Traveler s Diarrhea at Sea: Three Outbreaks of Waterborne Enterotoxigenic Escherichia coli on Cruise Ships

1491 CONCISE COMMUNICATION Traveler s Diarrhea at Sea: Three Outbreaks of Waterborne Enterotoxigenic Escherichia coli on Cruise Ships Nicholas A. Daniels, 1,4,6 Jakob Neimann, 7 Adam Karpati, 5 Umesh D. Parashar, 3 Katherine D. Greene, 1 Joy G. Wells, 1 Anjali Srivastava, 2 Robert V. Tauxe, 1 Eric D. Mintz, 1 and Robert Quick 1 1 Foodborne and Diarrheal Diseases Branch and 2 Biostatistics and Information Management Branch, Division of Bacterial and Mycotic Diseases, and 3 Viral Gastroenteritis Section, Respiratory and Enterovirus Branch, National Center for Infectious Diseases, 4 Epidemic Intelligence Service, Epidemiology Program Office, and 5 Health Studies Branch, Division of Environmental Hazards and Health Effects, National Center for Environmental Health, Centers for Disease Control and Prevention, Atlanta, Georgia; 6 Division of General Internal Medicine, Department of Medicine, University of California, San Francisco; 7 Danish Zoonosis Center, Danish Veterinary Laboratory, Copenhagen, Denmark Enterotoxigenic Escherichia coli (ETEC) has become the leading bacterial cause of gastroenteritis outbreaks on cruise ships. Investigation of recent outbreaks of ETEC gastroenteritis on 3 cruise ships indicated that all were associated with consuming beverages with ice cubes on board the ship (relative risk [RR], 1.4, 95% confidence interval [CI], 1.0 1.9, P=.02; RR, 1.9, 95% CI, 1.3 2.9, P!.001; and RR, 1.3, 95% CI, 1.0 1.6, P!.01), and 2 were associated with drinking unbottled water (RR, 2.7, 95% CI, 1.8 4.1, P!.001; RR, 1.7, 95% CI, 1.3 2.3, P!.001). Multiple ETEC serotypes were detected in patients stool specimens in each of the 3 outbreaks, and 12 (38%) of 32 isolates were resistant to 3 antimicrobial agents. ETEC appears to be emerging as a waterborne pathogen on cruise ships. Water bunkered in overseas ports was the likely source of ETEC infection in these outbreaks. To ensure passenger safety, cruise ships that take on water in foreign ports must ensure that water treatment and monitoring systems function properly. Outbreaks of gastroenteritis occur on commercial cruise ships at a rate of 1 6 outbreaks per 1000 cruises [1, 2]. In recent years, passenger-carrying vessels have increased in number and size to accommodate a growing number of holiday and business travelers, exposing more people to the risk of developing gastroenteritis. Enterotoxigenic Escherichia coli (ETEC), a common cause of traveler s diarrhea [3, 4], is the leading bacterial cause of outbreaks of gastroenteritis on cruise ships [1]. Although contaminated water and food are both implicated as vehicles of transmission of ETEC infection [1, 5 8], before 1996 there were no confirmed reports of waterborne outbreaks of ETEC gastroenteritis on cruise ships. From 1986 to 1993, the number of waterborne outbreaks of Received 5 October 1999; revised 28 December 1999; electronically published 13 April 2000. Presented in part: 36th annual meeting, Infectious Diseases Society of America, Denver, November 1998 (abstract 533). This work is in compliance with the human experimentation guidelines of the US Department of Health and Human Services. Reprints or correspondence: Dr. Eric Mintz, Foodborne and Diarrheal Diseases Branch, Division of Bacterial and Mycotic Diseases, Centers for Disease Control and Prevention, 1600 Clifton Rd., Mailstop A38, Atlanta, GA 30333 (edm1@cdc.gov). The Journal of Infectious Diseases 2000;181:1491 5 2000 by the Infectious Diseases Society of America. All rights reserved. 