water supply. Chlorine levels were adequate in all water samples determined by a manual chlorine test kit.
Stool and vomit specimens
Stool specimens were collected from six persons (two passengers and four crew members.) One vomitus specimen was collected from a passenger. Five of the six stool specimens tested positive for NLV by PCR using a “region B” primer set. The sixth stool specimen, from a crew member, and the vomitus specimen tested negative for NLV.
Potable water was bunkered on board the vessel in Whittier. Chlorination levels of the potable water supply were obtained daily by crew members. VSP measured the free residual chlorine concentration in the potable water holding tanks and distribution system and performed a visual inspection and record review to identify possible direct or indirect cross-connections between potable and nonpotable water stores.
The water bunkered on board the vessel measured 2.2 ppm or greater free available chlorine. Water samples taken from the galley hand wash sink and the food preparation sink measured 1.2 ppm and 2.2 ppm, respectively. Chlorination levels of water samples from two passenger cabins measured 0.08 ppm and 0.07 ppm, and in two crew cabins, 0.03 ppm and 0.05 ppm (normal range 0.2-4.0 ppm for holding tanks; any measurable amount for tap water).
The following deficiencies were noted during the environmental health inspection of the vessel on June 4 in accordance with VSP Operations Manual, 1989.
Potable Water System
1. Potable water hoses were stored without the ends capped or coupled together to protect them from contamination. Additionally, potable and nonpotable water fittings were stored together.
2. Shower hoses (long enough to reach the deck) in passenger and crew cabins were not equipped with backflow preventers. There have been numerous incidents in the past when the sewage system has backed up flooding bathroom decks with sewage.
3. Clothes washing machines were not equipped with backflow preventers.
4. The potable water chlorine analyzer (far point) received water 4 feet from where the chlorine was injected.
5. The potable water system had not been periodically super chlorinated.
Food Service Areas
6. The food preparation/dish wash sink was not equipped with an air break on the drain line.
7. Several food and equipment storage cabinets in the galley were heavily soiled with food residue.
8. Numerous bain marie pans, knives, and other food-contact surfaces of equipment and utensils were heavily soiled with food residue. Knives and utensils in a nylon knife carrying case on a storage shelf appeared to have been replaced without being washed, rinsed, and sanitized.
9. The grill grease drip pan had accumulated food debris, grease, and water.
10. Oven interiors were heavily soiled with food residue.
11. Nonfood items, such as vitamins, were stored with food items.
12. No sneeze guard was available for food items provided as self-service.
13. Ice machine components were soiled with mold.
14. Nonfood items were stored on top of the freezers by the laundry, and this area was soiled with dust and dirt.
Alaska Sightseeing cooperated fully with the investigation. The corporate office cancelled the next three scheduled cruises to accommodate the environmental health inspectors as well as to complete a thorough cleaning and sanitizing of the ship.
Chicken served at a lunchtime barbecue on May 29 was associated with an 86% increased risk for illness. Given the small numbers of well persons in the cohort study, this association could have been spurious. The chicken had been thoroughly cooked, and therefore any contamination with virus would have had to occurred when the chicken was served.
On a cruise ship where there was a large amount of shared space (e.g., in public restrooms, the bar, hand rails), ill crew members or passengers may have contaminated one of these surfaces that were later touched by other passengers or crew members. Additionally, hors d’oeuvres at cocktail hours and certain breakfast items were self-service and may have provided opportunities for persons to spread virus to through contaminated nonfood items such as utensils. One factor that might have obscured the ability to unequivocally identify a single point source of contamination was the high attack rate resulting in relatively low numbers of well passengers and crew, which limited statistical power.
NLVs cause an acute, self-limited gastroenteritis characterized by diarrhea, nausea, vomiting, abdominal cramps, muscle aches, and fever. The incubation period and the average duration of illness are from 24 to 48 hours. The most common routes of NLV transmission are foodborne or person-to-person contact.
In conclusion, almost 70% of the passengers and crew members aboard the ship suffered acute gastroenteritis caused by NLV. The source of the virus and method of its transmission among persons on board the ship were not found. Deficiencies aboard the ship noted by the inspection team, although not directly responsible for transmission of the virus, could have facilitated its survival.
Potable Water System
1. Store potable and nonpotable water hoses and fittings separately. Store hoses with ends either capped or coupled together.
2. Install backflow preventers on shower hoses at the fixed point to protect against contamination of the potable water system.
3. Install backflow preventers on water supply lines to washing machines.
4. Ensure chlorine analyzer receives water from a point on an upper/forward part of the vessel to more accurately measure chlorine residual throughout the vessel.
5. Super chlorinate entire potable water system at least once a year by raising the free chlorine level throughout the vessel to 50 ppm chlorine for 4 hours or 100 ppm for 1 hour. Then flush system until the free chlorine level is below 5.0 ppm.
Food Service Areas
6. Install a physical separation (air break) in the sink drain line.
7. Clean cabinets as often as necessary to keep them free of food and dirt accumulation.
8. Wash, rinse, and sanitize all food service equipment before storing. The nylon knife carrying case should be used only during transportation of knives. Install a lockable stainless steel knife locker if additional storage is required.
