In the United States, Shigella sonnei causes an estimated 360,000 cases of gastroenteritis each year; an estimated 59% of these cases occur in children <10 years old.1 Although many cases present with watery diarrhea, not infrequently, diarrhea may also contain blood and mucus. Illness in children is usually brief and self-limited, but among daycare center attendees hospitalization rates up to 3% have been reported.2,3 During the past 2 decades, numerous outbreaks of S. sonnei infection have been associated with daycare centers.4 This is because S. sonnei has a very low infectious dose, and shigellosis can be transmitted from person to person or through contact with contaminated objects.5,6 This situation is exacerbated in daycare centers, where children exhibit different levels of toileting skills and may not be able to wash their own hands adequately, and staff members may be inconsistent in their handwashing practices.7–10
Control measures for shigellosis outbreaks in daycare centers focus primarily on hygiene promotion, including instructions on proper handwashing and diapering practices.11 Many states also require children with shigellosis to be excluded from the daycare center until documentation of 2 consecutive follow-up stool specimens that do not yield S. sonnei. Although antimicrobial agents are not required for this generally mild disease, these exclusion policies promote antimicrobial treatment because antimicrobial agents can shorten the duration of carriage, and therefore hasten return of a child to the daycare center.11
Antimicrobial treatment for shigellosis among children attending daycare centers has become a common practice.12 Concurrently, shigellosis caused by multidrug resistant (MDR) strains has emerged. For example, among 1278 S. sonnei isolates received by the National Antimicrobial Resistance Monitoring System at the Centers for Disease Control and Prevention between 1999 and 2003, 80% were resistant to ampicillin, 47% to trimethoprim-sulfamethoxazole, and 38% to both agents.13 In 2004, 39.4% of isolates were resistant to both drugs. (available at: http://www.cdc.gov/narms/NARMSAnnualReport2004.pdf, p 30). This has led to treatment controversies in children with shigellosis. Ampicillin and trimethoprim-sulfamethoxazole are the safest, least expensive, and in the past were the most common drugs used to treat pediatric shigellosis, but given the increases in resistance they can no longer be used empirically. Fluoroquinolones are now recommended by the American Academy of Pediatrics as an option for the treatment of shigellosis, but they have an uncertain safety profile in this age group.11,14,15 Macrolides, particularly azithromycin, are also recommended by the American Academy of Pediatrics for treatment of shigellosis, although data about clinical effectiveness are limited and no standardized guidelines for monitoring azithromycin resistance among Shigella species are currently available.16 In addition, the half-life of azithromycin excretion in stool is approximately 72 hours.17 Therefore, follow-up stool cultures may not give an accurate result until azithromycin is no longer excreted, thus extending the time required for follow-up testing. This delay might complicate patient follow-up and prolong exclusion from daycare centers. Ceftriaxone is effective, but is administered parenterally for 5 days.18 Although laboratories frequently report susceptibility to amoxicillin-clavulanate, the absorption of amoxicillin is too rapid to be effective against shigellae in the intestinal tract.19 As a result, physicians have very limited options for treating shigellosis caused by MDR strains in children.
These issues were highlighted by a large shigellosis outbreak caused by MDR S. sonnei among children in multiple daycare centers in 7 counties in the Kansas City metropolitan area in northwest Missouri that began in May 2005. Despite periodic visits by the local health authorities to affected daycare centers and the distribution of 2 health alerts during summer and fall of 2005, the outbreak persisted for >5 months and spread through multiple daycare centers. In October 2005, an investigation was conducted to determine risk factors for transmission within daycare centers, antimicrobial resistance of outbreak-associated strains, and antimicrobial treatment practices during the outbreak.