0022-1899/2000/18104-0039$02.00 gastroenteritis from any cause on cruise ships decreased from the previous 10-year period [1]. This decline was attributed in part to the efforts of the Vessel Sanitation Program (VSP) of the Centers for Disease Control and Prevention (CDC) [1]. The VSP was established in 1975 to help cruise ships achieve and maintain a high level of sanitation and thereby lower the risk of outbreaks of gastrointestinal disease. VSP requires that 24 h before arrival at a US port, the captains of ships with international itineraries report the number of passengers and crew members who have visited the ship s physician for diarrhea. If 3% of the passengers or crew members report gastroenteritis or if severe or unusual illness occurs, the CDC may choose to investigate further. Since 1990, the CDC has investigated 10 cruise ship outbreaks of ETEC gastroenteritis. Contaminated water was implicated as the source of infection in 4 of these outbreaks, 3 of which occurred after 1997. Here we report the findings of the investigations of the 3 most recent outbreaks. Methods Outbreak 1. On 10 April 1997, the VSP was notified of an outbreak of gastroenteritis on a cruise ship that departed from Acapulco, Mexico, on 5 April 1997 and was scheduled to arrive in New York City on 19 April. On 12 April, the ship s physician reported that 83 (7.1%) of 1169 passengers and 14 (2.6%) of 532

1492 Daniels et al. JID 2000;181 (April) crew members were ill with diarrhea and abdominal cramps. On 14 April, CDC investigators met the ship in St. Thomas, Virgin Islands, and began an investigation. Outbreak 2. On 22 December 1997, the Florida State Health Department contacted the VSP after receiving phone calls from passengers who reported diarrheal illness that began during and after a cruise. The ship had embarked from Tampa Bay, Florida, on 17 December. It arrived in Playa Del Carmen and Cozumel, Mexico, on 19 December and returned to Florida on 21 December. The VSP received a report from the ship s physician of diarrheal illness in 18 (2.5%) of 717 passengers and 1 (0.25%) of 400 crew members during the cruise. On 8 January 1998, ETEC was isolated from 4 patients, and a retrospective investigation was initiated. Outbreak 3. On 22 May 1998, the VSP received a report of diarrheal illness in 189 (31%) of 611 passengers and 8 (2%) of 410 crew members on a cruise ship that embarked from Montego Bay, Jamaica, on 17 May. The CDC initiated an investigation when the ship returned to Montego Bay on 24 May. Epidemiologic investigations. In all 3 investigations, the ships medical logs were reviewed and hypothesis-generating interviews were conducted with ill passengers. On the basis of this information, questionnaires about consumption of food, water, and other beverages on ship and on shore during port calls were distributed to all passengers on board ship during outbreaks 1 and 3 and were mailed, 1 2 weeks after disembarkation, to 712 (99%) passengers on board ship during outbreak 2. In all 3 outbreaks, a case of gastroenteritis was defined as a person with 3 loose stools in a 24-h period. Parents responded for children who were passengers on the ship. A questionnaire was distributed to all crew members on board ship during outbreaks 2 and 3 but not to those on board ship during outbreak 1. This questionnaire asked about illness and water and ice consumed on the ships but not about food, since the crew did not have a standard food menu. Responses from passengers and crew members were analyzed separately. Environmental investigations. CDC investigators inspected all 3 implicated cruise ships. Records of food purchases, water bunkering (filling of the ship s water storage tanks), water chlorination, and medical and diarrheal logs were reviewed. Free-chlorine residuals were tested at different points in the potable water distribution system. Laboratory investigations. Stool specimens were obtained