Inoculum Preparation.

Properly prepared inocula are the key to any antimicrobial susceptibility testing method. Inconsistencies in inoculum preparation may lead to inconsistencies and inaccuracies in susceptibility test results. The two important requirements for correct inoculum preparation are use of a pure culture and use of a standard-sized inoculum.

Interpretation of results obtained with a mixed culture is not reliable and can substantially delay reporting of results. Pure inocula are obtained by selecting four or five colonies of the same morphology, inoculating them into a broth medium, and allowing the culture to achieve active growth (i.e., midlogarithmic phase), as indicated by observable turbidity in the broth. For most organisms this requires 3 to 5 hours of incubation. Alternatively, four to five colonies 16 to 24 hours of age may be selected from an agar plate and suspended in broth or 0.9% saline solution to achieve a turbid suspension.

Use of a standard inoculum size is as important as culture purity and is accomplished by comparing the turbidity of the organism suspension with a turbidity standard. McFarland turbidity standards, prepared by mixing 1% sulfuric acid and 1.175% barium chloride to obtain a solution with a specific optical density, are commonly used. The 0.5 McFarland standard, which is commercially available, provides an optical density comparable to the density of a bacterial suspension of 1.5 × 10⁸ colony forming units (CFU) per milliliter. Pure cultures are grown or are prepared directly from agar plates to match the turbidity of the 0.5 McFarland standard (Figure 12-1). The newly inoculated bacterial suspension and the McFarland standard are compared by examining turbidity against a dark background. Alternatively, any one of various commercially available instruments capable of measuring turbidity may be used to standardize the inoculum. If the bacterial suspension does not match the standard’s turbidity, the suspension may be further diluted or supplemented with more organisms as needed.

Selection of Antimicrobial Agents for Testing.

The antimicrobial agents chosen for testing against a particular bacterial isolate are referred to as the antimicrobial battery or panel. A laboratory may use different testing batteries, but the content and application of each battery are based on specific criteria. Although the criteria listed in Box 12-1 influence the selection of the panel’s content, the final decision should not be made by the laboratory independently; input from the medical staff (particularly infectious diseases specialists) and the pharmacy is imperative.

CLSI publishes up-to-date tables listing potential antimicrobial agents recommended for inclusion in batteries for testing against specific organisms or organism groups. Two tables are of particular interest: Table 1, “Suggested Groupings of U.S. FDA–Approved Antimicrobial Agents That Should Be Considered for Routine Testing and Reporting on Nonfastidious Organisms by Clinical Microbiology Laboratories,” and Table 1A, “Suggested Groupings of U.S. FDA–Approved Antimicrobial Agents That Should Be Considered for Routine Testing and Reporting on Fastidious Organisms by Clinical Microbiology Laboratories.” Because revisions are made annually, laboratory protocols should be reviewed and modified accordingly (see the Bibliography). Further considerations about antibiotics that may be used for a specific organism or group are presented later in this chapter and in various chapters in Part III of this text.

Testing profiles are considered for each of the common organism groupings:

Procedures.

The key features of broth dilution testing procedures are shown in Table 12-1. Because changes are made in these procedural recommendations, the CLSI M07 series, “Methods for Dilution Antimicrobial Susceptibility Tests for Bacteria that Grow Aerobically,” should be consulted annually.

Medium and Antimicrobial Agents.

With in vitro susceptibility testing methods, certain conditions must be altered when examining fastidious organisms to optimize growth and facilitate expression of bacterial resistance. For example, the Mueller-Hinton preparation is the standard medium used for most broth dilution testing, and conditions in the medium (e.g., pH, cation concentration, thymidine content) are well controlled by commercial manufacturers. However, media supplements or different media are required to obtain good growth and reliable susceptibility profiles for bacteria such as S. pneumoniae and H. influenzae. Although staphylococci are not considered fastidious organisms, media supplemented with sodium chloride (NaCl) enhance the expression and detection of methicillin-resistant isolates (see Table 12-1).

