ABSTRACT
Antimicrobial susceptibility testing for Pseudomonas aeruginosa is critical to determine suitable treatment options. Commercial susceptibility tests are typically calibrated against the reference method, broth microdilution (BMD). Imprecision of MICs obtained by BMD for the same isolate on repeat testing is known to exist. Factors that impact the extent of variability include concentration of the inoculum, operator effects, contents of the media, inherent strain properties, and the testing process or materials. We evaluated the variability of BMD for antipseudomonal beta-lactams (aztreonam, cefepime, ceftazidime, meropenem, piperacillin-tazobactam, ceftazidime-avibactam, and ceftolozane-tazobactam) tested against a collection of P. aeruginosa isolates. Multiple replicate BMD tests were performed, and MICs were compared to assess reproducibility, including the impact of the inoculum and operator. Overall, essential agreement (EA) was ≥90% for all beta-lactams tested. Absolute agreement (AA) was as low as 70% for some beta-lactams. Variability from the inoculum and operators impacted the reproducibility of MICs. Piperacillin-tazobactam exhibited the highest degree of variability with 74% AA and 94% EA. The implications of MIC variability are extensive, as the MIC is essential for multiple facets of microbiology, such as the development of new compounds and susceptibility tests, dose optimization, and pharmacokinetic/pharmacodynamic (PK/PD) targets for individual patients.
KEYWORDS: Pseudomonas aeruginosa, beta-lactams, susceptibility testing
INTRODUCTION
Beta-lactams are often the preferred therapeutic option for Pseudomonas aeruginosa infections due to their established efficacy and safety profile (1). However, widespread resistance to beta-lactams among contemporary isolates of P. aeruginosa highlights the important role of susceptibility testing in determining appropriate treatment options for these infections. In the United States, only 70 to 75% of P. aeruginosa isolates are susceptible to piperacillin-tazobactam and cefepime, and approximately 30% of isolates are multidrug resistant, demonstrating the need for susceptibility testing (2). Expression of AmpC, efflux pumps, and loss of the outer membrane protein, OprD, are the most common contributors of beta-lactams resistance in P. aeruginosa (3).
The accepted variability of the reference broth microdilution (BMD) method for antimicrobial susceptibility testing is ±1-log2 dilution, although some organisms may display a broader range of MICs upon repeat testing (4–6). Despite this inherent variability in the test method, clinicians may scrutinize the absolute MIC value over categorical interpretation (i.e., susceptible versus resistant) when making antimicrobial treatment decisions, particularly for difficult-to-treat organisms such as P. aeruginosa (7). Various methods of susceptibility testing have shown specific challenges or inaccuracies, which can vary depending on the organism and antimicrobial, adding additional uncertainty to the accuracy of an MIC result reported by clinical laboratories (8–11). While no “definitive” MIC exists for a given bacterial isolate, standard and regulatory bodies such as International Organization for Standardization (ISO), Clinical and Laboratory Standards Institute (CLSI), and the U.S. Food and Drug Administration (FDA) define BMD as the reference method for MIC determination. Antimicrobial susceptibility data, particularly performance between two methods, is often described in terms of essential agreement (EA) and categorical agreement (CA). Specifically, EA is defined as the MIC value being within 1-log2 dilution, and CA is defined as the MIC values being within the same categorical interpretation (i.e., susceptible, intermediate, susceptible dose dependent, resistant, or nonsusceptible). The accuracy of any test method can, at most, be as high as that of the reference against which it is calibrated. As such, we sought to understand the variability of BMD among antipseudomonal beta-lactams tested against a collection of P. aeruginosa isolates.
RESULTS
Intraassay precision (bacterial isolate set 1).
The modal MIC distributions of the isolates included in this set are provided in Fig. S1 in the supplemental material. One isolate did not have a unique modal MIC for ceftazidime, for which the MIC values ranged from 4 μg/ml to >64 μg/ml. A second isolate did not have a unique modal MIC for meropenem, for which the MIC values ranged from 0.0625 μg/ml to 0.5 μg/ml. Median and mean colony counts were 8.2 × 105 CFU ml−1 and 8.4 × 105 CFU ml−1, respectively. The inocula ranged from 3.0 × 105 CFU ml−1 to 1.5 × 106 CFU ml−1.
