Interpretation of MRSASelect Screening Agar at 24 Hours of Incubation

An incubation time of 24 h at 35 to 38°C is recommendedfor the optimal performance of MRSASelect (Bio-Rad) chromogenicscreening agar. An additional 24 h is required to capture slow-growingmethicillin-resistant Staphylococcus aureus (MRSA). However,the normal hours of operation for most laboratories cannot reliablyaccommodate 24-h interpretation intervals. Daily agar plateinterpretations are more likely to occur around 18 h and 42h of incubation, which may compromise the performance of thechromogenic agar and negatively impact patient infection controlefforts. In order to validate the importance of stringent incubationtimes to plate performance, we evaluated MRSASelect medium atcontrolled intervals of 24 and 48 h of incubation, using clinicalMRSA-screening swabs. A total of 1,071 MRSA-positive and 2,733MRSA-negative cultures were selected for analysis. Comparedto 48-h-incubation results, the sensitivity and specificityof MRSASelect at 24 h were 98.3% and 98.2%, respectively. Only19 of 1,071 (1.8%) MRSA-positive isolates required 48 h fordetection. Holding 24-h negative plates an additional 24 h resultedin the workup of 253 (6.6%) pink, yet non-MRSA, colonies. The24-h positive and negative predictive values of MRSASelect,assuming MRSA prevalences of 1, 5, and 10%, were 35.5 and 99.98%,74.2 and 99.9%, and 85.9 and 99.8%, respectively. MRSASelectmedium held for 24-h incubation is a highly sensitive and specificMRSA-screening tool. Further incubation prolongs the turnaroundtime for results and creates a significant amount of unnecessarywork in the laboratory.

INTRODUCTION

Top

INTRODUCTION

The Society of Healthcare Epidemiology of America (SHEA) currentlyrecommends active surveillance of methicillin-resistant Staphylococcusaureus (MRSA) infections and contact precautions to controlits spread (7). In Canada, MRSA infection and colonization ratesin hospitals have increased more than 10-fold, from less than1% of total S. aureus isolates in 1995 to 10% in 2002 (2, 10).Much of this increase can be attributed to surveillance, butthe infection rate also increased fivefold in this same period.As a result, many laboratories have experienced substantialincreases in their work loads to meet the demands of activeMRSA surveillance programs.

While rapid and reliable screening measures are important forthe implementation of appropriate and effective infection controlmeasures, they are also important for the efficiency of thework flow in the clinical laboratory. Chromogenic media forthe rapid isolation and identification of MRSA from surveillancespecimens have recently been developed. MRSASelect (Bio-Rad,Marnes La Coquette, France) is a sensitive and specific chromogenicagar medium used for MRSA screening. Compared to those of otherscreening-culture media, including other chromogenic media,the sensitivity and specificity of MRSASelect range between97 to 99% and 90 to 99%, respectively (1, 6, 12). MRSA is presumptivelyidentified as a small pink colony, while white or pink-tingedcolonies are disregarded. According to the manufacturer's instructions,MRSASelect plates recover the majority of MRSA isolates following24-h incubations at 35 to 38°C. However, the length of theincubation presents a challenge for most clinical laboratories,because the majority of specimens are not received until middayor later, which results in incubation periods closer to 18 (day1) and 42 h (day 2). In this setting, a large number of MRSA-positivecultures may not be identified on day 1. However, previous evaluationsof MRSA-screening media have shown that their specificity isreduced after 24 h of incubation due to the breakthrough growthof coagulase-negative staphylococci, which can significantlyincrease the laboratory work load.

The objective of this study was to evaluate the ability of MRSASelectto detect MRSA from clinical surveillance specimens at 24 hof incubation compared to that at 48 h of incubation in an acute-carediagnostic laboratory setting.