Broth dilution testing is divided into two general categories: microdilution and macrodilution. The principle of each test is the same; the only difference is the volume of broth in which the test is performed. For microdilution testing, the total broth volume is 0.05 to 0.1 mL; for macrodilution testing, the broth volumes are usually 1 mL or greater. Because most susceptibility test batteries require testing of several antibiotics at several different concentrations, the smaller volume used in microdilution allows this to be conveniently accomplished in a single microtiter tray (Figure 12-2).

The need for multiple large test tubes in the macrodilution method makes that technique substantially cumbersome and labor intensive when several bacterial isolates are tested simultaneously. For this reason, macrodilution is rarely used in most clinical laboratories, and subsequent comments about broth dilution focuses on the microdilution approach.

A key component of broth testing is proper preparation and dilution of the antimicrobial agents incorporated into the broth medium. Most laboratories that perform broth microdilution use commercially supplied microdilution panels in which the broth is already supplemented with appropriate antimicrobial concentrations. Therefore, antimicrobial preparation and dilution are not commonly carried out in most clinical laboratories (the details of this procedure are outlined in the CLSI M07-A6 document). In most instances, each antimicrobial agent is included in the microtiter trays as a series of doubling twofold dilutions. To ensure against loss of antibiotic potency, the antibiotic microdilution panels are stored at −20°C or lower, if possible, and are thawed immediately before use. Once thawed the panels should never be refrozen, which may result in substantial loss of antimicrobial action and potency. Alternatively, the antimicrobial agents may be lyophilized or freeze dried with the medium or drug in each well; upon inoculation with the bacterial suspension, the medium and drug are simultaneously reconstituted to the appropriate concentration.

Reading and Interpretation of Results.

After incubation, the microdilution trays are examined for bacterial growth. Each tray should include a growth control that does not contain antimicrobial agent and a sterility control that was not inoculated. Once growth in the growth control and no growth in the sterility control wells have been confirmed, the growth profiles for each antimicrobial dilution can be established and the MIC determined. The detection of growth in microdilution wells is often augmented through the use of light boxes and reflecting mirrors. When a panel is placed in these devices, bacterial growth, manifested as light to heavy turbidity or a button of growth on the well bottom, is more reliably visualized (Figure 12-3).

When the dilution series for each antibiotic is inspected, the microdilution well containing the lowest drug concentration that completely inhibits visible bacterial growth is recorded as the MIC. Once the MICs for the antimicrobials in the test battery for an organism have been recorded, they are usually translated into one of the interpretive categories, specifically susceptible, intermediate, or resistant (Box 12-2). The interpretive criteria for these categories are based on extensive studies that correlate the MIC with serum-achievable levels for each antimicrobial agent, particular resistance mechanisms, and successful therapeutic outcomes. The interpretive criteria for an array of antimicrobial agents are published in the CLSI M07 series document, “Methods for Dilution Antimicrobial Susceptibility Tests for Bacteria that Grow Aerobically (M100 supplements).” For example, using these standards, an isolate of P. aeruginosa with an imipenem MIC of less than or equal to 4 µg/mL would be classified as susceptible; one with an MIC of 8 µg/mL would be classified as intermediate; and one with an MIC of 16 µg/mL or greater would be classified as resistant to imipenem.

Box 12-2   Definitions of Susceptibility Testing Interpretive Categories*

Susceptible

Indicates that the antimicrobial agent in question may be an appropriate choice for treating the infection caused by the organism. Bacterial resistance is absent or at a clinically insignificant level.

Intermediate

Indicates a number of possibilities, including:

• The potential utility of the antimicrobial agent in body sites where it may be concentrated (e.g., the urinary tract) or if high concentrations of the drug are used

• Possible effectiveness of the antimicrobial agent against the isolate, but possibly less so than against a susceptible isolate.

• Use as an interpretive safety margin to prevent relatively small changes in test results from leading to major swings in interpretive category (e.g., resistant to susceptible or vice versa)

Resistant

Indicates that the antimicrobial agent in question may not be an appropriate choice for treatment, either because the organism is not inhibited with serum-achievable levels of the drug or because the test result highly correlates with a resistance mechanism that indicates questionable successful treatment.