For bacterial isolate set 1, where technician variability was controlled by using the same individual to read all BMDs and BMDs were performed over a short time period, the EA across MIC results was greater than 98%, irrespective of how the data were analyzed (Table 1). However, AA of interinoculum reproducibility (Table 1 [bottom]) was <90% for piperacillin-tazobactam (TZP) (87.58%), cefepime (FEP) (89.84%), and meropenem (88.56%). The difference between EA and AA when comparing the isolate mode to the individual inoculum mode ranged from 4.76% (ceftolozane-tazobactam [TZC], ceftazidime-avibactam [CZA]) to 8.82% (ceftazidime [CAZ]) (Table 1, top). The difference between EA and AA when assessing intrainoculum reproducibility ranged from 4.14% (TZC) to 6.73% (TZP). Lastly, the difference between EA and AA when assessing interinoculum reproducibility ranged from 5.73% (TZC) to 10.83% (TZP). Details of the reproducibility and distribution of the errors are provided in Tables 1 and 2 for the intraassay isolate set.
TABLE 1.
Reproducibility and agreement by antibiotic comparing the mode of the isolate to the mode of the individual inoculum (top) for the intraassay isolate set (bacterial isolate set 1), as well as for intrainoculum reproducibility (middle) and interinoculum reproducibility (bottom)
| Drug | Total no. of MICs | No. of MICs with a result of: |
%AA | %EA | Comment(s) | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| −5 | −4 | −3 | −2 | −1 | 0 | 1 | 2 | 3 | 4 | 5 | 6 | NDa | |||||
| Comparison of the mode of the isolate (n = 9) to the mode of the inoculum (n = 3) | |||||||||||||||||
| TZP | 104 | 0 | 0 | 0 | 0 | 2 | 97 | 4 | 0 | 0 | 0 | 0 | 1 | 1 | 93.27 | 99.04 | 1 isolate had no mode for 1 inoculum |
| ATM | 105 | 0 | 0 | 0 | 0 | 4 | 98 | 1 | 2 | 0 | 0 | 0 | 0 | 0 | 93.33 | 98.10 | |
| FEP | 103 | 0 | 0 | 0 | 0 | 6 | 96 | 0 | 1 | 0 | 0 | 0 | 0 | 2 | 93.20 | 99.03 | 2 isolates had no mode for 1 inoculum |
| CAZ | 102 | 0 | 0 | 0 | 0 | 2 | 93 | 7 | 0 | 0 | 0 | 0 | 0 | 3 | 91.18 | 100.00 | 1 isolate with more than 1 mode |
| MEM | 102 | 0 | 0 | 0 | 0 | 3 | 93 | 5 | 1 | 0 | 0 | 0 | 0 | 3 | 91.18 | 99.02 | 1 isolate with more than 1 mode |
| TZC | 105 | 0 | 0 | 0 | 0 | 3 | 99 | 2 | 1 | 0 | 0 | 0 | 0 | 0 | 94.29 | 99.05 | |
| CZA | 105 | 0 | 0 | 0 | 0 | 0 | 99 | 5 | 1 | 0 | 0 | 0 | 0 | 0 | 94.29 | 99.05 | |
| Comparison of individual replicate (n = 1) to the mode of the inoculum (n = 3) that the replicate came from | |||||||||||||||||
| TZP | 312 | 0 | 0 | 0 | 0 | 13 | 290 | 8 | 0 | 0 | 0 | 1 | 0 | 3 | 92.95 | 99.68 | 1 inoculum for 1 isolate had no mode |
| ATM | 315 | 0 | 0 | 0 | 0 | 8 | 297 | 10 | 0 | 0 | 0 | 0 | 0 | 0 | 94.29 | 100.00 | |
| FEP | 309 | 0 | 0 | 0 | 0 | 8 | 289 | 11 | 1 | 0 | 0 | 0 | 0 | 6 | 93.53 | 99.68 | 2 isolates had no mode for 1 inoculum |
| CAZ | 314 | 0 | 0 | 0 | 0 | 8 | 294 | 11 | 1 | 0 | 0 | 0 | 0 | 1 | 93.63 | 99.68 | 1 replicate had no MIC |
| MEM | 312 | 0 | 0 | 0 | 1 | 9 | 293 | 8 | 1 | 0 | 0 | 0 | 0 | 3 | 93.91 | 99.36 | 1 inoculum for 1 isolate had no mode |