MATERIALS AND METHODS

Top

MATERIALS AND METHODS

Over a 10-month period in a 750-bed tertiary care hospital withfour intensive care units, MRSA surveillance swabs were collectedfrom nares, perinea, and urine and/or wounds of adult and pediatrichospital patients who were at risk of carrying MRSA. Risk factorsincluded previous hospitalization, residence in a long-term-carefacility, and previous MRSA colonization or infection. For thefirst 5 months, each swab was inoculated directly onto a MRSASelectagar plate. During the latter 5 months, naris and perineum swabswere combined on the same plate, while other specimen typeswere inoculated individually. The plates were incubated at 37°Cin ambient air for 24 h and 48 h, plus or minus 1 h, and pinkcolonies were selected for further identification as per themanufacturer's instructions. MRSA and non-MRSA S. aureus strains(ATCC 43300 and ATCC 25923, respectively) were used for qualitycontrol of the media throughout the study. Confirmatory identificationof MRSA by using conventional bacteriology methods for all pinkcolonies isolated from the 48-h plate served as our referencestandard for comparative analyses. The identification of S.aureus was confirmed based on Gram stains; coagulase (slideand tube), catalase, and pyrrolidinyl-arylamidase reactions;and detection of the modified penicillin binding protein 2aby latex agglutination (Denka Seiken, Tokyo, Japan). Antimicrobialsusceptibility testing, including testing with cefoxitin, wasperformed in conformance with Clinical Laboratory StandardsInstitute recommendations (4). Specimen collection was designedto capture an approximate 3:1 ratio of true-negative to true-positivespecimens based on estimates of monthly MRSA surveillance swabvolumes and positivity rates. The sensitivity, specificity,and positive and negative predictive values (PPV and NPV) for24-h incubations were calculated in comparison to the 48-h results.

RESULTS

Top

RESULTS

A total of 1,071 confirmed MRSA-positive specimens (from 285patients) and 2,733 confirmed MRSA-negative specimens were selectedfor analysis. The results are summarized in Table 1. ATCC controlstrains verified the integrity of the media throughout the study.The sensitivity, specificity, PPV, and NPV of MRSASelect at24 h compared to the results at 48 h were 98.3, 98.2, 95.6 and98.3%, respectively. Only 19 (1.8%) of 1,071 MRSA-positive specimensrequired 48 h for isolation. Three specimens from three patientswould have been misidentified as MRSA negative at 24 h, as noother specimens were collected from these patients. The remaining16 specimens from 14 patients had 24-h MRSA-positive resultsfrom previous or concurrent specimens. At 24 h, there were 49chromogenic (pink) isolates selected that were ultimately identifiedas non-MRSA bacteria (false positives). Holding 24-h negativeplates an additional 24 h resulted in the identification of253 (6.6% of total isolates) false positives and a decreasein specificity from 98.2% at 24 h to 90.6% at 48 h. In the first4 months, false positives accounted for 10% of the total positiveisolates. Over subsequent months, the percentage of false positivesdecreased to 7%. The percentages of MRSASelect PPV and NPV (Table2), assuming an MRSA prevalence of 1, 5, and 10%, were 35.5and 100, 74.2 and 99.9, and 85.9 and 99.8, respectively.

View this table:
[in this window]
[in a new window]

TABLE 1. Results from MRSASelect at 24 and 48 h for 3,804 surveillance specimens

View this table:
[in this window]
[in a new window]