MATERIALS AND METHODS
Routine Surveillance Data
In Missouri, a confirmed shigellosis case was defined as a person with any Shigella species isolated from stool. All cases were routinely interviewed by the local health departments using a standardized questionnaire that included information on daycare center exposure and self-reported treatment for the illness. Shigella isolates were analyzed at the state public health laboratory with pulsed-field gel electrophoresis. Results are uploaded to the National PulseNet database. Our investigation focused on S. sonnei infection in licensed daycare centers. We abstracted data from case interviews conducted between May 1 and October 31, 2005 in the Kansas City, Missouri, metropolitan area to determine the number and demographic characteristics of shigellosis cases in the outbreak period and to identify daycare centers associated with shigellosis cases. The affected daycare centers were matched with a database of licensed daycare centers (LDCs) in the state of Missouri. We analyzed available self-reported treatment data for patients ≤18 years.
Daycare Center Case-Control Study
From October to November 2005, we attempted to enroll all LDCs operating in 1 of the 7 affected counties of the Kansas City metropolitan area in northwest Missouri with one or more confirmed shigellosis case during the outbreak period from May to October 2005. We calculated the daycare-specific secondary attack rate in each LDC by obtaining the number of laboratory-confirmed shigellosis cases in the LDC during the outbreak period, subtracting the index child from this number, and dividing by the number of children enrolled in the LDC at the time the first case within the daycare center was identified. To differentiate between daycare centers with high and low transmission of shigellosis, a case LDC was defined as a daycare center with a secondary attack rate of shigellosis ≥2% (median, 5%; range, 2%–25%) and control LDC as one with a secondary attack rate <2% (median, 0; range, 0%–1.2%).
We interviewed LDC directors using a standardized questionnaire administered over the telephone to determine the physical structure of the daycare center and hygiene practices within the facility. In addition, a staff member from each LDC was interviewed on-site about the daily operations of the daycare center, pertaining to procedures and protocols used for handwashing, diapering, and cleaning procedures at the time the first case within the LDC was identified. We also performed structured on-site inspections within the LDCs, focusing on classrooms, bathrooms, and diapering stations.
Healthcare Provider Survey
We sought to characterize attitudes and practices regarding the management of S. sonnei gastroenteritis among pediatric healthcare providers in Northwest Missouri. The Missouri Department of Health and Senior Services provided contact information for the 786 pediatric, family medicine, and emergency medicine physicians licensed in the state of Missouri and practicing within northwest Missouri zip codes 64012 to 64484. Because we wished to compare practices among the different specialties, we initially estimated a sample size of 130 and randomly selected healthcare providers to interview. The initial response rate was lower than anticipated among these physicians, so we attempted to contact an additional 132 healthcare providers. A standardized questionnaire about clinical management of shigellosis was administered over the telephone. Three attempts spanning at least 2 days were made to contact each healthcare provider.
We analyzed 31 S. sonnei isolates from patients who resided in northwest Missouri during the outbreak period. Ten of these isolates were obtained by gathering the final 2 isolates submitted to the state public health laboratory each month from June to October, and the remaining 21 isolates were sent to the CDC National Antimicrobial Resistance Monitoring System laboratory in Atlanta, Georgia as part of routine, national surveillance. An isolate was considered LDC associated if it was obtained from a person who attended or worked at an LDC, or lived with someone who attended or worked at an LDC. These isolates were tested at the National Antimicrobial Resistance Monitoring System laboratory for antimicrobial susceptibility to standard agents using the disk diffusion method according to Clinical and Laboratory Standards Institute for test performance and interpretation. Pulsed-field gel electrophoresis was performed using the PulseNet standard reference method,20 and patterns were compared with the PulseNet National Database for Shigella.
We analyzed data using Statistical Analysis System version 9.1 (SAS Institute, Cary, NC). Frequencies were generated for categorical data, whereas means, medians, and ranges were estimated for continuous variables. Univariate analysis of associations between selected exposures and shigellosis attack rates among the LDCs were assessed. We calculated 95% confidence intervals (95% CI) for odds ratios (OR) using the binomial exact distribution. An association was considered to be significant if the statistical test yielded a P value of <0.05 and the 95% CI did not include 1.
