Microbiologic Surveillance Using Nasal Cultures Alone Is Sufficient for Detection of Methicillin-Resistant Staphylococcus aureus Isolates in Neonates

Abstract

During an outbreak of methicillin-resistant Staphylococcus aureus (MRSA) in the neonatal intensive care units at two hospitals, we assessed several sites for detection of MRSA colonization. Nasal cultures found 32 of 33 MRSA-colonized patients (97%). Rectal cultures detected 29% of 24 MRSA-colonized patients identified by paired rectal and nasal samples and axillary samples found 22% of 9 MRSA-colonized patients identified by axillary samples paired with nasal swabs. There were no positive umbilical samples.

Methicillin-resistant Staphylococcus aureus (MRSA) has emerged as a cause of epidemics in the neonatal intensive care unit (NICU) (4, 9, 13). High rates of colonization (30 to 70%) may occur before clinical infections indicate a problem (14, 17). However, there are few data on the appropriate body site(s) for microbiologic surveillance of MRSA among neonates. As a result, there are widely differing practices for the detection of MRSA that involve culturing multiple sites, including the anterior nares, umbilicus, pharynx, catheter sites, rectum, groin, and axilla, as well as any open wounds (1, 6, 18). During a recent outbreak of MRSA in the NICUs at two hospitals, we compared (i) nasal versus rectal cultures and (ii) nasal versus axillary plus (where available) umbilical cultures for the detection of MRSA carriage.

Northwestern Memorial Hospital is a 720-bed teaching hospital with a 46-bed NICU that admits approximately 700 infants each year. An outbreak of MRSA in this NICU was first recognized in July 2001 after it was initially recovered from an endotracheal tube specimen. Subsequently, weekly surveillance cultures for MRSA colonization were obtained from the anterior nares and rectum of all neonates with sterile rayon-tip swabs (Culturette, BBL, Sparks, Md.). Samples were plated directly onto mannitol salt agar and incubated at 35°C for 48 h. Mannitol-fermenting colonies were subcultured onto 5% sheep blood agar plates. S. aureus was identified by a latex agglutination test (Pastorex Staph Plus, Sanofi Diagnostic Pasteur, Marnes La Coquette, France). S. aureus isolates for which the oxacillin MIC was ≥4 μg/ml by the agar dilution technique were classified as MRSA (11).

Evanston Northwestern Healthcare is an 804-bed primary and tertiary referral center with a 44-bed NICU that admits approximately 520 infants each year. An outbreak of MRSA here was first recognized in October 2001 after MRSA was recovered from the endotracheal tube and blood of a neonate. Weekly surveillance cultures for the detection of MRSA were obtained from the anterior nares and rectum from all neonates in the NICU. In December 2001, following an interim comparison between paired nasal and rectal swabs, it was decided to discontinue rectal cultures. Continued comparisons were performed between paired nasal and axillary cultures plus (where present) umbilical stump cultures. All cultures were plated directly onto Columbia-colistin-nalidixic acid-5% sheep blood agar plates (Remel, Inc., Lenexa, Kans.) and incubated at 35°C for 48 h. Isolates that were catalase- and coagulase-positive and demonstrated growth on a 6-μg/ml oxacillin salt agar screening plate were identified as MRSA.

An analysis of the laboratory cost of our surveillance testing was performed, which included the actual cost of bacteriologic media, reagents, and laboratory technologist time. Recovery of MRSA from different culture sites was compared using a two-sided exact McNemar's test. A P value of ≤0.05 was considered significant. The Institutional Review Board of Evanston Northwestern Healthcare approved the analysis of this testing method comparison.

The outbreak in the two NICUs involved a total of 38 neonates. The mean gestational age of the colonized and infected infants was 30 weeks (range, 24 to 40 weeks), there was an equal number of boys and girls, and the average birth weight was 1,243 g (range, 555 to 3,356 g). Thirty-three of the 38 neonates were colonized with MRSA, and 5 had clinical infections.