from 30 passengers who became ill during outbreak 1, 6 who became ill during outbreak 2, and 10 who became ill during outbreak 3. The specimens were sent to the CDC, and those from outbreaks 2 and 3 were cultured for Salmonella and Shigella organisms, Escherichia coli O157:H7, Campylobacter jejuni/coli, and Vibrio species. E. coli recovered from stool specimens from outbreaks 1, 2, and 3 were tested for heat-labile (LT) and heat-stable (ST) enterotoxin gene sequences by polymerase chain reaction (PCR) [9]. All ETEC isolates identified were serotyped and tested for antimicrobial susceptibility by the disk diffusion method. The antimicrobial resistance patterns of all ETEC isolates from outbreaks on cruise ships during 1975 1998 were reviewed. Results Epidemiologic investigations. Questionnaires were returned by 95% of passengers in outbreak 1, 77% of passengers in outbreak 2, and 87% of passengers in outbreak 3. Diarrhea was reported by 429 (39%) respondents in outbreak 1, 457 (84%) in outbreak 2, and 463 (87%) in outbreak 3. The median duration of illness reported by the passengers who had recovered by the time questionnaires were administered was 3 days (range, 1 7) for outbreak 1, 8 days (range, 1 30) for outbreak 2, and 2 days (range, 1 4) for outbreak 3. Questionnaires were returned by 67% of crew members in outbreak 2 (28 [11%] reported diarrhea) and by 96% of crew members in outbreak 3 (41 [10%] reported diarrhea). Illness onset for crew members peaked on the same day as that for passengers during both outbreaks. The most common symptoms reported by ill passengers in the 3 outbreaks were diarrhea (99%), abdominal cramps (78%), nausea (63%), headache (47%), myalgia (39%), fever (27%), and vomiting (22%). The dates of illness onset showed that the outbreaks began soon after passengers embarked and were sustained for several days, suggesting an ongoing and widespread exposure (figure 1). Illness was associated with consuming the ship s tap water (relative risk [RR], 2.7; 95% confidence interval [CI], 1.8 4.1; P!.001) and beverages with ice (RR, 1.4; 95% CI, 1.0 1.9; P=.02) in outbreak 1; beverages with ice (RR, 1.9; 95% CI, 1.3 2.9; P!.001) and ice water obtained from pitchers in the dining room (RR, 1.2; 95% CI, 1.0 1.4; P=.02) in outbreak 2; and unbottled water (RR, 1.7; 95% CI, 1.3 2.3; P!.001) and any beverage with ice (RR, 1.3; 95% CI, 1.0 1.6; P=.002) in outbreak 3. There was a positive association between illness and both the number of glasses of tap water consumed ( P!.001) and the number of beverages with ice consumed ( P=.007) in outbreak 1 and between illness and the number of beverages with ice consumed ( P=.01) in outbreak 2. Drinking bottled water did not protect against the development of illness in any of the outbreaks. Consumption of food or beverages while on shore was not associated with illness in any of the 3 outbreaks. Among crew members on board ship during outbreak 2, diarrheal illness was associated with consumption of any beverage containing ice (RR, undefined; Fisher s exact 95% CI, 4.7 4; P!.001) and with having an ice bucket in the cabin (RR, 3.3; P!.01). In outbreak 3, however, consumption of water and ice among crew members was not associated with gastroenteritis. Environmental investigations. On the ship on which outbreak 1 occurred, inspection of the galleys found no deficiencies in foodhandling and storage practices. Chlorination logs showed that water bunkered in Acapulco on 5 April before departure (day 1 of the cruise) was chlorinated to a free-chlorine residual of 2.0 3.0 ppm. Chlorine-recording charts and logs for bunkering and production of water were reviewed, and recorded free-chlorine residuals were acceptable. All water bunkered in Acapulco was used early in the cruise. Environmental inspection of the ship on which outbreak 2