---

^*Although these definitions are adapted from CLSI guideline M7-A3, Methods for Dilution Antimicrobial Susceptibility Tests for Bacteria that Grow Aerobically, they are commonly applied to results obtained by various susceptibility testing methods.

After the MICs are determined and their respective and appropriate interpretive categories assigned, the laboratory may report the MIC, the category, or both. Because the MIC alone will not provide most physicians with a meaningful interpretation of data, either the category result with or without the MIC is usually reported.

In some settings, the full range of antimicrobial dilutions is not used; only the concentrations that separate the categories of susceptible, intermediate, and resistant are used. The specific concentrations that separate or define the different categories are known as breakpoints, and panels that only contain these antimicrobial concentrations are referred to as breakpoint panels. In this case, only category results are produced; precise MICs are not available, because the full range of dilutions is not tested.

Procedures.

The key features of disk diffusion testing procedures are summarized in Table 12-3, with more details and updates available through CLSI.

Reading and Interpretation of Results.

Before results with individual antimicrobial agent disks are read, the plate is examined to confirm that a confluent lawn of growth has been obtained (see Figure 12-5). If growth between inhibitory zones around each disk is poor and nonconfluent, the test should not be interpreted and should be repeated. The lack of confluent growth may be due to insufficient inoculum. Alternatively, a particular isolate may have undergone mutation, and growth factors supplied by the standard medium are no longer sufficient to support robust growth. In the latter case, medium supplemented with blood and/or incubation in CO₂ may enhance growth. However, caution in interpreting results is required when extraordinary measures are used to obtain good growth and the standard medium recommended for testing a particular type of organism is not used. Plates should also be examined for purity. Mixed cultures are evident through the appearance of different colony morphologies scattered throughout the lawn of bacteria (Figure 12-7). Mixed cultures require purification and repeat testing.

A dark background and reflected light are used to examine a disk diffusion plate (Figure 12-8). The plate is situated so that a ruler or caliper can be used to measure the inhibition zone diameters for each antimicrobial agent. Certain motile organisms, such as Proteus spp., may swarm over the surface of the plate and complicate clear interpretation of the zone boundaries. In these cases, the swarming haze is ignored and zones are measured at the point where growth is obviously inhibited. Similarly, hazes of bacterial growth may be observed when testing sulfonamides and trimethoprim as a result of the organism population going through several doubling generations before inhibition; the resulting haze of growth should be ignored for disk interpretation with these agents.

In instances not involving swarming organisms or the testing of sulfonamides and trimethoprim, hazes of growth that occur in more obvious inhibition zones should not be ignored. In many instances, this is the only way clinically relevant resistance patterns are manifested by certain bacterial isolates when tested using the disk diffusion method. Key examples in which this may occur include cephalosporin resistance among several species of Enterobacteriaceae, methicillin resistance in staphylococci, and vancomycin resistance in some enterococci. In fact, the haze produced by some staphylococci and enterococci can best be detected using transmitted rather than reflected light. In these cases, the disk diffusion plates are held in front of the light source when methicillin and vancomycin inhibition zones are read (see Figure 12-8). Still other significant resistances may be subtly evident and appear as individual colonies in an obvious zone of inhibition (Figure 12-9). When such colonies are seen, purity of the test isolate must be confirmed. If purity is confirmed, the individual colonies are variants or resistant mutants of the same species, and the test isolate should be considered resistant.

Once zone sizes have been recorded, interpretive categories are assigned. Interpretive criteria for antimicrobial agent/organism combinations that may be tested by disk diffusion are provided in the CLSI-M2 series, “Performance Standards for Antimicrobial Disk Susceptibility Tests (M100 supplements).” The definitions of susceptible, intermediate, and resistant are the same as those used for dilution methods (see Box 12-2). For example, using the CLSI interpretive standards, an E. coli isolate that produces an ampicillin inhibition zone diameter of 13 mm or less is classified as resistant; if the zone is 14 to 16 mm, the isolate is considered intermediate to ampicillin; if the zone is 17 mm or greater, the organism is categorized as susceptible.