Routine Surveillance Data
From May 1 to October 31, 2005, the Missouri Department of Health and Senior Services reported 639 cases of laboratory-confirmed S. sonnei infection in Northwestern Missouri (Fig. 1). Among these cases, 490 (77%) occurred among children ≤18 years old. Among the 639 cases, 300 had a reported daycare center exposure; 176 (59%) reportedly attended or worked in an LDC, and 112 (37%) reportedly had a household contact who attended or worked in an LDC. Cases were associated with 44 LDCs. A total of 138 outbreak isolates were uploaded to the PulseNet National Shigella database; 30 homologous but nonidentical pulsed-field gel electrophoresis outbreak patterns were identified. Two patterns accounted for 28% of all isolates.
Daycare Center Case-Control Investigation
Of the 44 LDCs with shigellosis cases during the outbreak, 39 (89%) participated in the investigation. Among these 39 LDCs, 3 declined an on-site inspection, and 1 declined a staff member interview. Among participating LDCs, 18 were case-LDCs (secondary attack rate, 2%–25%) and 21 were control-LDCs (secondary attack rate, 0%–1%). No significant differences between case- and control-LDCs were found in the median number of enrolled children (case LDCs median, 66; range, 42–99 vs. control LDCs median, 73; range, 36–90), child-to-staff ratios (case and control LDCs median, 5; range, 4–6), total number of rooms (case LDCs median, 7; range, 6–14 vs. control LDC median, 8; range, 6–11), and total square footage per child (case LDCs median, 59; range, 39–93 vs. control LDCs median, 58; range, 47–117).
Three (16%) case-LDCs and 13 (68%) control-LDCs had at least 1 sink in each room (OR: 0.1; 95% CI: 0.02–0.5). Similarly, 6 (43%) case-LDCs and 15 (88%) control LDCs had a diapering station in all rooms where diapered children were present (OR: 0.1; 95% CI: 0.01–0.6) (Table 1). Scheduled handwashing was required before entry to 13 (95%) control-LDCs and 8 (86%) case-LDCs (OR: 0.3; 95% CI: 0.01–1.1). There were no significant differences between case and control LDCs in regards to other infrastructure, food handling, or cleaning, toileting, handwashing, or diapering procedures.
Healthcare Provider Survey
We were able to interview 31 of 262 healthcare providers (Table, Supplemental Digital Content 1, https://links.lww.com/A1139). Eighteen (57%) physicians reported they would use antimicrobial agents to treat children aged ≤18 years with shigellosis during the outbreak. Of these physicians, 9 (47%) reported they would prescribe azithromycin, 4 (24%) fluoroquinolones, and 1 (6%) trimethoprim-sulfamethoxazole. Twelve (38%) physicians reported ever having used a fluoroquinolone to treat children aged ≤18 years with diseases other than cystic fibrosis. None of the physicians reported they would treat a child <36 months with a fluoroquinolone for a disease other than cystic fibrosis.
Among the 13 healthcare providers that would not use antimicrobial agents to treat shigellosis caused by MDR strains, 10 (77%) based their decision on adverse drug effects, and 10 (77%) considered shigellosis to be a self-limited disease.
A total of 289 (45%) patients were interviewed by the Missouri Department of Health and Senior Services were reportedly treated with an antimicrobial agent during the diarrheal illness; 210 (73%) were children aged ≤18 years. Overall, during this outbreak 210 (43%) children ≤18 years old with shigellosis during this outbreak were treated with an antimicrobial agent. Azithromycin was the most commonly used antimicrobial agent, and was reportedly taken by 92 (44%) of the treated pediatric patients. The remaining pediatric patients reportedly took cephalosporins (n = 57, 27%), trimethoprim-sulfamethoxazole (n = 32, 15%), fluoroquinolones (n = 11, 5%), ampicillin (n = 3, 2%), or an unknown antimicrobial agent (n = 15, 7%) to treat the illness. Among the 11 pediatric patients who reported taking a fluoroquinolone for their illness, the median age was 2 years (range, 2–5).