JID 2000;181 (April) Enterotoxigenic E. coli Outbreaks 1493 On 16 December (1 day before cruise embarkation), the freechlorine residual at a distal site of water distribution was recorded as 0 ppm, but this was corrected in!4 h. After returning to the United States, the cruise ship bunkered fresh water in Florida before embarkation on the cruise during which the outbreak occurred. The fresh water tank inventory showed that only 1 of 14 total potable water tanks filled in Costa Rica had been used during the preceding cruise on 14 17 December. In outbreak 3, the environmental assessment of the cruise ship found several food storage, preparation, and handling deficiencies. Fresh water tank records confirmed that 475 tons of water were obtained from a road tanker at Puerto Limon, Costa Rica, on 15 May. After bunkering water in Costa Rica, freechlorine levels recorded on the ship were below the VSP recommended minimum of 0.2 ppm on several occasions between 15 and 17 May. On 17 May before embarkation (day 1 of the cruise), fresh water was bunkered in Jamaica. Laboratory investigations. In outbreak 1, 13 ETEC isolates were recovered from stool specimens from 11 (37%) of 30 ill passengers (table 1). Eight isolates were serotype O169:H41 positive for the ST gene, and 3 were serotype O148:H28 positive for the LT and ST genes. Two (15%) of the 13 ETEC isolates were resistant to 3 antimicrobial agents. Neither fecal (thermotolerant) coliform bacteria nor E. coli was detected in any of the 10 water samples collected aboard the ship on 14 April. In outbreak 2, 7 ETEC isolates were recovered from 4 (67%) of 6 stool specimens from ill passengers. Two patients had 2 Figure 1. Cases of gastroenteritis in passengers on 3 cruise ships with enterotoxigenic Escherichia coli outbreaks, by day during or after cruise on which onset occurred. occurred revealed no deficiencies in food storage, preparation, or handling. The ice machines aboard the ship were connected to the potable water system, and ice was made continuously throughout the day. Free-chlorine residuals in stored water were adequate, and all ice machines were working properly. The ship s captain reported bunkering water at Puerto Limon, Costa Rica, on 13 December 1997, before the cruise associated with the outbreak, and fresh water tank records confirmed that 595 tons of water were obtained from a road tanker. After bunkering water in Costa Rica, free-chlorine residuals recorded on 13 December by the ship s chlorine analyzer and chart recorder were below the VSP recommended minimum of 0.2 ppm. Table 1. Enterotoxigenic Escherichia coli stool culture results for passengers with diarrhea during 3 cruise ship outbreaks, 1997 1998. No. of stool samples yielding ETEC a Serotype Toxin type Antimicrobial resistance Outbreak 1 (n = 30) 8 O169:H41 ST TET 3 O148:H28 LT/ST TET 1 O27:H7 ST TET, SUL, S 1 O78:H12 ST C, SXT, TET, SUL, S 19 Negative Negative Not applicable Outbreak 2 (n =6) 2 O169:H41 ST TET 2 O8:H9 LT C, TET, AM, AmC 1 O167:H5 LT/ST SXT, SUL, S 1 O153:H45 ST TET, AM, SUL, S 1 O27:NM ST Sensitive 2 Negative Negative Not applicable Outbreak 3 (n = 10) 2 O169:H41 ST TET 2 O25:NM LT Sensitive 2 O6:H16 LT/ST SXT, TET, AM, SUL, S 1 O169:H41 ST SXT, TET, AM, SUL, S 1 O25:NM ST Sensitive 1 O64:NM LT SXT, TET, AM, SUL, S 1 O153:H45 ST SXT, TET, AM, SUL, S 1 O169:NM ST TET 1 O148:H28 LT/ST SXT, TET, AM, SUL, S 2 Negative Negative Not applicable NOTE. AM, ampicillin; AmC, amoxicillin/clavulanic acid; C, chloramphenicol; LT, heat-labile toxin; NM, nonmotile; S, streptomycin; ST, heat-stable toxin; SUL, sulfisoxasole; SXT, trimethoprim/sulfamethoxazole; TET, tetracycline. a Some stool samples from patients had 11 serotype isolated.

