Methicillin-resistant Staphylococcus aureus (MRSA) infection has long been associated with exposure in health-care settings but emerged in the late 1990s among previously healthy adults and children in the community. Community-associated MRSA (CA-MRSA) infections most commonly are skin and soft-tissue infections; however, certain cases can progress to invasive tissue infections, bacteremia, and death.1 This report describes two independent investigations by local health departments, assisted by CDC, into outbreaks of MRSA skin infection among otherwise healthy, full-term newborns delivered at hospitals in Chicago, Illinois, and Los Angeles County, California. In both locations, MRSA transmission likely occurred in the newborn nursery; however, laboratory testing identified the MRSA strain as one that was described initially in association with CA-MRSA infections and outbreaks and that differs from predominant health-care–associated MRSA (HA-MRSA) strains. The findings from these investigations underscore (1) the need for health-care providers to be aware that MRSA can cause skin infections among otherwise healthy newborns and (2) the importance of adhering to standard infection-control practices,* including consistent hand hygiene, in newborn nurseries.
In October 2004, the Chicago Department of Public Health was notified of a cluster of seven MRSA skin infections among otherwise healthy, full-term newborns delivered at a Chicago hospital (hospital A). The health department investigated, seeking to identify other cases among hospital A newborns who were hospitalized after discharge or brought to the hospital's emergency department or pediatric and well-baby clinics. A case was defined as an infection in a newborn aged <30 days delivered at hospital A during May-December 2004 with a skin lesion from which MRSA was isolated. A total of 11 cases were identified. Two patients had single and nine patients had multiple pustules, vesicles, or blisters on the neck (five patients), groin (five), perineum (four), ears (two), legs (two), chin (one), and trunk (one). Seven of the 11 patients had multiple affected body sites.
Births of nine (82%) of the infants were by cesarean delivery. Median age at symptom onset was 7 days (range: 4-23 days); nine (82%) infants were male. Median hospital stay after delivery was 4 days (range: 3-10 days). One infant had symptoms of infection while still hospitalized on day 6. Symptom onset for the other 10 infants occurred 1-18 days (median: 5 days) after discharge from the newborn nursery. Ten infants received topical antimicrobial therapy (i.e., mupirocin or neosporin), and three of those 10 also received oral antimicrobials (i.e., cefaclor, cephalexin, or clindamycin) for their skin infections; none were treated with incision and drainage. One infant was hospitalized as a result of his infection; all recovered without incident. None of the mothers of the infants reported having current or recent skin lesions. No likely sources of MRSA exposure were identified outside of the hospital environment (e.g., family members or close contacts who had skin lesions or recognized risk factors for MRSA infection†).
In January 2005, nasal cultures were obtained from 135 health-care workers (HCWs) in the labor and delivery, postnatal, and newborn nursery wards who were likely to have had direct contact with one or more of the patients. One physician and one nurse who attended to newborns in the nursery were found to have nasal MRSA colonization. Isolates from the two HCWs and six of the 11 infants were available for characterization by pulsed-field gel electrophoresis (PFGE) and identification of toxin genes by a CDC laboratory. All eight isolates were identified as pulsed-field type (PFT) USA300 and contained genes for the Panton-Valentine leukocidin toxin, which has been associated with necrotizing pneumonia and primary skin infections.2 Isolates from the two HCWs and five of the infants were indistinguishable from one another by PFGE and also indistinguishable from MRSA strain USA300-0114, which has been associated with CA-MRSA outbreaks and sporadic infections in multiple states.3 Another strain of PFT USA300 was isolated from the remaining infant.
To prevent further transmission of MRSA in the nursery, adherence to standard infection-control measures, hand hygiene, and environmental cleaning were reinforced through in-service training and direct observation. In addition, the two MRSA-colonized HCWs were restricted from work until they completed a course of intranasal mupirocin and a second nasal culture tested negative for MRSA. As of March 27, 2006, no subsequent cases had been reported.
