Evaluation of presumptive identification of Enterobacterales using CHROMagar Orientation medium and rapid biochemical tests

Abstract Background The use of matrix‐assisted laser desorption/ionization time‐of‐flight mass spectrometry is gradually spreading among large‐scale laboratories; however, this method is impractical for small‐scale laboratories. In laboratories without access to these rapid identification methods, problems related to them remain unsolved. In this study, we aimed to develop a rapid and inexpensive method to presumptively identify Enterobacterales using CHROMagar Orientation medium. Methods The algorithm for presumptive identification of Enterobacteriaceae using CHROMagar Orientation medium was based on our previous studies. Modified property tests for indole, lysine decarboxylase, ornithine decarboxylase, and hydrogen sulfide were performed to evaluate the differentiation of the bacterial species. Results Using the type strains and clinical isolates, it was possible to conduct the property tests at a low cost, within 4 hours. The spot indole test was performed without any nonspecific reactions for the bacteria forming colored colonies. The presumptive identification of bacteria was thereby possible within 24 hours after specimen submission. Conclusion All these results suggest that the rapid presumptive identification of Enterobacterales is possible with this new identification method using CHROMagar Orientation medium. This is therefore a prompt and economical method that can be used in routine laboratory work.


| INTRODUC TI ON
Rapid and accurate identification of bacteria is extremely important in clinical microbial laboratory testing. The use of matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) for rapid identification of bacteria has been gradually spreading in recent years and is becoming an increasingly available and widely used method. 1,2 In many clinical cases, the introduction of MALDI-TOF MS has enabled the reporting of identification results the day after culture. In addition to the promotion of antimicrobial stewardship, the significance of rapid reporting has increased. Facilities that cannot acquire MALDI-TOF MS equipment due to installation costs are still required to report identification results promptly. As an alternative, the use of simplified conventional tests to assess biochemical properties can provide these laboratories with rapid identification capabilities. The Clinical and Laboratory Standards Institute (CLSI) document M35-A2 and other guidelines are resources that support the rapid identification of bacteria. 3 In addition, the operation of routine work using selective medium is highly useful. 4,5 Our research group has previously proposed an algorithm for the presumptive identification of Enterobacterales, which are frequently isolated in the extraintestinal infection using the CHROMagar Orientation medium. 6 This is an inexpensive and useful method that was developed by combining the colony color and property tests. However, the considerably long time required to perform the property tests for metallic blue and brown colonies is a disadvantage. To solve this problem, it is necessary to quickly determine bacterial properties using biochemical methods including the indole, lysine decarboxylase (LDC), ornithine decarboxylase (ODC), and hydrogen sulfide (H 2 S) tests. Therefore, the aim of our study is to establish a rapid method using this algorithm.

| Algorithm for presumptive identification
The algorithm for presumptive identification of Enterobacterales using CHROMagar Orientation medium was based on previous studies (Table 1). Briefly, this method interrogates properties required for each color tone of the colonies growing on the medium and performs presumptive identification by combining their results. In this study, four property tests (indole, LDC, ODC, and H 2 S) were performed to distinguish each bacterial species. Since brown color colonies already indicate indole pyruvic acid (IPA) production, the property of IPA was deleted in this algorithm. We conducted the following study so that the above property tests could be performed quickly and inexpensively.

| Investigation of medium volume and incubation time in the LDC and ODC tests
The media for the LDC and ODC tests were prepared using Moeller decarboxylase broth base (Becton-Dickinson), and each medium was adjusted to a 2% concentration using l-lysine monohydrochloride (FUJIFILM Wako Pure Chemical Corporation) or l-ornithine monohydrochloride (FUJIFILM Wako Pure Chemical Corporation). The LDC and ODC test media were prepared at liquid volumes of 0.1 mL (lowvolume method) and 3 mL (conventional method). For the low-volume method, a 0.2 mL tube was used as a medium container. Bacteria adhered to the tip of an inoculating needle were inoculated into the medium (approximately 10 6 CFU), and the results were determined at 2, 4, 6, 8, and 24 hours after incubation (35°C, aerobic condition).

