Foodborne Outbreak of Cryptosporidiosis—Spokane, Washington, 1997 FREE

MMWR. 1998;47:565-567

ON DECEMBER 29, 1997, the Spokane Regional Health District received reports of acute gastroenteritis among members of a group attending a dinner banquet catered by a Spokane restaurant on December 18. The illness was characterized by a prolonged (3-9 days) incubation period and diarrhea, which led public health officials to suspect a parasitic cause of the illness. Eight of 10 stool specimens obtained from ill banquet attendees were positive for Cryptosporidium using both modified acid-fast and auramine-rhodamine staining of concentrated specimens. This report summarizes the epidemiologic investigation of the outbreak, which suggests that foodborne transmission occurred through a contaminated ingredient in multiple menu items.

In a retrospective cohort study, a case was defined as diarrhea or abdominal cramping in a banquet attendee with onset within 10 days after the banquet. Of the 62 attendees, 54 (87%) had illnesses meeting the case definition; they became ill a median of 6 days (range: 3-9 days) after the banquet. Symptoms included diarrhea (98%), fever/chills (61%), headache (59%), body ache (54%), abdominal cramps (50%), nausea (28%), and vomiting (11%). Based on information from initial interviews, the median length of illness was 5 days (range: 1-13 days), but subsequently several persons reported that they had symptoms intermittently for a month or longer. Two persons were hospitalized, and six others sought health care for their illness.

The banquet buffet included 18 separate food and beverage items; seven items contained uncooked produce. No single food was significantly associated with illness. When menu items that contained green onions were combined, foods containing uncooked green onions (au gratin potatoes, romaine salad, and pasta salad) were reportedly eaten by all 51 case-patients who could recall and by three of four persons who were not ill and could recall (undefined relative risk, p=0.07).

The banquet food items were prepared or served by 15 food workers. Stool specimens were available from 14 food workers within 3-4 weeks of the banquet; specimens from two tested positive for Cryptosporidium. One of the two food workers was symptomatic at the same time as banquet attendees; the other was asymptomatic. A stool specimen from another food worker was not available for testing until 5 weeks after the outbreak and was negative; he reported that he worked for 2 days in December while experiencing diarrhea but he could not remember the dates of his illness. All three of these food workers reportedly ate food items served at the banquet associated with the outbreak.

The green onions were not washed before delivery at the restaurant. Food workers at the restaurant reported they did not consistently wash green onions before using them to prepare food or serving them to patrons.

To determine the extent of the outbreak, the health district requested by fax that Spokane area physicians report any patients with symptoms typical of cryptosporidiosis. No other cryptosporidiosis-like illnesses were identified at the time of the outbreak. Two other banquets catered by the restaurant on December 18 and 19 had menus similar to the banquet where the outbreak occurred; no illness was reported in either of these groups.

K Quinn, MPA, G Baldwin, P Stepak, MD, K Thorburn, MD, Spokane Regional Health District; C Bartleson, MPH, M Goldoft, MD, J Kobayashi, MD, P Stehr-Green, DrPH, State Epidemiologist, Washington Dept of Health. Div of Parasitic Diseases, National Center for Infectious Diseases, CDC.

Since 1993, three foodborne outbreaks of cryptosporidiosis have been reported in the United States. In 1993, an outbreak was associated with drinking unpasteurized, fresh-pressed apple cider¹; the apples used for the cider probably were contaminated when they fell to the ground in a cow pasture. In 1995, an outbreak was associated with eating chicken salad that may have been contaminated by a food worker who operated a day care facility in her home.² In 1996, an outbreak was associated with drinking commercially produced, unpasteurized apple cider³; the apples used for the cider may have become contaminated when they were washed with well water that had fecal contamination.

The outbreak described in this report had characteristics similar to others in the United States caused by enteric coccidian parasites (Cryptosporidium parvum and Cyclospora cayetanensis) in that case-patients had prolonged diarrhea; the incubation period averaged 6 days; and the attack rates were high.^4- 5 Physicians and public health officials should have a high index of suspicion for infection with coccidian parasites in patients with severe or prolonged watery diarrhea. Because most laboratories do not routinely test stool for either Cryptosporidium or Cyclospora,⁶ specific testing for these organisms generally must be ordered by a physician.

The high attack rate among banquet attendees made finding a statistically significant association with a particular menu item difficult. The strongest association between illness and eating a menu item was observed for food items containing uncooked green onions. This suggests that the onions were a possible source, but the data are inadequate to conclusively implicate them as the vehicle of infection. Available data do not exclude the possibility that multiple menu items may have been contaminated before arriving at the restaurant, contaminated by a food worker, or by cross-contamination during preparation.

This outbreak highlights several key issues for food workers. Uncooked produce should be throughly washed before being placed on kitchen work surfaces to prevent contamination of these surfaces. The FDA Food Code prohibits further bare-handed contact with fruits and vegetables after washing when they are intended for use in "ready-to-eat" foods except where approved by the regulating authority.⁷ Food preparation surfaces should be washed between preparation of different produce to prevent cross-contamination. Food workers should not work when experiencing a gastrointestinal illness. Persons infected with Cryptosporidium may intermittently shed oocysts in stool and remain infectious for up to 60 days after diarrhea has resolved; however, most persons will cease shedding within 2 weeks after resolution of their diarrhea.⁸ Therefore, food workers should be particularly meticulous about handwashing. Asymptomatic shedding probably occurs in persons exposed to the parasite who have developed some immunity, but the frequency of asymptomatic shedding is unknown.