Overall, 373 paired nasal and rectal cultures and 185 paired nasal and axillary cultures (53 of these also included umbilical cultures) were collected. Of these, 51% were from Northwestern Memorial Hospital and 49% from Evanston Northwestern Healthcare. Of 24 positive infants who had paired nasal and rectal cultures, both cultures were positive in 6 infants, only the nasal culture was positive in 17 infants, and only the rectal culture was positive in one (P < 0.01, nasal versus rectal). Of the 9 infants who had paired nasal and axillary cultures, both were positive in 2 infants and only the nasal culture was positive in 7 infants (P = 0.02, nasal versus axillary). There were no positive umbilical cultures. The most common sites of positive surveillance cultures for MRSA in the patients were the nares (97%), followed by the rectum (32%) and axilla (22%). A statistical analysis of the results is in Table 1.

TABLE 1.

Comparison of results from colonization sites in neonates with results from paired cultures collected from at least two areas

Sites of paired cultures	No. of neonates with positive results for:
	Nares	Rectum	Axilla	Umbilicus
Nares and rectum (n = 373)	23	7
Nares and axilla (n = 185)	9		2
Nares and umbilicus (n = 53)	9			0
Sensitivity (%)	95.8	29.2	22.2	0
Negative predictive value (%)	99.6	93.6	95.7	83.1

The strains of MRSA from the two hospitals were unrelated by pulse-field gel electrophoresis typing, as interpreted according to the criteria of Tenover and colleagues (15), and involved at least five distinct clones (data not shown). No infants were transferred between the two facilities during the study period.

The use of nasal surveillance cultures alone during both outbreaks would have resulted in an overall direct cost saving of U.S. $7,637.50 and would have detected 32 of 33 (97%) colonized infants.

Prevention and control of nosocomial infections in infants who start life in the NICU with immune systems compromised by prematurity, illness, and invasive procedures is an important challenge. Since MRSA is resistant to many antimicrobial agents, now including vancomycin via transfer of the vanA gene from Enterococcus faecalis in Michigan and Pennsylvania (2, 3), particular attention needs to be paid to prevent the spread of MRSA in the NICU. Furthermore, once established in an institution, MRSA outbreaks have proved difficult to eradicate (5).

Accurate identification of those infants who are colonized with MRSA is critical for managing the spread of the organism (7). A prior study on MRSA during an outbreak in the NICU reported that the addition of rectal cultures improved surveillance sensitivity (1). Of 28 infants with MRSA who had paired rectal and nasal cultures, both cultures were positive in 16 infants, with solely the nasal culture being positive in 3 infants and solely the rectal culture being positive in 9 infants. The difference was not statistically significant (P = 0.15), and no selective or differential culture media were employed.

Another study of NICU patients that compared the yield of weekly surveillance cultures of nares, groin, axilla, and wounds (if present) found that 88% of nares, 56% of umbilicus, 50% of groin, and 31% of axilla were colonized by MRSA (6). These results were more consistent with our findings.

The ecological niche of S. aureus is the anterior nares, from where the organism can spread to other parts of the body (8). Importantly, as nasal colonization with S. aureus appears to antedate bacteremic as well as nonbacteremic infection (16), detection of nasal carriage may be of great clinical relevance. Additionally, treatment of the anterior nares to eliminate nasal carriage also results in the disappearance of the organism from other areas in most cases (10, 12).

Karchmer and colleagues recently estimated that well over U.S. $10 in reduced health care cost is realized for every U.S. $1 spent on active surveillance cultures by comparing a MRSA outbreak in two NICUs (7), only one of which performed weekly active surveillance. However, limiting expenditures to what is necessary is always prudent. We found that the use of nasal surveillance cultures alone during both outbreaks would have resulted in an overall cost saving of U.S. $7,637.50 and detected 97% of MRSA-colonized infants.

Our findings indicate that nasal surveillance cultures alone are sufficiently sensitive and cost-effective for the detection of MRSA in neonates.