|Home | About | Journals | Submit | Contact Us | Français|
Previous studies have validated the properties and documented the utility of chromogenic agar for surveillance of methicillin-resistant Staphylococcus aureus (MRSA). In this study, we used one of the chromagars, MRSASelect (Bio-Rad), as one of the primary isolation media for selected wound and respiratory clinical specimens which, in our institution, were typically polymicrobial. We examined a total of 638 specimens; 142 (22%) MRSA isolates were recovered. Twenty-six of these isolates were recovered only on the MRSASelect plate, representing a 28% (15/54) increase for endotracheal aspirates/sputa and a 15% increase for superficial wounds/ulcers (11/73) compared to the results with conventional culture. One isolate (1 CFU) was recovered by conventional medium alone. MRSASelect has generally been used for surveillance cultures; however, we document that an additional 21% of MRSA isolates would have gone unreported in these selected clinical specimens using only standard culture media. For 40% (6/15) of inpatients, MRSA isolated from the MRSASelect plate was the sole indicator of MRSA. Although these isolates can represent either colonization or infection, they are a potential reservoir of infection and nosocomial transmission. Our data support the focused use of chromogenic selective media for the increased detection of MRSA in polymicrobial wound and respiratory specimens, which could have an impact on both clinical treatment and infection control.
Methicillin-resistant Staphylococcus aureus (MRSA) strains were first reported in 1961, and since then, MRSA strains have spread across the world and are a significant source of hospital-acquired infections (5, 9). In 2007, the Veteran's Health Administration implemented a universal surveillance program for MRSA that began in the intensive care units and was quickly expanded to include all inpatient areas of the hospitals (Methicillin-resistant Staphylococcus aureus [MRSA] prevention initiative, 12 January 2007 [internal document]). The goal of this nationwide program is to prevent the transmission of MRSA from colonized and infected patients to other hospital patrons and staff.
Various methods are available for the detection of MRSA, including real-time PCR, latex agglutination, and phenotypic methods. Several different FDA-approved chromogenic agar formulations for the rapid detection of MRSA from nares specimens are available commercially, including Bio-Rad MRSASelect agar. Using chromogenic media, MRSA colonies can be quickly differentiated from methicillin-susceptible S. aureus (MSSA) and other bacteria based on colorometric phenotype (3, 6, 10, 15). MRSASelect agar has been validated for the isolation of MRSA from nares surveillance specimens but has not been FDA approved for specimens other than nares or used as a primary culture medium for the direct detection of MRSA from clinical specimens (3, 6). In this study, we initially validated MRSASelect agar as a primary culture medium for the rapid detection of MRSA from a large variety of clinical specimens. Based on the results from the initial phase of the study, we implemented the use of MRSASelect agar in our standard operating procedure for a selected group of specimen types that are typically polymicrobial, including endotracheal aspirates, sputa, superficial wounds, and ulcers.
A MRSASelect agar plate (Bio-Rad, Redmond, WA) was added as a primary isolation medium in addition to standard culture media for 333 consecutive samples from sterile and nonsterile sites (April to June 2008). Sample sites included wounds, fluids, aspirates, tissues, sputa, endotracheal aspirates, and other specimens (Table (Table1).1). Blood and urine specimens were excluded. The standard culture media included sheep blood agar (SBA), colistin-nalidixic acid blood agar (CNA), chocolate agar (CA), and MacConkey agar (MAC) (all from Remel, Lenexa, KS). SBA, CNA, and CA cultures were incubated for 18 to 24 h in CO2 at 35°C, and MAC and MRSASelect agar cultures were incubated for 18 to 24 h in ambient air at 35°C. On MRSASelect agar, differentiation of MRSA from other organisms was made based on the colorimetric phenotype according to the manufacturer's instructions. MRSA colonies appear mauve colored, while non-MRSA colonies appear white. Conventional methods, including the Gram stain and Staphaurex Staphylococcus aureus latex agglutination reagent (Remel, Lenexa, KS), were used to confirm the identification of S. aureus from both the MRSASelect agar and standard culture media. Methicillin resistance was determined by the Kirby-Bauer disk diffusion method with a cefoxitin disk or by the Vitek 2 ASTGP66 card (bioMeriuex, Durham, NC). To prevent bias, the MRSASelect agar plates were interpreted by the authors, while standard culture media were interpreted using routine personnel and laboratory protocols. The results of the first phase of the study suggested that there was no benefit in the use of a selective plate for sterile specimens, including deep wound specimens, tissues, and fluids. Thus, in the second phase of the study (January to May 2009), a MRSASelect agar plate was included in the standard operating procedure as a primary isolation medium in addition to standard culture media for 305 consecutive samples from a selected group of specimen types, which were typically polymicrobial, including endotracheal aspirates, sputa, superficial wounds, and ulcers. All plates were interpreted by laboratory technologists, and the MRSASelect plates were reviewed again by the authors. Standard and chromogenic culture methods were performed as previously described. Additionally, confirmation of mauve-colored colonies as MRSA was done using the Vitek 2 GP identification card and the ASTGP66 card (bioMerieux, Durham, NC). From specimens where S. aureus was isolated as the predominant organism on standard culture media and on the MRSASelect plate, the organism identification was reported as Staphylococcus aureus (MRSA) and the susceptibility result reported as oxacillin/cefoxitin resistant. From specimens where MRSA was isolated on the MRSASelect plate only, the organism identification was reported likewise, with an additional comment to indicate isolation in low quantities. All other organisms were reported according to laboratory protocol. Statistical analysis was performed using the Fisher's exact test.