The 31 isolates tested for antimicrobial susceptibility came from patients with a median age of 3 years (range, 1–55 years). Among the 10 isolates from patients that had been interviewed by the local health departments, 9 (90%) were from patients associated with an LDC; the median age of these 9 patients was 45 months (range, 16–67 months). Among all 31 isolates, 30 (95%) were resistant to ampicillin, 28 (90%) were resistant to trimethoprim-sulfamethoxazole, and 26 (84%) were resistant to chloramphenicol. Resistance to both ampicillin and trimethoprim-sulfamethoxazole was seen in 28 (90%) isolates. No resistance to amoxicillin/clavulanic acid, ceftriaxone, ciprofloxacin, gentamicin, or nalidixic acid was observed.
We report on an investigation conducted to identify risk factors for transmission and characterize treatment practices for shigellosis during one of the first, largest, and longest-lasting shigellosis outbreaks caused by MDR S. sonnei and affecting children in daycare centers. Local public health authorities implemented a variety of control measures, including educating daycare staff to promote good handwashing, diapering, and food preparation practices; excluding infected children and staff until 2 consecutive stool cultures obtained ≥24 hours apart and ≥24 hours after completing antimicrobial treatment were negative for S. sonnei; and issuing 2 public health alerts to healthcare providers about the local S. sonnei antimicrobial resistance patterns and advising avoidance of ampicillin and trimethoprim-sulfamethoxazole as treatment for patients during the outbreak. Despite these measures, the outbreak spread quickly and affected more than 40 daycare centers in 7 counties. Daycare centers without handwashing facilities in every room, and those enrolling diapered children but without diapering stations in every room were at a much greater risk of disease spread among attendees.
We also found that the outbreak strain was resistant to ampicillin and trimethoprim-sulfamethoxazole, the 2 drugs that, until recently, were most commonly recommended for empiric treatment of shigellosis in children. During this outbreak, children with shigellosis were frequently treated with azithromycin and occasionally with fluoroquinolones. There are no standardized guidelines for interpreting azithromycin susceptibility test results for Shigella,16 and increased use of azithromycin for treatment of shigellosis in the absence of surveillance for resistance to this agent may allow resistance to this drug to emerge unnoticed. The emergence of shigellosis caused by MDR strains highlights the urgent need for development of standards for azithromycin antimicrobial resistance monitoring and clarification of the optimal timing for follow-up stool cultures of patients after treatment with azithromycin.
This investigation also suggests that, given the limited antimicrobial options to treat shigellosis caused by MDR strains in young children, physicians, and patients may be more willing to accept the risks of potentially serious adverse drugs effects to treat the disease.14,15,21 Moreover, the use of fluoroquinolones for a generally mild disease like shigellosis may lead to fluoroquinolone resistance among Shigellae species and many other more virulent enteric or respiratory pathogens.
Our findings are subject to several limitations. First, because we enrolled relatively few LDCs, our power to detect differences among LDCs was limited. Next, we included only LDCs in the case-control study; thus, the factors influencing shigellosis transmission in unlicensed daycare centers during this outbreak are unknown. Third, little information was available on the socioeconomic status, level of education, and healthcare access of the families of children enrolled in the daycare centers, which limited our capacity to control for these characteristics when comparing case and control LDCs. We were also unable to assess directly the actual hygiene and diapering practices within daycare centers during the outbreak. Additionally, physician participation in the healthcare provider community survey was low, and the findings of this survey may not be representative of healthcare providers in the region. Finally, we were unable to assess the degree to which health alerts distributed to healthcare providers before this investigation began altered clinical practice and perceptions regarding shigellosis.