| TZC | 314 | 0 | 0 | 0 | 0 | 2 | 301 | 11 | 0 | 0 | 0 | 0 | 0 | 1 | 95.86 | 100.00 | 1 replicated had no MIC |
| CZA | 315 | 0 | 0 | 0 | 0 | 12 | 296 | 5 | 1 | 1 | 0 | 0 | 0 | 0 | 93.97 | 99.37 | |
| Comparison of individual replicate to the mode of the isolate | |||||||||||||||||
| TZP | 314 | 0 | 0 | 0 | 0 | 16 | 275 | 18 | 1 | 0 | 0 | 2 | 2 | 1 | 87.58 | 98.41 | 1 replicate had no MIC |
| ATM | 315 | 0 | 0 | 0 | 0 | 16 | 284 | 9 | 6 | 0 | 0 | 0 | 0 | 0 | 90.16 | 98.10 | |
| FEP | 315 | 0 | 0 | 0 | 0 | 21 | 283 | 6 | 3 | 1 | 0 | 1 | 0 | 0 | 89.84 | 98.41 | |
| CAZ | 305 | 0 | 0 | 0 | 0 | 9 | 275 | 20 | 1 | 0 | 0 | 0 | 0 | 10 | 90.16 | 99.67 | 1 replicate had no MIC, 1 isolate had more than 1 mode |
| MEM | 306 | 0 | 0 | 0 | 0 | 13 | 271 | 20 | 2 | 0 | 0 | 0 | 0 | 9 | 88.56 | 99.35 | 1 isolate had more than 1 mode |
| TZC | 314 | 0 | 0 | 0 | 0 | 6 | 293 | 12 | 3 | 0 | 0 | 0 | 0 | 1 | 93.31 | 99.04 | 1 replicate had no MIC |
| CZA | 315 | 0 | 0 | 0 | 0 | 9 | 289 | 13 | 3 | 1 | 0 | 0 | 0 | 0 | 91.75 | 98.73 | |
ND, not determined.
TABLE 2.
Reproducibility and agreement by antibiotic comparing the mode of the isolate (n = 9 replicates) to the mode of the individual inoculum (top; n = 3 inocula per isolate) for isolates from intraassay isolate set (bacterial isolate set 1) with a modal MIC ± 1-log2 dilution of intermediate, as well as for intrainoculum reproducibility (middle) and interinoculum reproducibility (bottom)
| Drug | Total no. of MICs | No. of MICs with a result of: |
%AA | %EA | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| −5 | −4 | −3 | −2 | −1 | 0 | 1 | 2 | 3 | 4 | 5 | 6 | NDa | ||||
| Comparison of the mode of the isolate (n = 9) to the mode of the inoculum (n = 3) | ||||||||||||||||
| TZP | 24 | 0 | 0 | 0 | 0 | 1 | 23 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 95.83 | 100.00 |
| ATM | 51 | 0 | 0 | 0 | 0 | 3 | 46 | 1 | 1 | 0 | 0 | 0 | 0 | 0 | 90.20 | 98.04 |
| FEP | 36 | 0 | 0 | 0 | 0 | 1 | 35 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 97.22 | 100.00 |
| CAZ | 27 | 0 | 0 | 0 | 0 | 1 | 24 | 2 | 0 | 0 | 0 | 0 | 0 | 0 | 88.89 | 100.00 |
| MEM | 36 | 0 | 0 | 0 | 0 | 1 | 32 | 3 | 0 | 0 | 0 | 0 | 0 | 0 | 88.89 | 100.00 |
| TZC | 27 | 0 | 0 | 0 | 0 | 0 | 26 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 96.30 | 100.00 |
| CZA | 24 | 0 | 0 | 0 | 0 | 0 | 24 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 100.00 | 100.00 |
| Comparison of individual replicate (n = 1) to the mode of the inoculum (n = 3) that the replicate came from | ||||||||||||||||
| TZP | 72 | 0 | 0 | 0 | 0 | 3 | 68 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 94.44 | 100.00 |
| ATM | 153 | 0 | 0 | 0 | 0 | 5 | 145 | 3 | 0 | 0 | 0 | 0 | 0 | 0 | 94.77 | 100.00 |
| FEP | 108 | 0 | 0 | 0 | 0 | 1 | 105 | 2 | 0 | 0 | 0 | 0 | 0 | 0 | 97.22 | 100.00 |
| CAZ | 81 | 0 | 0 | 0 | 0 | 2 | 78 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 96.30 | 100.00 |
| MEM | 108 | 0 | 0 | 0 | 0 | 2 | 103 | 3 | 0 | 0 | 0 | 0 | 0 | 0 | 95.37 | 100.00 |