TABLE 2. PPV and NPV of MRSASelect at 24 h of incubation

DISCUSSION

Top

DISCUSSION

For the majority of clinical laboratories, routine "24-h" orovernight incubation for bacterial culture actually occurs for16 to 20 h. However, the product monograph for MRSASelect indicatesthat optimal performance requires 24 h of incubation at 35 to38°C (35°C is standard for most laboratories). To ourknowledge, this is the first study designed to evaluate theperformance of MRSASelect at controlled incubation intervals.At 24 h of incubation, 98.2% of MRSA isolates were identified.Prior to controlled 24-h incubation periods, our laboratoryincubated MRSASelect overnight ( $~$ 16 to 18 h) initially, followedby another 24 h of incubation. During this period, we recovered118 MRSA isolates from 5,242 specimens, with only 83 (70%) recoveredafter overnight incubation (data not shown), which indirectlyvalidates the importance of a complete 24-h incubation period.In previous studies, shorter incubation times of 16 to 20 hresulted in detection of 62 to 81% of MRSA isolates (3, 8, 13).Conversely, two other studies have reported $≥$ 95% MRSA detectionafter 18 to 24 h of incubation (1, 6). Louie et al. (6) recovered169 (95%) of 177 MRSA isolates at 18 to 24 h (from a total of6,199 specimens), and Ben Nsira et al. (1) recovered 99 (97%)of 102 MRSA isolates at 18 to 24 h (from a total of 699 specimens).The incubation intervals were not elaborated in either study.Stoakes et al. (12) reported that 98% (108 of 111) of the MRSAisolates were isolated at 18 h without further incubation.

At 24 h, only 49 specimens yielded chromogenic non-MRSA isolates(1.8% of the 2,733 negative specimens), underscoring the specificityof MRSASelect medium. However, reincubating plates for a further24 h had a minimal increase in yield; only 19 (1.8%) of 1,071MRSA isolates required 48 h for detection. It is not clear whythese specimens were not detected at 24 h; however, low bacterialconcentration, inhibitory ingredients in the media, or smallcolony variants have been implicated in other studies with delayedrecoveries from MRSASelect and other chromogenic media (5, 9,11, 12). The use of antimicrobials may also have contributedto the delayed recovery of these isolates (patient charts werenot reviewed).

Reincubation also created further unnecessary work; chromogenicisolates on 253 of 272 (93%) plates at 48 h of incubation provednot to be MRSA, accounting for 83.8% of the total false positivesevaluated. These false positives decreased the specificity from98.2 to 90.6% at 48 h, which mirrors the findings of similarstudies. Louie et al. (6) found a decrease in specificity from99.5 to 92%, detecting 28 false-positives by 24 h and another428 at 48 h, and van Loo et al. (13) found a decrease from 92%to 89%, detecting 4 false positives at 20 h and 12 more at 48h. While we did not further identify our false-positive isolates,false-positive pink colonies have been found to be coagulase-negativestaphylococci, corynebacteria, enterococci, or one of a smallnumber of Enterobacteriaceae, Acinetobacter, or Pseudomonasspecies. (1, 6, 13).

Chromogenic agar assays are subjective; they rely on the observerto decide if the colony is the "correct" color for further identificationof the organism in question. In the case of MRSASelect, coloniesthat are pink tinged or light pink are not considered suspiciousfor MRSA. However, when chromogenic media are initially implemented,it may require some time for observers to become familiar withcolor variations. Although the difference in the percentageof false positives in the first 4 months and that for the remainderof the study was only 3%, the trend each month was a decreasein the number of false positives selected, particularly at 24h, for further identification. This indicates a temporal improvementin the accuracy of observer interpretation.

The increasing prevalence and surveillance of MRSA infectionssubsequently impart significant constraints on laboratory resourcesand management. Incubation of MRSASelect plates for a full 24h is a highly sensitive and specific screening approach to MRSAdetection. Stringent adherence to this incubation time can minimizethe number of false-positive workups and decrease the time requiredto report a MRSA-negative result.

FOOTNOTES

* Corresponding author. Mailing address: Division of Medical Microbiology, Department of Laboratory Medicine and Pathology, University of Alberta Hospital Walter C. Mackenzie Centre 2B3.13, 8440 112 Street, Edmonton, Alberta T6G 2J2, Canada. Phone: (780) 407-2767. Fax: (780) 407-3864. E-mail: jeff.fuller{at}capitalhealth.ca

Published ahead of print on 14 January 2009.

REFERENCES

Top