Unlike MICs, inhibition zone sizes are used to produce a category interpretation and have no clinical utility. Therefore, when testing is performed by disk diffusion, only the category interpretation of susceptible, intermediate, or resistant is reported.

Broth Microdilution Methods.

Several systems have been developed that provide microdilution panels already prepared and formatted according to the guidelines for conventional broth microdilution methods (e.g., BBL Sceptor, BD Microbiology Systems, Cockeysville, Maryland; Sensititre, Trek Diagnostics Systems, Inc., Westlake, Ohio; MicroScan touch SCAN-SR, Dade Behring, Inc., West Sacramento, California). These systems enable laboratories to perform broth microdilution without having to prepare their own panels.

The systems may differ to some extent regarding the volume in the test wells, how inocula are prepared and added, the availability of different supplements for the testing of fastidious bacteria, the types of antimicrobial agents and dilution schemes, and the format of medium and antimicrobial agents (e.g., dry-lyophilized or frozen). Furthermore, the degree of automation for inoculation of the panels and the devices available for reading results vary among the different products. In general, these commercial panels are designed to receive the standard inoculum and are incubated using conditions and durations recommended for conventional broth microdilution. They are growth-based systems that require overnight incubation, and CLSI interpretive criteria apply for interpretation of most results. Reading of these panels is frequently augmented by the availability of semiautomated reading devices.

Agar Dilution Derivations.

One commercial system (Spiral Biotech Inc., Bethesda, Maryland) uses an instrument to apply antimicrobial agent to the surface of an already prepared agar plate in a concentric spiral fashion. Starting in the center of the plate, the instrument deposits the highest concentration of antibiotic and from that point drug application proceeds to the periphery of the plate. Diffusion of the drug in the agar establishes a concentration gradient from high (center of plate) to low (periphery of plate). Starting at the periphery of the plate, bacterial inocula are applied as a single streak perpendicular to the established gradient in a spoke-wheel fashion. After incubation, the distance is measured between the point where growth is noted at the edge of the plate to the point where growth is inhibited toward the center of the plate (Figure 12-10). This value is used to calculate the MIC for the antimicrobial agent against each of the bacterial isolates on the plate.

Diffusion in Agar Derivations.

One test has been developed that combines the convenience of disk diffusion with the ability to generate MIC data. The Etest (bioMérieux, Durham, North Carolina) uses plastic strips; one side of the strip contains the antimicrobial agent concentration gradient, and the other contains a numeric scale that indicates the drug concentration (Figure 12-11). Mueller-Hinton plates are inoculated as for disk diffusion, and the strips are placed on the inoculum lawn. Several strips may be placed radially on the same plate so that multiple antimicrobials may be tested against a single isolate. After overnight incubation, the plate is examined and the number present at the point where the border of growth inhibition intersects the E-strip is taken as the MIC (Figure 12-11). The same MIC interpretive criteria used for dilution methods, as provided in CLSI guidelines, are used with the Etest value to assign an interpretive category of susceptible, intermediate, or resistant. This method provides a means of producing MIC data in situations in which the level of resistance can be clinically relevant (e.g., penicillin or cephalosporins against S. pneumoniae).

Another method (BIOMIC, Giles Scientific, Inc., New York, New York) combines the use of conventional disk diffusion methodology with video digital analysis to automate interpretation of inhibition zone sizes. Automated zone readings and interpretations are combined with computer software to produce MIC values and to allow for data manipulations and evaluations for detecting unusual resistance profiles and producing antibiogram reports.

Automated Antimicrobial Susceptibility Test Systems.

