JSTOR

Methods

In response to the increase in reported clusters of MRSA infection and to the diverse measures enacted by the affected hospitals to control neonatal transmission of this organism, the CDPH convened a working group of infectious diseases physicians and infection control professionals from 9 Chicago‐area hospitals with level III NICUs. Level III NICUs in Illinois are defined as units able to treat infants who need the highest‐acuity intensive care services. In 2000, the median number of available beds per NICU reported by 8 of the 9 hospitals was 38 (range, 10‐48 beds), and the median number of patient‐days per year was 10,247 (range, 2,715‐17,033 patient‐days per year). The Chicago Area Neonatal MRSA Working Group (CANMWG), which first convened in 2002, initiated a forum in which experiences, research, and data pertaining to outbreaks of MRSA infection in NICUs could be discussed. As part of this process, standardized data collection instruments were designed and distributed to participating hospitals for reporting of clusters of MRSA infection and for reporting of infection control responses. The CDPH and the EDHHS collected demographic, clinical, microbiologic, and epidemiological information about individual cases and clusters of MRSA infection. In addition, participating hospitals submitted MRSA isolates to the Illinois Department of Public Health Division of Laboratories for genotyping, when isolates were available. Pulsed‐field gel electrophoresis (PFGE) was performed using the restriction enzyme Sma1 according to methods described elsewhere.⁵

An invasive case of MRSA infection was defined as isolation of MRSA from a normally sterile site (blood, spinal fluid, abscess, or soft tissue) associated with a change in patient’s clinical status, such as ventilator‐setting changes, blood pressure changes, and skin and soft‐tissue changes. A noninvasive case of MRSA infection was defined as isolation of MRSA from a nonsterile site (such as sputum, the nares, the umbilicus, the rectum, or a medical device). A cluster was defined as detection of 2 or more patients with a culture positive for identical types of MRSA (defined as isolates with identical antibiograms or indistinguishable PFGE patterns) within a 14‐day period in a single NICU. For each cluster of MRSA infection, hospital staff completed a standardized report form that described infection control strategies implemented at the time of recognition of the cluster. Data were analyzed in Epi Info, version 6.04b (Centers for Disease Control and Prevention). Statistically significant differences between the characteristics of invasive and noninvasive cases were determined using either the Mantel‐Haenszel χ² statistic or the Fisher exact test, for dichotomous variables. The Mann‐Whitney test was used for analyses involving continuous variables.

Results

Overview of Clusters and Infection Control Response

Thirteen clusters of MRSA infection in NICUs were reported between June 2001 and September 2002. Overall, 149 MRSA‐positive infants were reported. Thirty‐one infants (21%) developed invasive infections (bacteremia in 24 and abscess in 7). Six (19%) of the 31 infected infants died. Cases were identified during all but 1 month throughout the period under study, and approximately half of the clusters began during the months from November 2001 through March 2002 (Figure). Of the 8 clusters for which census information was reported, 7 of the affected NICUs were below bed capacity and 1 NICU was at bed capacity at the time the first positive culture result was obtained. Molecular typing of isolates from the 8 hospitals demonstrated that they had 34 distinct PFGE patterns. Six hospitals had MRSA isolates with PFGE patterns that matched those from another hospital, and 6 PFGE patterns were shared by isolates from more than 1 hospital.

(53 KB)

Figure  Epidemic curve of 149 cases of methicillin‐resistant Staphylococcus aureus (MRSA) infection in Chicago‐area level III neonatal intensive care units, by the month that MRSA was first isolated, June 2001 through September 2002 (months are indicated with numbers). Black bars denote invasive cases, grey bars denote noninvasive cases, and each asterisk denotes a hospital‐based cluster that began during the month indicated.