| TZC | 81 | 0 | 0 | 0 | 0 | 1 | 78 | 2 | 0 | 0 | 0 | 0 | 0 | 0 | 96.30 | 100.00 |
| CZA | 72 | 0 | 0 | 0 | 0 | 2 | 70 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 97.22 | 100.00 |
| Comparison of individual replicate to the mode of the isolate | ||||||||||||||||
| TZP | 72 | 0 | 0 | 0 | 0 | 6 | 65 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 90.28 | 100.00 |
| ATM | 153 | 0 | 0 | 0 | 0 | 11 | 136 | 3 | 3 | 0 | 0 | 0 | 0 | 0 | 88.89 | 98.04 |
| FEP | 108 | 0 | 0 | 0 | 0 | 3 | 104 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 96.30 | 100.00 |
| CAZ | 81 | 0 | 0 | 0 | 0 | 2 | 75 | 4 | 0 | 0 | 0 | 0 | 0 | 0 | 92.59 | 100.00 |
| MEM | 108 | 0 | 0 | 0 | 0 | 4 | 93 | 11 | 0 | 0 | 0 | 0 | 0 | 0 | 86.11 | 100.00 |
| TZC | 81 | 0 | 0 | 0 | 0 | 0 | 77 | 4 | 0 | 0 | 0 | 0 | 0 | 0 | 95.06 | 100.00 |
| CZA | 72 | 0 | 0 | 0 | 0 | 2 | 70 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 97.22 | 100.00 |
ND, not determined.
The number of isolates close to the breakpoints with a modal MIC ± 1-log2 dilution of the intermediate category (Table S1) were 8 (23%) for TZP, 17 (49%) for aztreonam (ATM), 12 (34%) for FEP, 9 (26%) for CAZ, 12 (34%) for meropenem (MEM), 9 (26%) for TZC, and 8 (23%) for CZA. The overall number of isolates within this subset was relatively small, and the differences in AA and EA agreement compared to that of the total isolate population of the intraassay isolate set were minimal (Table 2).
Table 3 provides the range and mean standard deviations of log2 MICs for each antibiotic for each set of isolates evaluated. The mean standard deviations of log2 MICs for each beta-lactam tested ranged from 0.13 to 0.29, with piperacillin-tazobactam having the largest mean standard deviation. Overall, the maximum standard deviation observed for any beta-lactam for a given isolate was less than 2, with the exception of piperacillin-tazobactam and ceftazidime.
TABLE 3.
Range and mean of the standard deviations of log2 MICs observed for each beta-lactam tested
| Beta-lactam | Intraassay (bacterial isolate set 1) |
Interassay (bacterial isolate set 2) |
||
|---|---|---|---|---|
| Mean SD | Range | Mean SD | Range | |
| Piperacillin-tazobactam | 0.29 | 0.00–2.92 | 0.54 | 0.00–2.06 |
| Aztreonam | 0.18 | 0.00–1.23 | 0.38 | 0.00–1.89 |
| Cefepime | 0.20 | 0.00–1.87 | 0.37 | 0.00–1.19 |
| Ceftazidime | 0.24 | 0.00–2.26 | 0.50 | 0.00–3.33 |
| Meropenem | 0.23 | 0.00–1.00 | 0.44 | 0.00–3.71 |
| Ceftolozane-tazobactam | 0.13 | 0.00–1.04 | NAa | NA |
| Ceftazidime-avibactam | 0.17 | 0.00–1.17 | NA | NA |
NA, not available.
Interassay set (bacterial isolate set 2).
The modal MIC distributions of the isolates included in this set are provided in Fig. S2. Three isolates had no unique mode for one antibiotic, which include two isolates for ceftazidime, for which the MIC values ranged from 4 μg/ml to 8 μg/ml and 2 μg/ml to 32 μg/ml. Another isolate did not have a unique modal MIC for meropenem, for which the MIC values ranged from 0.125 μg/ml to 0.5 μg/ml.