In the initial phase of the study, a total of 333 consecutive clinical specimens from both inpatients and outpatients were evaluated by direct plating onto MRSASelect agar. Among the diverse group of specimens evaluated (Table (Table1),1), there was complete agreement for both methods for all specimens from sterile sites; however, in comparison with routine clinical culture, MRSASelect agar was more sensitive for the direct identification of MRSA from specimens that had heavy and complex microbial flora, such as sputa, endotracheal specimens, superficial wounds, and ulcers. MRSA was isolated from 54 (16%) specimens using any plate and on MRSASelect agar alone from 10/54 specimens. The use of MRSASelect agar resulted in a 19% overall increase in the detection of MRSA, with a 64% increase for endotracheal aspirates and sputum specimens and an 11% increase for superficial wound and ulcer specimens, when compared with the results of conventional culture. On conventional culture media, all of the discrepant specimens showed moderate to heavy mixed growth of other organisms, which can obscure S. aureus and resulted in its being overlooked by the technologists.
Based on the results from the initial phase of the study, the MRSASelect agar plate was implemented in the standard evaluation of only typically polymicrobial specimens, including endotracheal aspirates, sputa, superficial wounds, and ulcers. In this second phase of the study, a total of 305 consecutive clinical specimens from both inpatients and outpatients were evaluated by direct plating onto MRSASelect agar. MRSA was isolated from 88 (29%) specimens. Use of the MRSASelect plate resulted in an 18% overall increase in the detection of MRSA, with a 19% increase for endotracheal and sputum specimens and an 18% increase for superficial wound and ulcer specimens when compared with the results of conventional culture. One isolate was recovered on a standard culture medium alone (1 CFU on SBA only, 1/4 plates). A summary of the combined data for both study phases is found in Table Table22.
The patient data were divided into inpatient and outpatient groups. Outpatients are not captured by the VA MRSA surveillance program and do not routinely have nares surveillance cultures performed. Of the 228 inpatients in our study, only 29% (67/228) had more than one clinical culture performed. Of the 48 inpatients that had MRSA isolated from a clinical culture, 77% (37/48) had MRSA isolated from only 1 clinical culture workup. Table Table33 shows the number of inpatients with MRSA isolated from clinical culture and nares surveillance culture. Thirty-two inpatients had concordant MRSA-positive nares surveillance and clinical culture results. In addition, we wanted to determine whether the MRSASelect plate was useful in detecting MRSA that had not been detected by nares surveillance culture or by routine culture using standard culture media. Of the 15 inpatients' culture results where MRSA was isolated only from the MRSASelect plate, 40% (6/15) had a negative nares culture or a nares culture was not performed and the MRSASelect plate was the sole identifier of MRSA (Table (Table33).
The differences in the results of the two study phases were not statistically significant, with two exceptions. In the first phase, 7/11 (64%) MRSA isolates recovered from endotracheal aspirates and sputa were isolated only on the MRSASelect plate, and in the second phase, 8/43 (19%) were isolated only on the MRSASelect plate (P = 0.006). On review, we appeared to be equally diligent in plate observation for both study phases and suggest that this difference is due to the variation in the distribution of MRSA in the individual specimens. Compliance improved for the collection of nares surveillance cultures from inpatients in the second phase of the study, with an increase from 73% to 91% (P = 0.0003).