The automated antimicrobial susceptibility test systems available for use in the United States include the Vitek Legacy and Vitek 2 systems (bioMérieux, Inc., Durham, North Carolina), the MicroScan WalkAway system (Dade International, Sacramento, California), and the Phoenix system (BD Microbiology Systems, Cockeysville, Maryland). These different systems vary with respect to the extent of automation of inoculum preparation and inoculation, the methods used to detect growth, and the algorithms used to interpret and assign MIC values and categorical findings (i.e., susceptible, intermediate, resistant).

For example, the Vitek 2 AST inoculum is automatically introduced by a filling tube into a miniaturized, plastic, 64-well, closed card containing specified concentrations of antibiotics (Figure 12-12). Cards are incubated in a temperature-controlled compartment. Optical readings are performed every 15 minutes to measure the amount of light transmitted through each well, including a growth control well. Algorithmic analysis of the growth kinetics in each well is performed by the system’s software to derive the MIC data. The MIC results are validated with the Advanced Expert System (AES) software, a category interpretation is assigned, and the organism’s antimicrobial resistance patterns are reported. Resistance detection is enhanced with the sophisticated AES software, which can recognize and report resistance patterns using MICs. In summary, this system facilitates standardized susceptibility testing in a closed environment with validated results and recognition of an organism’s antimicrobial resistance mechanism in 6 to 8 hours for most clinically relevant bacteria (Figure 12-13).

The MicroScan WalkAway system uses the broth microdilution panel format manually inoculated with a multiprong device. Inoculated panels are placed in an incubator-reader unit, where they are incubated for the required time and then the growth patterns are automatically read and interpreted. Depending on the microdilution tray used, bacterial growth may be detected using spectrophotometry or fluorometry (Figure 12-14).

Spectrophotometric analyzed panels require overnight incubation, and the growth patterns may be read manually as described for routine microdilution testing. Fluorometric analysis is based on the degradation of fluorogenic substrates by viable bacteria. The fluorogenic approach can provide susceptibility results in 3.5 to 5.5 hours. Either full dilution schemes or breakpoint panels are available. In addition to speed and facilitation of workflow, the automated systems provide increasingly powerful computer-based data management that can be used to evaluate the accuracy of results, manage larger databases, and interface with the pharmacy to improve and advance the utility of antimicrobial susceptibility testing data.

The Phoenix system provides a convenient, albeit manual, gravity-based inoculation process. Growth is monitored in an automated fashion based on a redox indicator system with results available in 8 to 12 hours. Supplemental testing (e.g., confirmatory extended spectrum beta-lactamase [ESBL] test for E. coli) is included in each panel, reducing the need for additional or repeat testing. Interpretation of results is augmented by a rules-based data management expert system.

Supplemental Testing Methods.

Table 12-4 highlights some of the features of supplemental tests that may be used to enhance resistance detection. For certain strains of staphylococci, conventional and commercial systems may have difficulty detecting resistance to oxacillin and the related drugs methicillin and nafcillin. The oxacillin agar screen provides a backup test that may be used when other methods provide equivocal or uncertain profiles. Growth on the screen correlates highly with the presence of oxacillin (or methicillin) resistance, and no growth is strong evidence that an isolate is susceptible. This is an important determination; strains that are classified as resistant are considered resistant to all other currently available beta-lactam antibiotics, indicating the need for therapy to include the use of vancomycin. The agar screen plates can be made in-house, and are available commercially (e.g., Remel, Lenexa, Kansas; BBL, Cockeysville, Maryland). Additionally, other commercial tests designed to detect oxacillin resistance more rapidly (i.e., 4 hours) have been developed and may provide another approach to supplemental testing (e.g., Crystal MRSA ID System, BBL, Cockeysville, Maryland). In addition to the agar screen, 30-µg cefoxitin disks have been developed for disk diffusion to improve the detection of oxacillin-resistant staphylococci (CLSI M100-22). According to this method, cefoxitin inhibitory zones less than or equal to 24 mm indicate oxacillin resistance in staphylococci. The cefoxitin disk test is especially helpful in detecting oxacillin resistance in coagulase-negative staphylococci.