For isolate set 2, where sources of known variability were less controlled (i.e., reader, multiple lots of media, antibiotic powder), the EA was lower than what was observed in the intraassay isolate set but still remained >93%, irrespective of how the data were analyzed (Table 4). Piperacillin-tazobactam AA was consistently the lowest across all three data assessments (Table 4). The difference between EA and AA when comparing the isolate mode to the individual inoculum mode was between 10.29% (MEM) and 17.71% (CAZ). The difference between EA and AA when assessing intrainoculum reproducibility was between 5.60% (CAZ) and 9.98% (TZP). Lastly, the difference between EA and AA when assessing interinoculum reproducibility was between 13.91% (MEM) and 21.91% (CAZ). Details of the reproducibility and distribution of the errors are provided in Tables 4 and 5 for the intraassay isolate set.
TABLE 4.
Reproducibility and agreement by antibiotic comparing the mode of the isolate (n = 9 replicates) to the mode of the individual inoculum (top; n = 3 inocula per isolate) for interassay isolate set (bacterial isolate set 2), as well as intrainoculum reproducibility (middle) and interinoculum reproducibility (bottom)
| Drug | Total no. of MICs | No. of MICs with a result of: |
Comment(s) | ||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| −10 | −5 | −4 | −3 | −2 | −1 | 0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | NDa | %AA | %EA | |||
| Comparison of the mode of the isolate (n = 9) to the mode of the inoculum (n = 3) | |||||||||||||||||||||
| TZP | 182 | 0 | 0 | 0 | 0 | 4 | 14 | 142 | 16 | 2 | 1 | 3 | 0 | 0 | 0 | 0 | 0 | 4 | 78.02 | 94.51 | 4 isolates had no mode for 1 inoculum |
| ATM | 186 | 0 | 0 | 0 | 0 | 1 | 12 | 155 | 16 | 1 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 83.33 | 98.39 | |
| FEP | 183 | 0 | 0 | 0 | 0 | 1 | 15 | 153 | 11 | 3 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 3 | 83.61 | 97.81 | 3 isolates had no mode for 1 inoculum |
| CAZ | 175 | 0 | 0 | 0 | 0 | 0 | 13 | 144 | 18 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 11 | 82.29 | 100.00 | 2 isolates had more than 1 mode for the entire isolate, 2 isolates had no mode for 2 inocula, 1 isolate had no mode for 1 inoculum |
| MEM | 175 | 0 | 0 | 0 | 0 | 0 | 9 | 154 | 9 | 1 | 1 | 1 | 0 | 0 | 0 | 1 | 0 | 10 | 88.00 | 98.29 | 7 isolates had no mode for 1 inoculum, 1 isolate had more than 1 mode for the entire isolate |
| Comparison of individual replicate (n = 1) to the mode of the inoculum (n = 3) that the replicate came from | |||||||||||||||||||||
| TZP | 541 | 0 | 0 | 1 | 0 | 0 | 23 | 485 | 31 | 0 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 17 | 89.65 | 99.63 | 5 replicates had no MIC, 4 isolates with no mode for 1 inoculum |
| ATM | 554 | 0 | 0 | 0 | 0 | 0 | 17 | 508 | 27 | 1 | 0 | 0 | 1 | 0 | 0 | 0 | 0 | 4 | 91.70 | 99.64 | 4 replicates had no MIC |
| FEP | 546 | 0 | 0 | 0 | 0 | 0 | 14 | 508 | 23 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 12 | 93.04 | 99.82 | 3 replicates had no MIC, 3 isolates with no mode for 1 inoculum |
| CAZ | 536 | 1 | 0 | 0 | 0 | 0 | 14 | 503 | 16 | 3 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 21 | 93.84 | 99.44 | 3 replicates had no MIC, 2 isolates had no mode for 2 inocula, 2 isolates had no mode for 1 inoculum |
| MEM | 528 | 0 | 0 | 0 | 0 | 0 | 13 | 485 | 26 | 2 | 0 | 2 | 0 | 0 | 0 | 0 | 1 | 29 | 91.86 | 99.24 | 5 replicates had no MIC, 8 isolates with no mode for 1 inoculum |
| Comparison of individual replicate to the mode of the isolate | |||||||||||||||||||||
| TZP | 552 | 0 | 0 | 0 | 0 | 11 | 53 | 405 | 60 | 11 | 4 | 8 | 0 | 0 | 0 | 0 | 0 | 6 | 73.37 | 93.84 | 6 replicates had no MIC |
| ATM | 554 | 0 | 0 | 0 | 0 | 2 | 37 | 445 | 62 | 4 | 2 | 1 | 1 | 0 | 0 | 0 | 0 | 4 | 80.32 | 98.19 | 4 replicates had no MIC |
| FEP | 553 | 0 | 0 | 0 | 0 | 4 | 51 | 439 | 48 | 11 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 5 | 79.39 | 97.29 | 5 replicates had no MIC |
| CAZ | 534 | 1 | 0 | 0 | 0 | 0 | 49 | 406 | 68 | 7 | 2 | 1 | 0 | 1 | 0 | 0 | 0 | 23 | 76.03 | 97.94 | 5 replicates had no MIC, 2 isolates had more than 1 mode for the entire isolate |
| MEM | 532 | 0 | 0 | 0 | 0 | 0 | 32 | 443 | 42 | 7 | 3 | 5 | 0 | 0 | 0 | 3 | 1 | 22 | 83.27 | 97.18 | 13 replicates had no MIC, 1 isolate had more than 1 mode for the entire isolate |
ND, not determined.