| Investigation of medium volume and incubation time in the hydrogen sulfide production test
The SIM medium for the H 2 S production test was prepared according to past literature. 7 The composition of the medium was as fol-

| Evaluation of the accuracy of new identification methods using clinical isolates
Based on the algorithm (Table 1) Table 1).
The LDC test medium and ODC test medium were aliquoted at volumes of 0.1 mL. As a medium for the H 2 S production test, 0.1 mL of agar-free medium was prepared. For the LDC, ODC, and H 2 S tests, bacteria adhered to the tip of a platinum needle were inoculated into the medium, and the results were determined at 4 hours after incubation (35°C, aerobic condition). In the spot indole test, reagents whose colony color did not affect the test results were selected.
Therefore, Kovac's reagent was used for the metallic blue colonies and p-dimethylaminocinnamaldehyde (DMACA) reagent was used for the brown colonies.

| Effect of medium volume and incubation time on the LDC, ODC, and H 2 S tests
The LDC and ODC test results for three bacterial species of the ATCC strain could be determined 4 hours after inoculation using the low-volume method even if they could not be determined by the conventional method ( Figure S1). Four hours was chosen as the test time point because after 24 hours incubation, a positive reaction was also observed in the negative control of the low-volume method and at 2 hours the positive reaction after incubation in the low-volume method was too weak.
In the H 2 S-producing strains P. mirabilis and P. vulgaris, blackening was observed 4 hours after inoculation in 0.1 mL of SIM medium without agar but not in the other media at the same incubation time ( Figure S2). A positive reaction at 2 hours after culture was not observed in any sample (data not shown). The H 2 S-non-producing strain M. morganii showed negative reactions in all media. Therefore, it was decided that evaluation of H 2 S production can be judged in 0.1 mL of agar-free SIM medium at 4 hours.

| Evaluation of the accuracy of new identification methods using clinical isolates
The new identification method, which is a combination of the im-  (Table 2). This method showed misidentification at a rate of 2.3% (n = 11) in the bacterial species included in the algorithm. Misidentification was caused by the lack of typical characteristics observed in these strains (Table 3). We confirmed that there is no difference between the results of the new method and the conventional method in these strains (data not shown).
For the spot indole test, to avoid confusion between colony color and color tone indicating a positive result, the blue colony test was performed with Kovac's reagent and the brown colony test was performed with DMACA reagent. Nonspecific reactions were not observed for any strain, and it was therefore not included as a cause of misidentification ( Figure S3; Table 3). The flowchart of this method based on these results and previous findings is shown in Figure 1. reported until the third day, which negatively impacts reporting speed. In this study, several investigations were carried out with the objective of shortening the identification reporting time. Our presumptive identification method using CHROMagar Orientation medium is a useful technique that can be easily performed at a low cost. It is useful in the presumptive identification of Enterobacterales, which are frequently isolated in extraintestinal infections, and is well adapted for the examination of specimens such as urine, sputum, pus, ascitic fluid, and blood. In addition, this algorithm has sufficient accuracy to be used as a presumptive identification method. 6 The use of the LDC and ODC tests allowed for correct judgment at 4 hours in ATCC strains; it is necessary to keep this 4-8 hours evaluation time point, as false-positive reactions are observed after 24 hours. In the study of H 2 S productivity using ATCC strains, H 2 S production was visible at 4 hours in 0.1 mL of agar-free SIM medium.

| D ISCUSS I ON
In all these tests, it was possible to perform evaluations without issues, even in the study using clinical isolates. In the study using clinical isolates, 11 of the strains included in the algorithm could not be identified. These were strains that do not exhibit typical properties, highlighting a limit of this method. In particular, it should be noted that there are some S. marcescens that form metallic blue colonies.
With the spot indole test, which is generally not recommended for to 20 years ago 9 and that of carbapenemase-producing bacteria is also increasing, 10,11 there is a significant need for the implementation of this method as a screening test. Due to the importance of antimicrobial stewardship in recent years, 12
CLSI M35-A2 is often cited as a manual for simplified identification methods, containing descriptions of methods for the identification of 37 bacterial species, most of which can be performed at low cost and in a short time frame. These methods are considered to be useful not only for convenience but also as alternative methods in unexpected cases such as disasters and machine failures. In order for this new method to be used as commonly as those described in this manual, we intend to continue our studies and provide more information in the future.

CO N FLI C T O F I NTE R E S T
The authors declare that they do not have any competing financial interests.