Similarly, reduced staphylococcal susceptibility to vancomycin (i.e., MICs from 4 to 16 µg/mL) can be difficult to detect by disk diffusion and some commercial methods. Although the therapeutic relevance of staphylococci with vancomycin MICs in this range is currently uncertain, the diminished susceptibility is outside the normal MIC range for susceptible strains; therefore, this phenotype needs to be detected. The agar screen used for this purpose is outlined in Table 12-4 and is essentially the same as that outlined for enterococci, also in Table 12-4. Strains that grow on the screen should be tested by broth microdilution to obtain a definitive MIC value.

Similarly, detection of enterococcal resistance to vancomycin can be difficult by some conventional and commercial methods, and the agar screen may be helpful in confirming the resistance pattern (see Table 12-4). However, as a screen, not all enterococcal isolates capable of growth are resistant to vancomycin at clinically relevant levels. Therefore, strains detected using this method should also be characterized using a broth microdilution method to determine the isolate’s MIC.

Aminoglycosides also play a key role in therapy for serious enterococcal infections, and acquired high-level resistance, which essentially destroys the therapeutic value of these drugs for combination therapy with ampicillin or vancomycin, is not readily detected by conventional methods. Therefore, screens using high concentrations of aminoglycosides (see Table 12-4) have been developed and are available commercially (e.g., Remel, Lenexa, Kansas; or BBL, Cockeysville, Maryland).

With the emergence of penicillin resistance in S. pneumoniae, the penicillin disk diffusion test became insufficiently sensitive to detect subtle but significant changes in susceptibility to penicillin. To address this issue, the oxacillin disk screen described in Table 12-4 is useful but has a notable limitation. Although organisms identified with zones greater than or equal to 20 mm can be accurately characterized as penicillin susceptible, the penicillin susceptibility status of those with zones less than 20 mm remains uncertain, and an MIC value must be determined through an additional test.

With regard to macrolide (e.g., erythromycin, azithromycin, clarithromycin) and lincosamide (e.g., clindamycin) resistance among staphylococci, interpretation of in vitro results can also be complicated by the different underlying mechanisms of resistance that have very different therapeutic implications. Isolates that produce a profile of resistance to a macrolide (e.g., erythromycin) and susceptibility to clindamycin may do so as a result of two different resistance mechanisms. If this profile is the result of the efflux (msrA gene) mechanism, the isolate can be considered susceptible to clindamycin. However, if this profile resulted from the inducible macrolide-lincosamide-streptogramin-B (MLSB) mechanism, which results in an altered ribosomal target, clindamycin-resistant mutants may readily arise during therapy with this agent. Currently such strains should be reported as resistant to clindamycin. The D test that is used to distinguish between these two different resistance mechanisms is outlined in Table 12-4.

Undoubtedly, as complicated resistance mechanisms requiring laboratory detection continue to emerge, screening and supplemental testing methods will continue to be developed. Some of these will be maintained as the primary method for detecting a particular resistance mechanism, whereas others may tend to fade away as adjustments in conventional and commercial procedures enhance resistance detection and preclude the need for a supplemental test.

Predictor Antimicrobial Agents.

Another approach that may be used to ensure accuracy in resistance detection is the use of “predictor” antimicrobial agents in the test batteries. The basic premise of this approach is to use antimicrobial agents (predictor drugs) that are the most sensitive indicators of certain resistance mechanisms. The profile obtained with such a battery is used to deduce the underlying resistance mechanism. A susceptibility report then is produced based on the likely effect the resistance mechanisms would have on the antimicrobials being considered for therapeutic use. The use of predictor drugs is not a new concept, and this approach has been taken in a number of cases, such as the following:

β-Lactamase Detection.

β-lactamases play a key role in bacterial resistance to beta-lactam agents, and detection of their presence can provide useful information (see Chapter 11). Various assays are available to detect β-lactamases, but the most useful in clinical laboratories is the chromogenic cephalosporinase test. β-lactamases exert their effect by opening the β-lactam ring (see Figure 11-9). When a chromogenic cephalosporin is used as the substrate, this process results in a colored product. The Cefinase disk is an example of a commercially available chromogenic test (BD Microbiology Systems, Cockeysville, Maryland). The disk incorporates nitrocefin as the substrate (Figure 12-15).

