During June 2001–March 2003, outbreaks of Shigella sonnei infections were reported in Delaware, Maryland, New Jersey, North Carolina, South Carolina, and Virginia (Figure). Five- to fortyfold increases in statewide shigellosis rates were observed during this period. These increases were attributed primarily to outbreaks in multiple day care settings that became prolonged and communitywide. S. sonnei isolates from these states, as well as from New York City and Philadelphia, were similar genetically by pulsed-field gel electrophoresis (PFGE). Many of these isolates lacked the capacity to ferment rhamnose, which is unusual for S. sonnei. This report summarizes these outbreaks and describes the laboratory characteristics that link them. The findings underscore the importance of rapid and coordinated public health responses to isolated outbreaks of shigellosis.
Shigellosis is a nationally reportable disease. During June 2001–March 2003, approximately 3,081 laboratory-confirmed cases of S. sonnei were reported from the six states through the Public Health Laboratory Information System (PHLIS). Each state or city health department investigated outbreaks independently and submitted case counts, including cases not reported through PHLIS; demographic information and laboratory data also were submitted. A day care–related case was defined as S. sonnei infection in a child attending day care or in a close contact of a child attending day care. The extent of laboratory testing, including PFGE, varied substantially from state to state on the basis of available resources and health department policies. Selected isolates and PFGE patterns were submitted to the National Molecular Subtyping Network for Foodborne Disease Surveillance (PulseNet) for comparison among states.
State Reports
Virginia. A day care–related outbreak of shigellosis attributed to S. sonnei began in southeastern Virginia in June 2001 and subsequently became regional. During June 2001–March 2003, a total of 876 laboratory-confirmed cases of shigellosis were reported. The median age of patients was 7 years (range: 0–87 years); 57% were female.
Maryland. During November 2001–March 2003, a total of 1,222 culture-confirmed cases were reported, a tenfold increase above baseline levels during 1998–2000. The outbreak was concentrated in Baltimore City (727 [59%] cases) and Baltimore County (296 [24%]). The median age of patients was 6 years (range: 0–101 years); 54% were female, and 72% were non-Hispanic black. A total of 250 (20%) cases were known to be day care–related. One death occurred in a boy aged 12 years.
New Jersey. During February–October 2002, a total of 453 culture-confirmed cases of S. sonnei in three adjacent counties were reported, representing a fortyfold increase compared with previous years. The majority (73%) of these cases occurred in the Trenton area. The median age of patients was 5 years (range: 1–79 years); 59% were female, and 56% were day care, camp, or elementary school attendees.
South Carolina. During June 2002–March 2003, a total of 172 laboratory-confirmed cases of shigellosis were reported. The median age of patients was 5 years (range: 0–88 years); 54% were female, and 25% were non-Hispanic white. Approximately 55% of cases were day care–related.
Delaware. During June 2002–March 2003, a total of 506 culture-confirmed cases were reported, representing a twentyfold increase above baseline during 1997–2001. A total of 457 (90%) of these cases occurred in New Castle County, including 324 (64%) in Wilmington. The median age of patients was 4 years (range: 0–69 years); 54% were female, and 52% were non-Hispanic black. A total of 200 (40%) cases were day care–related.
North Carolina. During August–December 2002, a total of 206 laboratory-confirmed cases of shigellosis were reported in three counties. A separate outbreak during October 2002–March 2003 in Mecklenburg County resulted in 729 cases. Epidemiologic investigations indicated that these outbreaks began in day care facilities and progressed to elementary schools. Among these patients, the median age was 5 years (range: 0–56 years); 52% were female, and 75% were non-Hispanic black. Statewide, 1,705 cases were reported during June 2001–March 2003, a total of 935 (55%) of which were linked to these two outbreaks.
New York City. During September 2002–April 2003, a total of 115 culture-confirmed cases were reported within traditionally observant Jewish communities in two Brooklyn neighborhoods, representing a two- to tenfold increase compared with previous years. The median age of patients was 2 years (range: 0–89 years); 57% were male. No cases associated with common schools or day care facilities were identified.
Pennsylvania. During 2002, a total of 317 cases of shigellosis were reported to the Pennsylvania Department of Health. This did not represent a statistically significant increase compared with previous years. Of these cases, 117 (37%) occurred in Philadelphia County. Statewide, 10 tested isolates were indistinguishable from the outbreak strain by PFGE. The median age of patients was 5 years (range: 0–57 years); 50% were female, and 60% were non-Hispanic black. A total of five cases were day care–related. Eight patients resided in Philadelphia but were not linked epidemiologically. However, the Philadelphia Department of Public Health reported a new day care–related outbreak of shigellosis beginning in April 2003, resulting in 706 laboratory-confirmed cases communitywide, 298 (42%) of which were linked to day care facilities.
Laboratory Characteristics
During June 2001–March 2003, a total of 1,349 S. sonnei isolates from the affected states were reported to PulseNet. Among these isolates, two dominant PFGE patterns differing by a single band were identified. These two patterns accounted for 505 (37%) and 382 (28%) isolates. Seven other distinct patterns differed from the dominant pattern by no more than three bands and accounted for an additional 271 (20%) cases.
PFGE and rhamnose fermentation results were available for 386 isolates. Among 246 isolates with either of the two dominant PFGE patterns, 241 (98%) were rhamnose-negative. In contrast, among 87 isolates with PFGE patterns that differed from the dominant pattern by more than three bands, seven (8%) were rhamnose-negative. Two states (Delaware and Virginia) provided rhamnose fermentation results for 627 isolates that did not undergo PFGE testing; 94% of these isolates also were rhamnose-negative.