TABLE 5.
Reproducibility and agreement by antibiotic comparing the mode of the isolate (n = 9 replicates) to the mode of the individual inoculum (top; n = 3 inocula per isolate) for isolates from interassay isolate set (bacterial isolate set 2) with a modal MIC ± 1 log2 dilution of intermediate, as well as intrainoculum reproducibility (middle) and interinocula reproducibility (bottom)
| Drug | Total no. of MICs | No. of MICs with a result of: |
%AA | %EA | Comment(s) | ||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| −10 | −5 | −4 | −3 | −2 | −1 | 0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | NDa | |||||
| Comparison of the mode of the isolate (n = 9) to the mode of the inoculum (n = 3) | |||||||||||||||||||||
| TZP | 68 | 0 | 0 | 0 | 0 | 3 | 6 | 48 | 8 | 1 | 1 | 1 | 0 | 0 | 0 | 0 | 0 | 1 | 70.59 | 91.18 | 1 isolate had no mode for 1 inoculum |
| ATM | 102 | 0 | 0 | 0 | 0 | 1 | 9 | 81 | 9 | 1 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 79.41 | 97.06 | |
| FEP | 74 | 0 | 0 | 0 | 0 | 1 | 5 | 65 | 2 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 87.84 | 97.30 | 1 isolate had no mode for 1 inoculum |
| CAZ | 30 | 0 | 0 | 0 | 0 | 0 | 2 | 25 | 3 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 83.33 | 100.00 | |
| MEM | 36 | 0 | 0 | 0 | 0 | 0 | 2 | 30 | 4 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 83.33 | 100.00 | |
| Comparison of individual replicate (n = 1) to the mode of the inoculum (n = 3) that the replicate came from | |||||||||||||||||||||
| TZP | 202 | 0 | 0 | 1 | 0 | 0 | 12 | 177 | 12 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 5 | 87.62 | 99.50 | 2 replicates had no MIC, 1 isolate with no mode for 1 inoculum |
| ATM | 304 | 0 | 0 | 0 | 0 | 0 | 10 | 278 | 15 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 2 | 91.45 | 99.67 | 2 replicates had no MIC |
| FEP | 221 | 0 | 0 | 0 | 0 | 0 | 6 | 207 | 8 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 4 | 93.67 | 100.00 | 1 replicate had no MIC, 1 isolate with no mode for 1 inoculum |
| CAZ | 90 | 0 | 0 | 0 | 0 | 0 | 5 | 82 | 3 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 91.11 | 100.00 | |
| MEM | 107 | 0 | 0 | 0 | 0 | 0 | 4 | 97 | 5 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 90.65 | 99.07 | 1 replicate had no MIC |
| Comparison of individual replicate to the mode of the isolate | |||||||||||||||||||||
| TZP | 204 | 0 | 0 | 0 | 0 | 9 | 22 | 140 | 25 | 4 | 2 | 2 | 0 | 0 | 0 | 0 | 0 | 3 | 68.63 | 91.67 | 3 replicates had no MIC |
| ATM | 304 | 0 | 0 | 0 | 0 | 2 | 31 | 231 | 33 | 4 | 2 | 1 | 0 | 0 | 0 | 0 | 0 | 2 | 75.99 | 97.04 | 2 replicates had no MIC |
| FEP | 224 | 0 | 0 | 0 | 0 | 3 | 21 | 183 | 14 | 3 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 81.70 | 97.32 | 1 replicate had no MIC |
| CAZ | 90 | 0 | 0 | 0 | 0 | 0 | 10 | 69 | 11 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 76.67 | 100.00 | |
| MEM | 107 | 0 | 0 | 0 | 0 | 0 | 7 | 86 | 13 | 0 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 80.37 | 99.07 | 1 replicate had no MIC |
ND, not determined.