Useful application of tests to directly detect β-lactamase production is limited to organisms producing enzymes whose spectrum of activity is known. This also must include the β-lactams commonly considered for therapeutic eradication of the organism. Examples of useful applications include detection of:

The actual utility of this approach, even for the organisms listed, is decreasing. As β-lactamase–mediated resistance has become widespread among N. gonorrhoeae, H. influenzae, and staphylococci, other agents not affected by the β-lactamases have become the therapeutic antimicrobials of choice. Therefore, the need to know the β-lactamase status of these bacterial species has become substantially less urgent. Whereas several Enterobacteriaceae and P. aeruginosa produce β-lactamases, the effect of these enzymes on the various β-lactams depends on which enzymes are produced. Therefore, even though such organisms would frequently produce a positive β-lactamase assay, very little, if any, information would be gained about which antimicrobial agents are affected. It is recommended that detection of β-lactam resistance among these organisms be accomplished using conventional and commercial systems that directly evaluate antimicrobial agent/organism interactions.

Chloramphenicol Acetyltransferase Detection.

Chloramphenicol modification by chloramphenicol acetyltransferase (CAT) is one mechanism by which bacteria may express resistance to this agent. This, coupled with the substantially diminished use of chloramphenicol in today’s clinical settings, significantly limits the utility of the CAT detection test. Commercial assays provide a convenient method for establishing the presence of this enzyme. If positive, chloramphenicol resistance can be reported, but a negative test result does not rule out resistance that may be mediated by other mechanisms, such as decreased uptake.

Minimal Bactericidal Concentration.

The MBC test involves continuation of the procedure for conventional broth dilution testing. After incubation and determination of the antimicrobial agent’s MIC, an aliquot from each tube or well in the dilution series demonstrating inhibition of visible bacterial growth is subcultured to an enriched agar medium (usually sheep blood agar). After overnight incubation, the plates are examined and the CFUs determined. With the volume of the aliquot and the number of CFUs obtained, the number of viable cells per milliliter for each antimicrobial dilution can be calculated. This number is compared with the known CFU/mL in the original inoculum. The antimicrobial concentration resulting in a 99.9% reduction in CFU/mL compared with the organism concentration in the original inoculum is recorded as the MBC.

Although the clinical significance of MBC results is uncertain, applications of this information include considering whether treatment failure could be occurring as the result of an organism’s MBC exceeding the serum-achievable level for the antimicrobial agent. Alternatively, if an antibiotic’s MBC is greater than or equal to 32 times higher than the MIC, the organism may be tolerant to the drug. Tolerance, a phenomenon most commonly associated with bacterial resistance to beta-lactam antibiotics, reflects an organism’s ability to be inhibited by an agent that is usually bactericidal. Although the physiologic basis of tolerance has been studied in several bacterial species, the actual clinical relevance of this phenomenon has not been well established.

Time-Kill Studies.

Another approach to examining bactericidal activity involves exposing a bacterial isolate to a concentration of antibiotic in a broth medium and measuring the rate of killing over a specified period. By this time-kill analysis, samples are taken from the antibiotic-broth solution immediately after addition of the inoculum and at regular intervals afterward. Each time-sample is plated to agar plates; after incubation, CFU counts are performed as described for MBC testing. The number of viable bacteria from each sample is plotted over time to determine the rate of killing. Generally, a 1000-fold decrease in the number of viable bacteria in the antibiotic-containing broth after a 24-hour period, compared with the number of bacteria in the original inoculum, is interpreted as bactericidal activity. Although time-kill analysis is frequently used in the research environment to study the in vitro activity of antimicrobial agents, the labor intensity and technical specifications of the procedure preclude its use in most clinical microbiology laboratories to determine the proper treatment of a patient’s infection.

Serum Bactericidal (Schlichter Test).