Open New Window

The CDPH and the EDHHS compared infection control practices in 7 hospitals that reported information pertaining to 10 clusters of MRSA infection in their NICUs during the study period. MRSA‐positive infants were cohorted in all 10 clusters. Healthcare workers (HCWs) were cohorted in 8 of 10 clusters. In 2 clusters, nurses and physicians were cohorted; in 1 cluster, nurses and respiratory therapists were cohorted; and in 4 clusters, only nurses were cohorted. Cohorting of HCWs did occur in the remaining cluster, but HCW types were not specified. In response to the clusters of infection, hospitals performed surveillance cultures of infant specimens weekly for 7 clusters and as needed for 3 clusters. Specimens from neonates that were used for surveillance culture included nares specimens (in 5 clusters), combined swab specimens from the nares, throat, and umbilicus (in 2), nares and rectum specimens (in 1), nares, rectum, and umbilicus specimens (in 1), and combined swab specimens from the throat and umbilicus (in 1). Specimens for surveillance culture were obtained from HCWs in 8 of 10 clusters. Nurses, respiratory therapists, and physicians provided specimens for culture in all 8 of the clusters, and radiology technicians and phlebotomists provided specimens for culture in 7 and 4 clusters, respectively. Twenty‐seven (2.4%) of 1,130 HCWs screened were MRSA positive. In 5 clusters, both infants and HCWs who were infected or colonized with MRSA were treated with nasal mupirocin. Environmental samples were cultured in 4 clusters, and MRSA was isolated from the environmental cultures in 2 clusters.

Risk Factors for Invasive Infection

Compared with infants who had noninvasive MRSA infection, those with invasive infection were more likely to have a lower birth weight ( ) and shorter duration between admission and the first culture positive for MRSA (P = .03) (Table 1). In addition, patients with invasive MRSA infection were significantly more likely to be parenterally fed: among patients for whom the means of nutrition was reported, 89% of those with invasive infection received parenteral nutrition (with or without enteral nutrition), compared with 40% of those with noninvasive infection (P < .01). Presence of a percutaneous central venous catheter ( ) and presence of an endotracheal tube ( ) were significantly associated with invasive disease.

Table 1.  Risk Factors for Invasive Infection With Methicillin‐Resistant Staphylococcus aureus Among 140 Patients in Neonatal Intensive Care Units for Whom Complete Epidemiologic Data Were Reported

Open New Window

Discussion

Eliminating MRSA transmission within the NICU is important because MRSA can be the cause of serious infections in the highly vulnerable population of the NICU. The outbreaks of MRSA infection described in the Chicago‐area NICUs exemplify the problems, issues, and potential solutions associated with this pathogen. Our results demonstrate that MRSA infection in neonates is an important problem and that localized infection control practices at individual hospitals are not sufficient to stem the tide of MRSA. Spread of MRSA within the NICU can be controlled, but only by enforcing and mandating infection control practices already known to be effective in preventing horizontal spread.

The epidemiology of this regional outbreak was complex, and it is likely that MRSA was transmitted both nosocomially and exogenously. These findings support the notion that the incidence of nosocomial MRSA infection is increasing among intensive care unit patients and the observation that emergence of community‐associated MRSA infection has been increasing in certain geographic areas.^11‐20 It is likely that other geographic regions may begin to experience increases in the incidence of neonatal MRSA infection. The need for standardized recommendations regarding management and control of MRSA infection and colonization in NICUs is increasing in importance. Our recommendations provide a sensible yet rigorous approach to infection control in the NICU that can be used to protect these neonates from serious disease due to MRSA. More studies—for example, to examine mupirocin use, to calculate the cost‐effectiveness of various infection control practices, to characterize MRSA strains associated with severe disease, and to develop rapid molecular tests that detect MRSA—will be important to help guide future recommendations, which will have to evolve as this problem intensifies.

Consensus Statements

Since there are no current published guidelines for the control of MRSA infection in neonates, the CANMWG developed management recommendations for the prevention and control of MRSA infection and colonization in the NICU; these are presented in Table 2. These recommendations reflect the collective experience of the group, as well as the findings of the pertinent literature.^1‐8^,10,21‐42 Individual recommendations are rated according to the designations developed by the Centers for Disease Control and Prevention and the Hospital Infection Control Practice Advisory Committee (HICPAC)³³ to categorize recommendations based on the existing scientific data, theoretical rationale, applicability, and economic impact; the rating categories are defined in the note to Table 2. The consensus statements reflect the opinions of the CANMWG and are intended to serve as guidelines for the prevention and control of MRSA infection and colonization in level III NICUs. For the purpose of these guidelines, a cluster of MRSA infection was defined as detection of 2 or more patients with a culture positive for identical genotypes of MRSA (defined as isolates with identical antibiograms or indistinguishable PFGE patterns) within a 14‐day period in a single NICU.

Table 2.  Consensus Statement of the Chicago‐Area Neonatal MRSA Working Group Regarding Control of Methicillin‐Resistant Staphylococcus aureus (MRSA) in Neonatal Intensive Care Units (NICUs)

Open New Window