In January and June 2004, the Los Angeles County Department of Health Services was notified of two clusters of MRSA skin infections among newborns delivered at a Los Angeles County hospital (hospital B). A case was defined as a culture-confirmed MRSA skin infection in a newborn delivered at hospital B with onset during November 1, 2003–June 14, 2004, within 21 days after discharge from the well-baby nursery. Eleven cases were identified during two outbreak periods: November-December 2003 and May-June 2004. All 11 patients were males with pustular-vesicular lesions in the groin region; births of seven (64%) were by cesarean delivery. Median nursery stay after delivery was 4 days (range: 2-5 days). Symptom onset occurred at a median of 3 days (range: 1-17 days) after discharge from the nursery.
Eight (73%) of the 11 infants were hospitalized and treated with parenteral antimicrobials; all recovered without adverse sequelae. The remaining three infants were either administered topical antimicrobial agents or not treated. Characterization of the seven available MRSA isolates by PFGE, performed by the Public Health Laboratory of the Los Angeles County Department of Health Services, confirmed that the outbreak strain was USA300-0114, the same MRSA outbreak strain as identified in Chicago.
Investigators elected not to test HCWs for MRSA colonization at hospital B because no single HCW had more contact than others with all of the infected infants. Sources outside the hospital (e.g., family members or household contacts) were excluded. Hospital HCWs were provided education to reinforce routine hospital infection-control practices, including proper hand hygiene. In addition, use of gloves and gowns for all patient contacts was instituted, newborns were bathed with antibacterial soap before discharge, and the frequency and intensity of environmental cleaning of the nursery were increased. As of March 27, 2006, no subsequent cases had been reported.
J Watson, MD, RC Jones, MPH, C Cortes, SI Gerber, MD, Chicago Dept of Public Health; RG Golash, MS, J Price, MS, Illinois Dept of Public Health Laboratory. E Bancroft, MD, L Mascola, MD, Los Angeles County Dept of Health Svcs. RJ Gorwitz, MD, DB Jernigan, MD, Div of Healthcare Quality Promotion, National Center for Infectious Diseases; L James, MMED, DM Nguyen, MD, EIS officers, CDC.
The outbreaks described in this report involved otherwise healthy, full-term newborns who had onset of MRSA skin infections before or shortly after discharge from common nurseries. The 22 cases in this report are similar to six cases among newborns in a New York City hospital in 2002.4 As occurred in the New York City outbreak, an MRSA strain associated with community transmission was identified as the outbreak strain in Chicago and Los Angeles County.
Outbreaks of CA-MRSA have been reported among children in child-care settings, prisoners, military trainees, athletes, and men who have sex with men.3,5 To date, MRSA isolates from CA-MRSA infections have been genetically and phenotypically distinct from isolates from HA-MRSA infections.6 Whereas isolates from HA-MRSA infections generally are resistant to multiple classes of antimicrobial agents, those from CA-MRSA infections typically have been resistant only to beta-lactams (e.g., penicillins and cephalosporins) and macrolides (e.g., erythromycin).1,6 PFGE and other testing methods have identified a limited number of molecular types that have accounted for most isolates from CA-MRSA infections characterized in the United States.7 Health-care–associated transmission of MRSA strains with bacteriologic properties characteristic of CA-MRSA has recently been reported among postpartum women8 and infants in a neonatal intensive-care unit.9
The clusters of CA-MRSA infection described in this report involved skin and soft-tissue infections that appeared superficial; however, 41% of patients were hospitalized for treatment of their infections. Infection is often treated aggressively in newborns because of their immature immune function and potential for rapid deterioration. MRSA strains are resistant to all beta-lactam agents, which have been used for standard first-line antimicrobial therapy for skin infections in the community. Several potential alternative agents (e.g., tetracyclines or trimethoprim-sulfamethoxazole) are contraindicated or not recommended in newborns. Isolates from CA-MRSA infections also are commonly susceptible to gentamicin, rifampin, linezolid, and clindamycin. However, some isolates that appear clindamycin-susceptible and erythromycin-resistant on routine susceptibility testing can be induced to express resistance to clindamycin in vitro, and clindamycin treatment failure has been reported in association with invasive infections caused by such isolates. This inducible clindamycin resistance can be detected using a specialized laboratory test known as the D-zone test.‡ Vancomycin remains a first-line therapy for severe infections potentially caused by MRSA. Incision and drainage is considered standard therapy for purulent skin lesions. Some minor CA-MRSA skin infections can resolve without antimicrobial therapy.