1494 Daniels et al. JID 2000;181 (April) ETEC-serotype isolates (ST-positive O169:H41 and LT-positive O8:H9). Stool cultures were negative for E. coli O157:H7 and for Salmonella, Shigella, Vibrio, and Campylobacter organisms. Of 7 ETEC isolates, 4 (57%) were resistant to 3 antimicrobial agents. In outbreak 3, 12 ETEC isolates were recovered from 8 of 10 stool specimens sent to the CDC. Seven ETEC serotypes were isolated, including ST-producing O169:H41 from 3 patients and LT- and ST-producing O148:H28 from 1 patient. Shigella sonnei also was isolated from a single specimen. All 10 stool cultures were negative for Salmonella, Vibrio, and Campylobacter organisms and for E. coli O157:H7. Six of 12 ETEC isolates were resistant to 3 antimicrobial agents. Nine (43%) of 21 water and ice samples collected from the ship on 24 May, after hyperchlorination of the water system, were positive for coliform bacteria. Of these, 4 (44%) of 9 were positive for E. coli. None of the E. coli isolates from water or ice possessed the ETEC toxin gene. In total, 12 (38%) of 32 isolates from these 3 outbreaks were resistant to 3 antimicrobial agents. Since 1990, 39 (46%) of 84 ETEC isolates from cruise ship outbreaks tested at the CDC have been resistant to 3 antimicrobial agents. Before 1990, none of 38 ETEC isolates obtained from cruise ship outbreaks was resistant to 3 antimicrobial agents. ETEC isolates were susceptible to ciprofloxacin, nalidixic acid, kanamycin, gentamicin, and ceftriaxone. Discussion These recent investigations suggest that, after a hiatus of more than a decade, waterborne disease outbreaks on cruise ships are reemerging as a serious problem. The 3 outbreaks of ETEC gastroenteritis described here affected 11300 people. The implicated vehicles of ETEC transmission were contaminated water bunkered by cruise ships in overseas ports and ice made from this water. ETEC, like most other bacterial organisms, is readily killed by chlorination. ETEC was not isolated from water on these ships because, by the time the investigative team arrived, the ship s water had been highly chlorinated. To ensure microbiologic safety of water, the VSP recommends that freechlorine residuals be maintained at 2 3 ppm during bunkering and 0.2 ppm at all times throughout the ship s water distribution system [10]. Although each ship attempted to comply with these recommendations, brief failures in water treatment systems on 2 ships appear to have allowed the ETEC to survive. These temporary failures suggest that the reliability of chlorination systems on ships needs to be improved. By taking prudent precautions to ensure the microbiologic safety of water bunkered in foreign ports, cruise lines can further protect the health of passengers and reduce the risk of outbreaks of waterborne diseases. ETEC gastroenteritis is characterized by watery diarrhea accompanied by abdominal cramps and infrequent vomiting. The infection often resolves without specific treatment, although antimicrobial agents, if given early, may shorten the duration and decrease the severity of illness [11, 12]. Unlike other bacterial pathogens, ETEC is not identified by routine stool cultures. Identification of ETEC requires special techniques, such as PCR, immunoassays, DNA probes, cell culture, or animal assays. Physicians and laboratories should consider ETEC as a possible etiologic agent of gastroenteritis in cruise ship passengers when routine laboratory tests fail to identify a pathogen, and state or local health departments should be notified. Multidrug-resistant ETEC is increasingly common in developing countries [13], and resistance in ETEC isolated during US outbreaks has also increased [8]. ETEC isolates resistant to ampicillin, streptomycin, sulfisoxazole, tetracycline, and trimethoprim/sulfamethoxazole have been documented in cruise ship outbreaks [8]. The improper use of antimicrobial agents has probably contributed to the emergence and dissemination of multidrug-resistant ETEC. In the absence of more-prudent policies on antimicrobial use worldwide, the prevalence of multidrug-resistant ETEC can be expected to increase. ETEC infections usually are not treated with antimicrobials. If they are treated with antimicrobials, a quinolone is usually the drug of choice for traveler s diarrhea. In the 3 outbreaks described, all ETEC isolates tested were susceptible to both ciprofloxacin and nalidixic acid. Cruise ship passengers who are at risk for ETEC exposure both aboard ships and during on-shore visits overseas eventually could benefit from ETEC vaccines that are currently under development [14]. Although in some instances only 1 ETEC serotype has been identified from cruise ship outbreaks, more commonly, as in the outbreaks reported here, multiple serotypes are isolated from patients [8]. This may reflect sewage contamination of source water or implicated food vehicles of transmission. Stool cultures from some patients may have been negative for ETEC because of prior antimicrobial therapy, collection of stool specimens after illness recovery, or isolation and culture sensitivity. The absence of fecal coliforms usually is conclusive evidence that ETEC is not present. It is also possible that some of the diarrheal illness reported may have been caused by another microbial agent, since background diarrheal illness is estimated to be 1% 2% among cruise ship passengers. Although ETEC and other bacteria can survive freezing [15], this is the first report in which epidemiologic evidence suggests that ice was a vehicle for ETEC transmission. These investigations underscore the importance of conducting surveillance for gastrointestinal illness and investigating outbreaks on cruise ships. Improved water-handling practices on cruise ships, which include ensuring the safety of water sources, maintaining uninterrupted water disinfection, and careful monitoring of water quality with state-of-the-art systems, are critical to the prevention of similar outbreaks in the future.