9. Clean the grill drip pan at least daily.
10. Clean oven interiors as often as necessary to keep them free of food accumulation.
11. Do not store nonfood items with food items.
12. Provide sneeze guards on self-service areas.
13. Clean the ice machine as often as necessary to keep it free from accumulations of mold.
14. Keep area on top of freezers clean, and do not use them for storage.
15. During outbreak conditions, use a 10% chlorine solution to sanitize hard surfaces of all passenger and crew cabins, not only those used by ill persons. With the same solution, clean all handhold surfaces through out the vessel, such as handrails and doorknobs.
16. Use separate toilet brushes for each toilet to prevent cross-contamination of cabins.
17. Provide daily reports of gastrointestinal illness to the appropriate public health agency until deemed no longer required. Adopt the “Gastrointestinal Illness” log set forth by VSP. An example of this log is available on the VSP website: http://www.cdc.gov/nceh/vsp.
18.Restrict ill food handlers from handling food or dishes until symptom-free for at least 48 and preferably 72 hours.
19. Post signs in bathrooms used by food handlers indicating they must wash their hands before returning to work.
20. Recommend having cruise personnel, chefs, hotel managers, and chief engineers attend VSP training seminars held in Miami, FL. Seminars are given four times per year. Contact Ms. Catherine Arthur, Cruise Institute International, 1809 Silver Valley Ct, Apopka, Florida 32712. Telephone 407.884.1301, Fax 407.884.1302, e-mail firstname.lastname@example.org.
The next excursion (June 11-15) sailed by Spirit of Glacier Bay was unremarkable. No illnesses were reported among passengers or crew. However, gastrointestinal illnesses were reported among both passengers and crew members of Spirit of Columbia on June 16. Both ships were owned by Cruise West. Spirit of Columbia (passenger capacity, 78 persons) sailed in Prince William Sound during June 12-16 with the same itinerary as Spirit of Glacier Bay.
Two members of the original crew from Spirit of Glacier Bay worked on Spirit of Columbia during June 12-16. Both had been ill during the outbreak, but were asymptomatic by June 2, and did not experience any further gastrointestinal symptoms. It is therefore unlikely that they could have been the source for the outbreak of gastrointestinal illness on Spirit of Columbia.
We received an anecdotal report that a passenger, traveling in another area of Alaska before the cruise, boarded the Spirit of Columbia and became ill shortly thereafter. This person could have been the source of illness for the other passengers and crew who experienced onset of illness approximately 72 hours after boarding the ship. Not all ill passengers were contacted after the cruise ended; however, cruise personnel actively sought reports of illness among the passengers while still on board and reported that 10 of 65 passengers and seven of 21 crew were ill with acute gastroenteritis.
Twelve passengers were available to be interviewed by the principal investigator on the evening of June 16. Six of them had been ill, and one submitted a stool sample. All seven crew members who had been ill were interviewed later in the evening and four submitted stool samples. Information about an additional four passengers was obtained on June 24; only one of these four reported illness.
Diarrhea was reported by six (86%) of seven case-passengers, abdominal cramps by five (71%), fever by five (71%), vomiting by four (57%), nausea by four (57%), and muscle aches by four (57%). Other symptoms included headache reported by three (43%) of seven case-passengers and a sore throat by one (14%). No ill passengers reported chills or bloody diarrhea. The median onset time of symptoms was between midnight through 6 a.m. on June 16. One case-passenger sought health care after disembarking from the ship but was not hospitalized.
Seven (100%) of seven case-crew members reported abdominal cramps, headache and muscle aches. Vomiting was reported by six (86%) of seven case crew-members, diarrhea by six (86%), nausea by six (86%), and a sore throat by three (43%). No ill crew members reported chills, fever, or bloody diarrhea. The median onset time of symptoms was the same as for the passengers, between midnight through 6 a.m. on June 16. None of the case-crew members sought health care after disembarking from the ship.
All stool specimens tested negative for Shigella, Salmonella and Campylobacter at the Alaska State Public Health Laboratory. Three of the five specimens tested for NLV at CDC were positive--one from a passenger and two from crew members. A crew member who had been ill on the Spirit of Glacier Bay and then sailed on Spirit of Columbia tested negative for NLV.
Unfortunately further transmission of NLV is expected as passengers and/or crew circulate around the various Alaska tourist attractions. The best disease prevention and control measures are to maintain good sanitary practices and recommend that ill persons adequately rest in an area where they do not have the opportunity to infect other persons. ADEC personnel were notified of the illnesses associated with Spirit of Columbia and will continue to inspect and monitor restaurants and hotel facilities in the areas of Alaska where most land-based tours travel. Health-care providers in that area were also alerted to report clusters of gastroenteritis among tourists or employees to the Section of Epidemiology.
Louisa Castrodale, D.V.M., M.P.H.
Epidemic Intelligence Service Officer
Epidemiology Program Office
Assigned to Section of Epidemiology,
Alaska Department of Health and Social Services
Elaine H. Cramer, MD, MPH
Vessel Sanitation Program
National Center for Environmental Health
Centers for Disease Control and Prevention
Click here for larger view
*2-tailed Fisher’s exact test.