The serum bactericidal test (SBT) is analogous to the MIC-MBC test except the medium used is the patient’s serum containing the therapeutic antimicrobial agents the patient has been receiving. Using the patient’s serum to detect bacteriostatic and bactericidal activity also allows observation of the antibacterial impact of factors other than the antibiotics (e.g., antibodies and complement).

Two serum samples are required for each test. One is collected just before the patient is to receive the next antimicrobial dose; this is the trough specimen. The other sample is collected when the serum antimicrobial concentration is highest; this is the peak specimen. The appropriate time to collect the peak specimen varies with the pharmacokinetic properties of the antimicrobial agents and the route by which they are being administered. Peak levels for intravenously, intramuscularly, and orally administered agents are generally obtained 30 to 60 minutes, 60 minutes, and 90 minutes after administration, respectively. The trough and peak levels should be collected for the same dose and tested simultaneously.

Serial twofold dilutions of each specimen are prepared and inoculated with the bacterial isolate from the patient (final inoculum of 5 × 10⁵ CFU/mL). Dilutions are incubated overnight. The highest dilution that inhibits visibly detectable growth is the serum-static titer (e.g., 1:8, 1:16, 1:32). Aliquots of known size are then taken from each dilution at or below the serum-static titer (i.e., dilutions that inhibited bacterial growth) and are plated on sheep blood agar plates. After incubation, the CFUs per plate are counted, and the serum dilution resulting in a 99.9% reduction in the CFU/mL, compared with the original inoculum, is recorded as the serum-cidal titer. For example, if a bacterial isolate showed a serum-static titer of 1:32, the tubes containing dilutions of 1:2, 1:4, 1:8, 1:16, and 1:32 would be subcultured. If the 1:8 dilution was the highest dilution to yield a 99.9% decrease in CFUs, the serum-cidal titer would be recorded as 1:8.

The SBT was originally developed to assist in the prediction of the clinical efficacy of antimicrobial therapy for staphylococcal endocarditis. Peak serum-cidal titers of 1:32 to 1:64 or greater have been thought to correlate with a positive clinical outcome. However, even though the test is performed on the patient’s serum, many differences go unaccounted for between the in vitro test environment and the in vivo site of infection. Therefore, although the test is used to evaluate whether effective bactericidal concentrations are being achieved, the predictive clinical value for staphylococcal endocarditis or any other infection caused by other bacteria is still uncertain.

Details regarding the performance of these bactericidal tests are provided in the CLSI document M26-A, “Methods for Determining Bactericidal Activity of Antimicrobial Agents.”

Data Review.

Recognition of unusual resistance profiles is primarily accomplished by carefully reviewing the daily laboratory susceptibility data. Examples of unusual susceptibility profiles for gram-positive and gram-negative bacteria are given in Table 12-7. The examples are a mixture of profiles that clearly demonstrate a likely error (i.e., clindamycin-resistant, erythromycin-susceptible staphylococci); profiles that have rarely been encountered but if observed require immediate attention (i.e., vancomycin resistance in staphylococci); and profiles that have been described but may not be common (i.e., imipenem resistance in Enterobacteriaceae).

The data review process for evaluation of profiles should not be the responsibility of a single person in the laboratory. Furthermore, the process requires checks and balances that do not impede the workflow or increase the time required to get the results to the physicians. The way this is established varies, depending on a particular laboratory’s division of labor and workflow, but several key aspects must be considered:

Resolution.

The importance of having strategies for resolving unusual profiles cannot be overstated. However, developing detailed procedures for every contingency is not possible or practical. Most resolution strategies should focus on certain general approaches, with supervisory or laboratory director consultation always being among the options available to technologists. Although the steps taken to investigate and resolve an unusual profile often depend on the organism and antibiotics involved, most protocols for resolution should include one or more of the following approaches:

Often a quick review of the data recording and interpretation aspects, or purity of culture, will reveal the reason an unusual profile was obtained. Other times more extensive testing, perhaps by more than one method, may be needed to establish the validity of an unusual or unexpected resistance profile.