Overall, there was a 96% correlation between the results for MRSASelect agar and standard culture media. MRSASelect agar had a sensitivity of 99.3% and a specificity of 100%.
We studied the use of Bio-Rad MRSASelect agar as a selective and differential primary medium for the isolation of MRSA from a large variety of clinical, nonsurveillance specimens from a VA patient population, a practice that was not FDA approved at the time of our study. Several manufacturers' formulations of chromogenic media have been approved by the FDA for the identification of MRSA from nares specimens; however, none are intended to diagnose MRSA infection from nonsurveillance specimens or to guide treatment. Nearly all studies that have evaluated chromogenic media for the detection of MRSA have used surveillance specimens, predominantly from the nares but also including specimens from axilla, groin, rectum, perineum, throat, and wounds (1-3, 7, 8, 10-12, 14, 15), and several studies have found additional reservoirs of MRSA by comparing alternate anatomical sites (1, 2, 7, 11). Comparatively, there have only been two studies that have evaluated the use of chromogenic media for the isolation of MRSA from nonsurveillance clinical specimens (4, 13). A British study by Davies et al. compared two chromogenic media, MRSASelect agar (Bio-Rad) and CHROMagar MRSA (BioConnections), with four other, nonchromogenic formulations for MRSA detection (4). Overall, 192 swabs from wounds, nares, perineum/groin, eye, axilla, and throat were evaluated, with a positivity rate of 53.1% (62 swabs were taken from previously evaluated specimens known to be positive for MRSA) (4). Both chromogenic media had sensitivities of 96% by 48 h, although the extended incubation time did result in some false positives (4). The second study, by Pape et al., found BBL CHROMagar MRSA to be 100% sensitive and specific at 48 h for the isolation of MRSA from 124 positive blood cultures (13). For both of these studies, the sample type and number of specimens evaluated was limited in scope. Our study is the first large-scale use of a chromogenic medium for the recovery of MRSA in nonnares clinical specimens from a VA patient population in a hospital setting with an active surveillance program.
MRSASelect compared favorably to standard culture media for the identification of MRSA. A diverse set of specimen types was evaluated, for which MRSASelect achieved a specificity of 100% and had excellent performance in the inhibition of Gram-negative organisms, such as Pseudomonas species and Proteus species. In characteristically monomicrobial infections, such as those found associated with abscesses, tissue, and synovial fluid, MRSASelect agar provided no increase in sensitivity for the detection of MRSA; however, for polymicrobial specimens with heavy mixed growth, where S. aureus may be easily overlooked due to the predominance of other organisms, the selective medium resulted in a significant increase in the detection of MRSA (P = 0.003). From specimens where S. aureus was isolated as the predominant organism on standard culture media and on the MRSASelect plate, the organism identification could be immediately reported as Staphylococcus aureus (MRSA) with an oxacillin/cefoxitin-resistant susceptibility result, which is a one-day improvement in turnaround time for the identification of MRSA in our laboratory. From polymicrobial specimens where MRSA was isolated on the MRSASelect plate only, an additional comment was attached to indicate isolation of MRSA in low quantities. This comment allows clinicians to consider MRSA colonization, as well as infection. All other organisms were reported according to laboratory protocol. Identification of MRSA in any amount indicates a potential reservoir of infection and nosocomial transmission, and the addition of the MRSASelect plate can be done within the routine scope of laboratory work without adding an additional hospital-wide surveillance component. MRSA recovered from the MRSASelect plate alone may represent colonization and not necessarily clinical infection, and although detection of additional patients who carry or are infected with MRSA does not necessarily lead to a decrease in nosocomial acquisition, the detection of otherwise unknown reservoirs of MRSA contributes to an effective infection control strategy as outlined in the VA MRSA Initiative. Our results demonstrate that when incorporated into a routine culture protocol as a primary medium for nonsurveillance polymicrobial specimens, MRSASelect agar provides more sensitive detection of MRSA, which might affect both therapeutic decisions and infection control.
We thank the people of the microbiology Laboratory for their cooperation and excellent skills.
This material is the result of work supported in part by resources from the VA Puget Sound Health Care System, Seattle, WA.
Published ahead of print on 24 February 2010.