Births of 16 of the 22 infants were by cesarean delivery. Because neonates whose births are by cesarean delivery remain in the hospital longer than neonates delivered vaginally, the role of cesarean delivery in MRSA infection is unclear. Comparative or prospective studies are needed to identify specific risk factors for MRSA acquisition and transmission among neonates.
A total of 20 of 22 infants with CA-MRSA infection in the Chicago and Los Angeles County outbreaks were male. Although the role of male sex in these outbreaks is unclear, male sex has been identified as a risk factor for staphylococcal colonization and infection among newborns in previous studies.10 A proclivity for involvement of the groin and perineal areas also was noted in the outbreaks described in this report. Skin in the diaper area might be particularly prone to staphylococcal infection because of the moist environment and friction from diapers, causing disruption of the epidermal barrier.
The implications of two HCWs colonized with the outbreak strain in Chicago are unclear. One or both of these HCWs might have introduced the MRSA strain into the nursery and transmitted the organism directly to the infants, or the HCWs might have become colonized as a result of contact with already colonized or infected newborns. In addition, because the same MRSA strain has been implicated in outbreaks in multiple states, colonization of the HCWs might have been unrelated to transmission in the nursery.
Clinicians should be aware that MRSA can cause skin infections and potentially more serious infections among healthy full-term newborns. These infections might be confused with other rash illnesses in newborns, such as infection caused by herpes simplex virus. Obtaining cultures of purulent skin lesions is important to guide therapy. Caretakers of newborns with skin infections should receive guidance on measures to prevent further transmission, including washing hands frequently and applying clean, dry dressings to draining lesions. Adherence to standard infection-control practices and strict hand hygiene should be enforced in all newborn nursery settings. HCWs should be encouraged to seek treatment promptly for skin lesions.
When transmission of MRSA occurs in a newborn nursery, standard infection-control practices should be reviewed and reinforced. Surveillance for skin lesions among patients, staff members, and visitors should be considered. The necessity for using other measures, such as universal use of gowns and gloves, antiseptic bathing of newborns, and surveillance cultures of HCWs and the environment is less clear. Culture surveys are not routinely recommended for HCWs for whom an epidemiologic link to MRSA transmission has not been established.§ However, cultures sometimes are performed to rule out potential sources of transmission in novel settings. When surveillance cultures of HCWs are conducted, they should target staff members likely to have had direct contact with patients with MRSA infections; a clear plan of action that will be taken on the basis of culture results should be established and communicated to staff members. Additional information regarding HA-MRSA and CA-MRSA infections is available at http://www.cdc.gov/ncidod/dhqp/ar_mrsa.html.
This report is based, in part, on technical assistance from the Public Health Laboratory of the Los Angeles County Dept of Health Services; GE Fosheim, MPH, B Jensen, MMSc, SK McAllister, LK McDougal, MS, Div of Healthcare Quality Promotion, National Center for Infectious Diseases; TJ Török MD, Office of Workforce and Career Development, CDC.
*Available at http://www.cdc.gov/ncidod/dhqp/gl_isolation_standard.html.
†Risk factors for HA-MRSA infection as defined in CDC's Active Bacterial Core Surveillance system include isolation of MRSA ≥2 days after hospitalization; history of hospitalization, surgery, dialysis, or residence in a long-term–care facility <1 year before the MRSA culture; presence of a permanent indwelling catheter or percutaneous medical device at the time of culture; or previous isolation of MRSA.
‡Available at http://www.phppo.cdc.gov/nltn/pdf/2004/2_hindler_d-test.pdf.
§Guidelines available at http://www.cdc.gov/ncidod/dhqp/gl_hcpersonnel.html.
Community-Associated Methicillin-Resistant. JAMA. 2006;296(1):36-38. doi:10.1001/jama.296.1.36