The emergence of shigellosis caused by MDR Shigella in the United States highlights the importance of prevention and rapid control of outbreaks in daycare centers. Appropriate handwashing and diapering infrastructure and practices are critical in minimizing the transmission of shigellosis in the child care environment.22 Scheduling handwashing sessions upon arrival at the daycare center, before meals, and after playing outdoors, supervising handwashing among young children, and eliminating water play areas have reduced the spread of shigellosis within daycare centers and to the community.23 Though current state regulations in Missouri do not allow alcohol-based hand sanitizers in daycare centers, these products have been associated with decreased enteric illness among families of children enrolled in daycare centers, and with appropriate supervision, could complement traditional handwashing with soap in daycare centers.24
Evidence of the efficacy of exclusion policies requiring 2 consecutive negative cultures rather than a single negative culture or simply resolution of symptoms to prevent transmission of shigellosis in daycare centers is limited. These policies, intended to reduce the likelihood of secondary transmission, may result in many days or weeks of daycare center exclusion, missed days of work for parents, and additional healthcare expenses. Rather than obtain 2 successive stool cultures, a parent may opt to take the child to another daycare center that is unaware of the child's illness. This could introduce the infection to another facility and further propagate the outbreak.25 Exclusion policies also implicitly encourage antimicrobial treatment of shigellosis in children who attend daycare centers,26–28 and may result in unnecessary adverse drug events in children and increased antimicrobial resistance among enteric and respiratory pathogens. Because shigellosis attack rates typically differ among age groups within daycare centers,2,29 treating or excluding children less likely to transmit the disease to others—that is, children who are fully toilet trained and capable of washing their own hands well—might have little impact on transmission rates within daycare facilities.
Cohorting convalescing children (ie, asymptomatic children who are culture-positive) by allowing them to attend the daycare center but prohibiting them from interacting with other well children has been used in combination with other measures to successfully control outbreaks associated with daycare centers and neonatal intensive care units.8,23,25,30 In these studies, additional measures like increased emphasis on handwashing and diapering hygiene, exclusion of persons with active diarrhea, and use of appropriate antimicrobial treatment may have all contributed to the decrease in disease transmission, although the relative contribution of each factor could not be defined. However, regulations in some states do not allow cohorting shigellosis cases in daycare centers, and in some instances cohorting may not be feasible due to limitations in the logistical and infrastructural capacity of the daycare center. Further research is needed to assess the effectiveness of cohorting within daycare centers and to establish comprehensive guidelines for implementing these measures. Moreover, in settings where cohorting measures can be appropriately implemented and enforced, the need for antimicrobial treatment should be determined.
Given the current frequency of resistance to ampicillin and trimethoprim-sulfamethoxazole among S. sonnei strains, the uncertain safety of administering fluoroquinolones to children, the difficulties in monitoring azithromycin resistance, and the lack of an appropriate vaccine, public health measures should focus on prevention of shigellosis outbreaks through appropriate hygiene practices and, where possible and allowed by state regulations, cohorting convalescing children in daycare centers. This investigation was presented in part at the American Academy of Pediatrics National Conference and Exhibition, Atlanta, October 2006, and at the National Foodborne Epidemiology Meeting, Miami, March 2006.
The authors thank the assistance of Julia Adams, Rex Archer, Ronda Charboneau, Thomas Emerson, Barbara Dawson, Kate Donaldson, Ellen Dorshow-Gordon, Patrick Franklin, Ron Griffin, Kimberly Henson, Lisa Hubbert, Shawna Jackson, Michele Lukenbill, Drew Pratt, Misha Odell, Jill Thompson, Tiffany Wilkinson, Joann Rudroff, Harvey Marx, Cheryl Bopp, and Mike Hoekstra.
1. Mead PS, Slutsker L, Dietz V, et al. Food-related illness and death in the United States. Emerg Infect Dis
2. Shane AL, Tucker NA, Crump JA, et al. Sharing Shigella
: risk factors for a multicommunity outbreak of shigellosis. Arch Pediatr Adolesc Med
3. Suspiro A, Menezes L. An outbreak of shigellosis in a child care institution in Queluz, Portugal-1995. Euro Surveill
4. Centers for Disease Control. Community outbreaks of shigellosis–United States. Morb Mortal Wkly Rep.
5. Gupta A, Polyak CS, Bishop RD, et al. Laboratory-confirmed shigellosis in the United States, 1989–2002: epidemiologic trends and patterns. Clin Infect Dis
6. DuPont HL, Levine MM, Hornick RB, et al. Inoculum size in shigellosis and implications for expected mode of transmission. J Infect Dis
7. Mohle-Boetani JC, Stapleton M, Finger R, et al. Communitywide shigellosis: control of an outbreak and risk factors in child day-care centers. Am J Public Health
8. Centers for Disease Control and Prevention. Day care-related outbreaks of rhamnose-negative Shigella sonnei
–six states, June 2001-March 2003. Morb Mortal Wkly Rep.