Antimicrobial susceptibility results were available for 379 isolates; 91% (342/375) were resistant to ampicillin, 89% (205/230) had either intermediate or full resistance to amoxicillin/clavulanate, 28% (106/379) had either intermediate or full resistance to trimethoprim/sulfamethoxazole, and 24% (89/375) were resistant to both ampicillin and trimethoprim/ sulfamethoxazole.
Public Health Interventions
All health departments excluded children with diarrhea from day care and did not allow them to return until their diarrhea had ceased. All but one health department did not allow children to return until two stool cultures testing negative for Shigella had been obtained at least 24 hours after completing antibiotics and 24 hours apart. Delaware allowed diapered children to return after completing antibiotic treatment and nondiapered children to return after 48 hours of antibiotic treatment (without culture). Several states encouraged cohorting convalescing children within day care facilities, but most facilities were unable to do so. State and local health department employees inspected day care centers, provided hand-washing instruction and flyers, and directly observed hand washing by day care employees and attendees. All health departments alerted clinicians, day care providers, and the community, including parents of day care attendees, about the outbreak and encouraged hand washing. Certain states also targeted schools, community pools, and other community centers. Treatment recommendations varied. Certain health departments did not recommend treatment with antimicrobials except in severe cases, whereas others encouraged treatment of all laboratory-confirmed cases to reduce bacterial shedding and transmission. All recommended using antimicrobial resistance data to guide the selection of treatment agents.
Reported by: J Totaro, Maryland Dept of Health and Mental Hygiene. C Tan, MD, New Jersey Dept of Health and Senior Svcs. V Reddy, MPH, New York City Dept of Health and Mental Hygiene, New York. K Dail, MEd, M Davies, MD, P Jenkins, EdD, JM Maillard, MD, Epidemiology Section, North Carolina Div of Public Health. DM Toney, PhD, Virginia Div of Consolidated Laboratory Svcs, Richmond; J Murphy, DVM, Virginia Dept of Health. A Beall, E Mintz, MD, Div of Bacterial and Mycotic Diseases, National Center for Infectious Diseases; M Drees, MD, A Shane, MD, EIS officers, CDC.
Editorial Note:
Multicommunity outbreaks of shigellosis are an ongoing public health challenge whose management and control demands considerable time, effort, and expense from health departments, day care staff, and affected communities. The prolonged multistate increase in shigellosis in the south and mid-Atlantic areas described in this report is representative of numerous similar S. sonnei outbreaks that have occurred during the previous two decades (1–4).
During 1996–2001, the number of S. sonnei isolates reported to PHLIS remained stable, averaging 9,024 isolates per year (range: 7,363–10,262 isolates per year) (5). The median age of patients with S. sonnei infection during this 6-year period was 8–9 years; 55% of patients were female. S. sonnei is the predominant cause of shigellosis in the United States, accounting for approximately 75% of all reported cases. A high proportion of these infections were likely associated with day care centers. Because few organisms are required to cause infection, shigellosis spreads easily from person to person when breaches in hand washing and sanitation occur. Intra- and intercommunity propagation of shigellosis is facilitated by the challenge of maintaining adequate hygeiene and sanitation in day care centers, the high proportion of mild and asymptomatic Shigella infections, and frequent contact between children who attend one day care center and their friends and relatives who attend other centers. In addition, the emergence of antimicrobial resistance reduces treatment options for children with moderate-to-severe clinical illness.
A combination of laboratory methods to characterize S. sonnei isolates proved useful in defining and monitoring these outbreaks. Biochemical profiling revealed an unusual trait that helped identify potential outbreak-associated isolates for subsequent molecular testing. In these outbreaks, 97% of tested isolates lacked the ability to ferment rhamnose, a trait observed in <15% of S. sonnei isolates received by the reference laboratory at CDC during 1974–2002. Molecular methods, including PFGE, provided information regarding the similarity of isolates. Nine related PFGE patterns were associated with these outbreaks; two patterns accounted for 66% of the isolates. In contrast to most common-source bacterial foodborne disease outbreaks, isolates from communitywide outbreaks caused by S. sonnei commonly demonstrate several different but highly related PFGE patterns.
Multitiered interventions are necessary to manage and control outbreaks of day care–associated shigellosis, and these must be tailored to each community (6). Notification through the news media and through direct communication with day care operators, staff, parents, and the medical community helps increase awareness of an outbreak and encourages use of effective control strategies such as supervised hand washing and the exclusion of symptomatic children from day care. Onsite educational efforts by health department staff, including observation of hand-washing and toilet facilities and activities in affected and high-risk day care centers, is labor-intensive but probably more effective than mass distribution of educational materials (7). Cohorting of asymptomatically infected children and staff in day care permits asymptomatic culture-positive day care attendees to remain under supervised care (8,9). When this approach is not feasible, the requirement for two negative stool cultures before a child can return to day care can be used to ensure that a child is no longer infectious. However, strict exclusion strategies might lead to propagation of an outbreak if excluded day care attendees are placed in alternative child care settings. For this reason, decisions about when a child with shigellosis is permitted to return to the licensed day care setting require balancing the responsibility to halt transmission within a facility with the needs of the child's family.