The number of isolates close to the breakpoints with a modal MIC ± 1-log2 dilution of the intermediate category were 23 (37%) for TZP, 34 (55%) for ATM, 25 (40%) for FEP, 10 (16%) for CAZ, and 12 (19%) for MEM. Numerical decreases in AA for piperacillin-tazobactam, aztreonam, and meropenem were observed compared to the full isolate set; however, most differences were less than 5% (Table 5). Essential agreement across all antibiotics was similar to that of the total isolate population for the interassay isolate set.
Overall, the mean standard deviations of log2 MICs for each beta-lactam tested ranged from 0.37 to 0.54, with piperacillin-tazobactam representing the largest standard deviation (Table 3). While mean standard deviations for the total isolate population were less than 1, standard deviations above 2 were observed for piperacillin-tazobactam, ceftazidime, and meropenem against some isolates as represented by the range (Table 3).
DISCUSSION
These data confirm that EA (i.e., MICs within 1 doubling dilution) is generally reliable for P. aeruginosa by the reference BMD method for commonly tested beta-lactam agents. Absolute agreement is much less reliable, particularly when the role of the inoculum is considered, and even more so as additional elements of variability are introduced as observed in bacterial isolate set 2. In bacterial isolate set 1, the difference between EA and AA ranged from 4% to 10% across antibiotics evaluated. This difference between EA and AA increased up to 21% when additional inoculum and reader variability was introduced (bacterial isolate set 2), with multiple antibiotics having an AA less than 80%. This ratifies that no single value exists as a definitive MIC, as others have also acknowledged (12).
When assessing a subset of isolates with MICs at the intermediate ±1-log2 dilution, the AA ranged from 68.63% to 81.70% (Table 5 [bottom]). While this is not considerably different than the AA observed for the total population of isolates (73.37% to 83.27%), this is still noteworthy, as many of these MICs outside AA would result in categorical errors since the intermediate breakpoint is often only 1 dilution (Table S1 in the supplemental material). Of note, the wild-type population of P. aeruginosa is within the susceptible breakpoints for all the beta-lactams evaluated except aztreonam (wild-type MIC ≤ 16 μg/ml) and ceftazidime-avibactam (wild type not defined). However, this relative impact on categorical errors is obviously dependent on the MIC distributions evaluated, which can vary among laboratories. Furthermore, while the mean standard deviation of log2 MICs for the overall isolate populations for each antibiotic was less than 1, this was not the case for every isolate (Table 3). Clinically, this is relevant for multiple reasons, but perhaps the most important is that these MIC measurements are often treated as absolute values, not estimates. This plays a more critical role for isolates that exhibit MICs at the breakpoint as various dosing modifications (e.g., higher doses or continuous or prolonged infusion) are often utilized at these higher MICs where there are concerns for insufficient drug exposure (13).
It is clear there are several sources of variability that can play into a single MIC measurement. For example, the media components, inoculum density, conditions of the incubation systems, and process of setting up the tests can all introduce variability that individually may seem marginal but play a considerable cumulative role (5, 12, 14–23). In the current analysis, data for bacterial isolate set 2 (Table 4) were generated over a few years; therefore, it is presumed that different lots of media were used, and multiple technicians generated the MIC data. While different manufacturers or brands of media will have greater variation compared with different lots from the same manufacturer, some lot-to-lot variability will occur (14, 15). Variation in media cation content has been demonstrated to impact MIC results, which can contribute to the interlaboratory variation that is often observed (16). This is one of the reasons that studies that establish quality control (QC) ranges for antimicrobials include multiple laboratories and manufacturers of media resulting in acceptable ranges for QC strains that can span 3 or 4 dilutions (24, 25).