9. Centers for Disease Control. Shigellosis in child day care centers–Lexington-Fayette County, Kentucky, 1991. Morb Mortal Wkly Rep.
10. Pickering LK, Bartlett AV, Woodward WE. Acute infectious diarrhea among children in day care: epidemiology and control. Rev Infect Dis.
11. American Academy of Pediatrics. Red Book: Report of the Committee on Infectious Diseases.
Elk Grove, IL: American Academy of Pediatrics; 2006.
12. Finkelstein JA, Metlay JP, Davis RL, et al. Antimicrobial use in defined populations of infants and young children. Arch Pediatr Adolesc Med
13. Sivapalasingam S, Nelson JM, Joyce K, et al. High prevalence of antimicrobial resistance
isolates in the United States tested by the National Antimicrobial Resistance
Monitoring System from 1999 to 2002. Antimicrob Agents Chemother
14. Leibovitz E. The use of fluoroquinolones in children. Curr Opin Pediatr
15. Velissariou IM. The use of fluoroquinolones in children: recent advances. Expert Rev Anti Infect Ther
16. Jain SK, Gupta A, Glanz B, et al. Antimicrobial-resistant Shigella sonnei
: limited antimicrobial treatment options for children and challenges of interpreting in vitro azithromycin susceptibility. Pediatr Infect Dis J
17. Jain R, Danziger LH. The macrolide antibiotics: a pharmacokinetic and pharmacodynamic overview. Curr Pharm Des
18. Agence Française de Sécurité Sanitaire des Produits de Santé. Antibiotic treatment of Shigella sonnei
gastroenteritis [in French]. Presse Med.
19. Nelson JA, Haltalin KC. Amoxicillin less effective than ampicillin against Shigella
in vitro and in vivo: relationship of efficacy to activity in serum. J Infect Dis.
20. Swaminathan B, Barrett TJ, Hunter SB, et al. CDC PulseNet Task Force. PulseNet: the molecular subtyping network for foodborne bacterial disease surveillance, United States. Emerg Infect Dis
21. Gendrel D, Chalumeau M, Moulin F, Raymond J. Fluoroquinolones in paediatrics: a risk for the patient or for the community? Lancet Infect Dis
22. Centers for Disease Control and Prevention. Outbreaks of multidrug-resistant Shigella sonnei
gastroenteritis associated with day care centers–Kansas, Kentucky, and Missouri, 2005. Morb Mortal Wkly Rep.
23. Hoffman RE, Shillam PJ. The use of hygiene, cohorting, and antimicrobial therapy to control an outbreak of shigellosis. Am J Dis Child
24. Sandora TJ, Taveras EM, Shih MC, et al. A randomized, controlled trial of a multifaceted intervention including alcohol-based hand sanitizer and hand-hygiene education to reduce illness transmission in the home. Pediatrics
25. Tauxe RV, Johnson KE, Boase JC, et al. Control of day care shigellosis: a trial of convalescent day care in isolation. Am J Public Health
26. Garrett V, Bornschlegel K, Lange D, et al. A recurring outbreak of Shigella sonnei
among traditionally observant Jewish children in New York City: the risks of daycare and household transmission. Epidemiol Infect
27. Mahoney FJ, Farley TA, Burbank DF, et al. Evaluation of an intervention program for the control of an outbreak of shigellosis among institutionalized persons. J Infect Dis
28. Bachrach SJ. Successful treatment of an institutional outbreak of shigellosis. Clin Pediatr (Phila)
29. Jonsson J, Alvarez-Castillo Mdel C, Sanz JC, et al. Late detection of a shigellosis outbreak in a school in Madrid. Euro Surveill.
30. Munoz FM, Campbell JR, Atmar RL, et al. Influenza A virus outbreak in a neonatal intensive care unit. Pediatr Infect Dis J