Evaluation of the Biomic V3 Microbiology System for Identification of Selected Species on BBL CHROMagar Orientation Agar and CHROMagar MRSA Medium

Abstract

The Biomic V3 microbiology system identifies bacteria by reading the color of colonies selected by the user. For CHROMagar orientation, Biomic results agreed with conventional methods for 94% of the strains assayed. For CHROMagar MRSA, Biomic correctly identified 100% of the strains tested and did not misidentify two methicillin-susceptible Staphylococcus aureus strains growing on the plates.

Use of CHROMagar orientation agar (BBL; Becton Dickinson, Sparks, MD) for primary isolation enables species identification comparably faster than the rapid abbreviated methods described in CLSI M35 (abbreviated identification of bacteria and yeast) and 4 to 24 h earlier than do traditional methods (2, 6, 7, 9, 10, 12). It can also facilitate picking colonies for subsequent conventional identification. Chromogenic agars use proprietary substrates that react with enzymes of specific species to impart colors to colonies. CHROMagar orientation is designed to allow technologists to recognize primary urinary tract pathogens without further manipulations. Expected colony appearances are as follows: Escherichia coli, dark rose to pink; Enterococcus, small blue-green; Klebsiella, Enterobacter, and Serratia group, larger dark blue; Proteus, Providencia, and Morganella group, pale to beige; Staphylococcus saprophyticus, small light pink to rose; and Streptococcus agalactiae, pinpoint light blue.

CHROMagar MRSA is useful for screening for methicillin/oxacillin-resistant Staphylococcus aureus (MRSA) colonization (1, 3, 4, 8). MRSA appears as medium size mauve colonies. Other organisms are not expected to grow well on this medium. Earlier organism identification may allow faster initiation of decolonization treatment and shorten the “transmission opportunity” time of a colonized patient.

The present study was designed primarily to determine whether the Biomic V3 automated plate reading instrument could interpret the colony color of isolates on primary urine culture plates and thus identify colonies of species for which CHROMagar orientation has indications as accurately as could an experienced microbiologist. Second, the Biomic V3 instrument was evaluated for its ability to accurately interpret the colony color of a group of S. aureus subcultured from original cultures to accurately differentiate between MRSA and methicillin-susceptible S. aureus.

CHROMagar orientation plates were inoculated with routine urine specimens sent for culture, yielding 182 isolates from pure cultures and 64 isolates from mixed cultures that included multiple species of pathogens and contaminants (skin or genital microbes) (Table 1). Isolates considered to be contaminants were not included in the present study. A total of 246 organism identifications were visually determined on routine CHROMagar orientation plates and subsequently confirmed (or not) by traditional methods in the Stanford University Medical Center laboratory. The Biomic V3 microbiology system (Giles Scientific. Inc., Santa Barbara, CA) was used in parallel to identify semiautomatically, based on color likeness, the same 246 organisms on CHROMagar orientation plates.

TABLE 1.

CHROMagar orientation organism identification by Biomic V3 compared to experienced technologists using visual and conventional methods

Organism(s)a	No. of isolates from:		Agreement between methods (%)	Color likeness (%)b
	Pure cultures	Mixed cultures		Avg	Range
Escherichia coli	44	8	98	71	24-100
Enterococcus species	37	23	88	87	35-100
Klebsiella, Enterobacter, and Serratia group	45	13	95	87	34-100
Proteus, Morganella, and Providencia group	23	3	100	95	67-100
Staphylococcus saprophyticusc	6	1	86	97	80-100
Group B streptococci	4	4	100	86	42-100
Other (staphylococci, yeast, Pseudomonas aeruginosa, others)d	23	12	83d	NA	NA
Total isolates	182	64	94

^aA positive result from a pyrrolidonyl aminopeptidase test (performed with blood agar) is required to distinguish Enterococcus spp. from group B streptococci. Some color overlap is common between these two species, so laboratory protocols should include the performance of PYR when such morphologies are seen. Very young (small) colonies of the Enterobacter group could also be mistaken for Enterococcus due to similar color ranges, although additional incubation results in increased colony size for Enterobacter group strains but not for enterococci, which can be distinguished by a microbiologist.

^bColor likeness is the percentage of color that matches the established color profile for each species. A higher percentage indicates a more typical color on CHROMagar orientation plates. A lower percent color likeness indicates greater heterogeneity in color reactions of a genus or species (e.g., E. coli). NA, not applicable.