JID 2000;181 (April) Enterotoxigenic E. coli Outbreaks 1495 Acknowledgments We acknowledge Linda Anderson, David Kim, and Daniel Harper from the Vessel Sanitation Program, Centers for Disease Control and Prevention (CDC), for technical assistance during these outbreak investigations. This research project received statistical support in part by an appointment to the Research Participation Program at the CDC, National Center for Infectious Diseases, Division of Bacterial and Mycotic Diseases, administered by the Oak Ridge Institute for Science and Education through an interagency agreement between the US Department of Energy and the CDC. References 1. Koo D, Maloney K, Tauxe R. Epidemiology of diarrheal disease outbreaks on cruise ships, 1986 through 1993. JAMA 1996;275:545 7. 2. Addiss DG, Yashuk JC, Clapp DE, Blake PA. Outbreaks of diarrheal illness on passenger cruise ships, 1975 85. Epidemiol Infect 1989;103:63 72. 3. Gorbach SL, Kean BH, Evans DG, Evans DJ, Bessudo D. Travelers diarrhea and toxigenic Escherichia coli. N Engl J Med 1975;292:933 6. 4. Merson MH, Morris GK, Sack DA, et al. Travelers diarrhea in Mexico: a prospective study of physicians and family members attending a congress. N Engl J Med 1976;294:1299 305. 5. Rosenberg ML, Koplan JP, Wachsmuth IK, et al. Epidemic diarrhea at Crater Lake from enterotoxigenic Escherichia coli. Ann Intern Med 1977;86: 714 8. 6. Hedberg CW, Savarino SJ, Besser JM, et al. An outbreak of foodborne illness caused by Escherichia coli O39:NM, an agent not fitting into the existing scheme for classifying diarrheogenic E. coli. J Infect Dis 1997;176:1625 8. 7. MacDonald KL, Eidson M, Strohmeyer C, et al. A multistate outbreak of gastrointestinal illness caused by enterotoxigenic Escherichia coli in imported semisoft cheese. J Infect Dis 1985;151:716 20. 8. Dalton CB, Mintz ED, Wells JG, Bopp CA, Tauxe RV. Outbreaks of enterotoxigenic Escherichia coli infection in American adults: a clinical and epidemiologic profile of an emerging pathogen. Epidemiol Infect 1999;123: 9 16. 9. Olsvik O, Strockbine NA. PCR detection of heat-stable, heat-labile, and Shiga-like toxin genes in Escherichia coli. In: Persing DH, Smith TF, Tenover FC, White TJ, eds. Diagnostic molecular microbiology. Washington, DC: American Society for Microbiology, 1993:271 76. 10. Centers for Disease Control and Prevention. Vessel Sanitation Program operations manual. Atlanta: US Department of Health and Human Services, US Public Health Service, 1989. 11. DuPont HL, Reves RR, Galindo E. Antimicrobial therapy for travelers diarrhea. Rev Infect Dis 1986;8:S217 22. 12. Gorbach SL. Bacterial diarrhea and its treatment. Lancet 1987;2:1378 82. 13. Lester SC, del Pilar Pla M, Wang F, Perez Schael I, Jiang H, O Brien TF. The carriage of Escherichia coli resistant to antimicrobial agents by healthy children in Boston, in Caracas, Venezuela, and in Qin Pi, China. N Engl J Med 1990;323:285 9. 14. Savarino SJ, Bassily S, Brown FM, et al. Oral, inactivated, whole cell enterotoxigenic Escherichia coli plus cholera toxin B subunit vaccine: results of the initial evaluation in children. PRIDE study group. J Infect Dis 1999;179:107 14. 15. Dickens DL, DuPont HL, Johnson PC. Survival of bacterial enteropathogens in the ice of popular drinks. JAMA 1985;253:3141 3.