MMWR. 1998;47:567-570

1 figure, 1 table omitted

ADENOVIRUSES are human pathogens that commonly infect the respiratory and gastrointestinal tracts.¹ Adenovirus infections are endemic, particularly among children, but also may cause epidemics of pharyngoconjunctival fever, keratoconjunctivitis, gastroenteritis, and acute respiratory disease (ARD) among military trainees. Outbreaks of ARD among adults in the civilian sector are rare.² In March 1997, an outbreak of acute respiratory disease (ARD) caused by adenovirus serotype 11 occurred among students at a job training facility in South Dakota. This report summarizes the epidemiologic and clinical features of this outbreak and discusses the change in availability of adenovirus vaccines for military use. The facility provides high school education and vocational training for 240 persons aged 16-21 years. New students matriculate year-round at 2-week intervals and remain for approximately 1-2 years. All students live on campus in one of four barracks-style dormitories (three for males and one for females). Sixty students are housed in each dormitory, with six to 10 persons per room sleeping in bunk beds. Students share a common dining hall. Routine medical care is provided by an infirmary nurse, who refers more severe illnesses to visiting physicians or local hospitals. Hospitalization discharge summaries are forwarded to the infirmary nurse.

Following the outbreak, a chart review was conducted at the facility's infirmary by the infirmary nurse. A case of lower respiratory tract infection (LRTI) was defined as physician-diagnosed pneumonia, an abnormal chest radiograph, or rales or wheezing on pulmonary auscultation in any student. A case of upper respiratory tract infection (URTI) was defined as coryza and sore throat without LRTI in any student. A case of ARD was defined as either URTI or LRTI in any student.

During March 8-28, a total of 146 (61%) students were diagnosed with ARD; 103 (71%) had URTI and 43 (29%) had LRTI. The ARD attack rate was higher among males than females (69% versus 37%, respectively, p <0.01). Although students with URTI and LRTI were similar in age and sex, frequencies of associated signs and symptoms differed between the two groups. Students with URTI were more likely than students with LRTI to have headache. Students with LRTI were more likely to have fever ≥101 F (≥38.3 C), pleuritic chest pain, shortness of breath, lymphadenopathy, vomiting, conjunctivitis, and dysuria (all p-values <0.05). Students with LRTI had higher fevers than students with URTI (median maximum temperatures: 103 F [39.4 C] versus 102 F [38.9 C], p <0.001). Five (12%) of 43 students with LRTI were hospitalized for 3 to 7 days each. One ill student with a poorly controlled seizure disorder suffered a respiratory arrest and required intensive care. Staff members at this facility also reported ARD symptoms during this time period.

Throat swab specimens were collected from seven ill students and inoculated into RMK and A549 cells. Six specimens yielded adenovirus, identified as subgenus B by the polymerase chain reaction assay, and as adenovirus 11 by microneutralization assays.^3,4 The sequences of a one kilobase region of the fiber gene were identical for all isolates, suggesting a single outbreak strain.

O Four Bear, Box Elder Job Corps, Nemo; LM Schaefer, LM Kellen-Anderson, SL Parker, DVM, State Epidemiologist, South Dakota Dept of Health. DP Schnurr, PhD, Viral and Rickettsial Diseases Laboratory, California Dept of Health Svcs. JC Gaydos, MD, Div of Preventive Medicine, Walter Reed Army Institute of Research, Washington, DC. Respiratory and Enteric Viruses Br, Div of Viral and Rickettsial Diseases, National Center for Infectious Diseases, CDC.

Although adenovirus-associated ARD outbreaks among military training populations are well-described, they have not been recognized among college students or other young adults in the civilian sector. However, the setting of this outbreak of adenovirus ARD is similar to settings of previous military ARD outbreaks. In both settings, young adults live in crowded conditions, and new groups of potentially susceptible persons are introduced regularly. This outbreak differed from military outbreaks because most adenovirus-associated ARD outbreaks among U.S. military trainees are associated with adenoviruses 4 and 7. Adenovirus 11 is most commonly recognized as a cause of hemorrhagic cystitis, acute hemorrhagic conjunctivitis, and illnesses among immunocompromised persons^1,5 and has rarely been associated with ARD in military trainees or in any other immunocompetent adult population.⁶

Outbreaks of adenovirus-associated ARD were common among U.S. military trainees before the 1970s, when routine vaccination of this group with oral vaccines against adenovirus serotypes 4 and 7 was instituted.^7,8 Although these vaccines were highly effective, their manufacture has been discontinued.⁹ Residual supplies of the vaccines will probably be exhausted in 1999, at which time large ARD outbreaks in military settings are expected, primarily in winter months.⁹ This outbreak underscores that adenoviruses can cause outbreaks of ARD among young adults, persons living in crowded conditions, and military recruits.

References: 9 available.