^cColony morphology, Gram staining, specimen source, and novobiocin 5-μg disk diffusion (resistance) are required to identify S. saprophyticus. Biomic misidentified one isolate as S. saprophyticus.

^dFour of thirty-five organisms had an incorrect color reading by the instrument, which led to incorrect (i.e., false-positive) identifications.

CHROMagar MRSA agar plates were inoculated with 105 previously identified S. aureus strains, primarily MRSA, isolated from routine cultures sent to our laboratory (Table 2). Whether plates are inoculated with nares swabs or from colony growth, only isolated colonies chosen by the technologist are evaluated in the Biomic system, so the use of subcultures from known strains does not negate the results.

TABLE 2.

CHROMagar MRSA identification by Biomic V3 compared to traditional identification methods

Organism	No. of isolates showing:		Agreement between Biomic V3 and conventional methods (%)	Color likeness (%)b
	Growth on CHROMagar	No growth on CHROMagar		Avg	Range
MRSA	90	0	100	92	29-100
MSSAa	2	13	100	0	aNA

^aTwo of fifteen MSSA isolates grew on CHROMagar MRSA media, but since they were not mauve and not typical MRSA colonies, they were not misidentified as MRSA.

^bSee Table 1, footnote b.

The user places a culture plate with visible colonies (usually after overnight incubation) onto the pull-out tray of the Biomic V3 and allows the instrument to show the image on the screen. The user manipulates the mouse to point to a colony on the plate image to obtain a probable species identification; the identification and plate image are then saved, and the next plate is inserted into the instrument. The Biomic V3 reports the probable species identification with a color likeness percentage that is relative to the ideal color profile established for each species. The color likeness percentage provides an objective quantitative measurement of colony color. A higher percent likeness indicates a more typical color for that organism on the specific CHROMagar plate being evaluated. The user chooses the CHROMagar plate type before obtaining a reading, since the instrument has been programmed to interpret colonies on several CHROMagar plate formulas.

The overall agreement for the Biomic between genus or species identification on CHROMagar orientation plates with traditional methods, including CLSI M35 rapid identification of bacteria and yeast, Vitek (Legacy; bioMérieux, Durham, NC) used for gram-negative rods, or MicroScan Walkaway (Dade MicroScan, Sacramento, CA) used for gram-positive cocci, was 94%. Gram-positive isolates in general had lower percent agreements with conventional identifications than did gram-negative strains. The color likeness of identified colonies ranged from 71 to 95%. Among 35 strains for which CHROMagar orientation has no specific identification claims, there were 4 (11.4%) for which Biomic assigned a color that led to an incorrect identification (Table 1). These strains demonstrated colors and morphologies that a technologist would have correctly interpreted as needing further tests for identification. The manufacturer plans to use these results to fine-tune the software. The 31 other strains were not given an identification by the Biomic instrument.

The overall agreement between identification on CHROMagar MRSA with routine methods, which included CLSI M35 or, for inconsistent results, the API Staph strip (bioMérieux), and for susceptibilities with the MicroScan Walkaway plus oxacillin agar screening was 100%. The average color likeness of identified MRSA colonies was 92% (Table 2). Two of fifteen MSSA organisms grew as clear colonies on CHROMagar MRSA but were not reported as MRSA. There were no false positives; that is, no MSSA isolates were reported as MRSA.

The slightly lower agreement between some species identifications on CHROMagar orientation and traditional identification methods suggests that selected species with lower color likeness as determined by the Biomic V3 may require additional identification methods beyond simple visual or automated color interpretation. However, because CHROMagar orientation was developed for use with urine cultures, it is expected that laboratories may be willing to sacrifice some accuracy of identification for the increased speed and efficiency gained by using that medium for routine urine specimens, which is consistent with good patient care practices, as determined with clinician consultation (2). A possible reason for lower agreement for S. saprophyticus could be that some isolates are overlooked on standard media, as suggested by Hengstler et al. (6).

The agreement between CHROMagar MRSA and traditional identification methods mirrors those of others and suggests that this medium is useful for screening primary specimens for the presence of MRSA (2-5, 8, 11). The Biomic instrument did not misidentify the two MSSA isolates (13% of the MSSA isolates tested) that yielded atypical colonies on the CHROMagar MRSA.

The Biomic V3 usually provided accurate color readings of colonies on CHROMagar orientation and MRSA media that were equivalent to those of an experienced technologist. The Biomic V3 system provides the advantage of reporting the objective quantitative percent likeness parameter and provides the potential to reduce technologist training, reduce reading variation between technologists, save time, and save plate images for later review.