Evaluation of screening media for detection of methicillin-resistant Staphylococcus aureus (MRSA) from surveillance swabs

Abstract

OBJECTIVES: Highly sensitive and cost-effective screening media that aid in a rapid turnaround time (TAT) are of prime importance in the control of MRSA. Good specificity is also paramount, as false positives significantly increase labor and material costs. METHODS: This study compares selective CHROMagar MRSA™ (BBL, Sparks, Maryland) with 6 ug/mL cefoxitin (Cfox) to commonly used Mannitol Salt Agar with 4 ug/mL Oxacillin (MSox). A total of 2561 MRSA surveillance specimens from 1313 patients (775 nasal swabs, 733 rectal, 266 wound, 682 nasal/axilla/groin/perineum, 105 other) were planted to Cfox and MSox. The order of inoculation was alternated every 200 swabs. Cfox were incubated in the dark at 35°C, and both media were read blinded at 24 and 48 hours. Mauve colonies from the Cfox and yellow from the MSox were tested for capsular antigens 5 and 8 (Pastorex Staph Plus, Sanofi Diagnostics Pasteur, Redmond, Washington), tube coagulase and PBP2a (Denka Seiken Co., Ltd., Tokyo) to identify MRSA directly from the medium when possible. Positive PBP2a reactions were confirmed using the NCCLS 6 ug/mL Oxacillin Screen Agar. RESULTS: From the 2561 swabs, 152 (5.9%) MRSA grew from at least one medium; 133 (87.5%) grew on Cfox and 104 (68.4%) grew on MSox, while 85 (55.9%) grew on both media. Of the total 152 MRSA, 19 (12.5%) failed to grow on Cfox and 48 (31.6%) failed on MSox. The sensitivities for Cfox and MSox were 88.9% and 76%, respectively. Cfox grew 17% of its 133 MRSA at 24 hours, and 83% at 48 hours, while MSox grew 25% of its 104 MRSA at 24 hours and 75% at 48 hours. From the 2407 MRSA-negative specimens, non-<i>S. aureus</i> mauve or yellow colonies (coagulase negatives, enterococci, or Gram-negative bacilli) required work-up on 46 from Cfox and 415 from MSox, resulting in specificities of 98.1% for Cfox and 85.2% for MSox, respectively. CONCLUSION: Compared to MSox, the Cfox plates were both more sensitive and specific (p < 0.0001). This resulted in fewer unnecessary work-ups on Cfox for a saving of ∼$2 per false positive. The impact of the breakthrough rate of 16% associated with MSox in this study represents an extremely significant potential for cost avoidance to laboratories processing large numbers of surveillance screens. However, the impact on overall cost to the laboratory depends on the cost differential between MSox and Cfox.