Variances in inoculum have proven to impact the in vitro activity of antimicrobials; however, most of these studies evaluate a defined “low” and “high” inoculum with approximately a 2-log difference (17, 18). What is less understood is the impact of the inoculum when made to a 0.5 McFarland and within the CLSI-recommended range of 2 × 105 CFU ml−1 to 8 × 105 CFU ml−1, which has more direct implications for the clinical laboratory. Smith and colleague sought to understand the inoculum effect within this range (19). They observed notable changes with meropenem MICs when tested against a collection of carbapenem-resistant Enterobacterales (CRE) isolates. Specifically, they reported a 34.8% minor error rate when using an inoculum at the lower end (2.9 × 105 CFU ml−1) of the CLSI acceptable range. While our study was not designed to fully understand the impact of inoculum, the change observed from Table 1 (middle) and Table 2 (middle) to Table 1 (bottom) and Table 2 (bottom) suggests that differences in inoculum concentrations play a role in variability. We also performed colony counts on bacterial isolate set 1, and although a photometric device was used to confirm bacterial suspensions were made to 0.5 McFarland standard, inocula ranged from 3.0 × 105 CFU ml−1 to 1.5 × 106 CFU ml−1, making the upper end slightly outside the CLSI recommended range. These slight inoculum variations likely reflect what is observed in clinical laboratories where colony counts may not be standard practice for each susceptibility test.
While not specifically for P. aeruginosa, a few studies have demonstrated the variation of MICs for various organisms and antimicrobials (5, 20–22). One study included 17 laboratories for which a set of Neisseria gonorrhoeae isolates were tested against ciprofloxacin, penicillin, and tetracycline using Etest (22). The investigators used a unique statistical approach incorporating the censoring of MIC data. The reason for that being the actual MIC, or in this case, the “event,” is actually unknown but, rather, a value in between an interval of two boundaries. Of the 17 laboratories evaluated, three laboratories had mean MICs greater than one doubling dilution different from the overall mean. Another study assessed the variability of daptomycin MICs when tested against a collection of Enterococcus faecium by testing 40 isolates across three laboratories by reference BMD and two gradient strips (5). Even when evaluating reference BMD, 20% of isolates tested in these data set spanned 6 dilutions, illustrating that even reference BMD is not always reproducible. In these scenarios, providing confident interpretive criteria (S/I [susceptible/intermediate] or SDD/R [susceptible dose dependent/resistant]) is very challenging. Wexler and colleagues assessed the variability observed when cefoxitin was tested against Bacteroides fragilis isolates via agar dilution (20). The testing was performed by one laboratory in multiple replicates over different days and utilized multiple readers. Despite readers having extensive training and experience, the data demonstrated a clear “reader effect” impacting the results and signifying that there is an operator impact in these assays, particularly when there is a manual component. The authors also concluded that the test day was also a significant source of variation, which would suggest that inoculum plays a role in variability as previously discussed. Lastly, Mouton and colleagues observed intra- and interlaboratory variability of linezolid when tested against a set of Staphylococcus aureus isolates (21). A collection of S. aureus isolates were sent to five laboratories and tested by Etest in quadruplicate at each site. A maximum average difference of about 0.65-log2 MIC was observed between laboratories, representing interlaboratory variability. The average standard deviation among the replicates tested ranged from 0.14 up to 0.51, representing intralaboratory variability. Another interesting observation was that, despite quadruplicate replicate testing, the majority of variable strains in the interlaboratory analysis could not be identified as variable by intralaboratory analysis.
Piperacillin-tazobactam consistently demonstrated the highest degree of variability in our data. MIC testing of piperacillin-tazobactam requires two separate suspensions (piperacillin and tazobactam), with tazobactam tested at a fixed concentration of 4 μg/ml (26). Altering the tazobactam concentration has previously been demonstrated to have an impact on the MICs of Enterobacterales, often affected by the amount of beta-lactamase production in the isolate (27, 28). It is unknown if this same effect would be observed with P. aeruginosa due to the inherent differences in predominant resistance mechanisms between the two organisms. Also, the addition of tazobactam to piperacillin is presumed to have limited in vitro activity on P. aeruginosa susceptibilities, although piperacillin testing alone in surveillance studies is often no longer conducted (29, 30). Additional data would be needed to fully understand the role of tazobactam to the variability of P. aeruginosa susceptibilities.
The variability of MICs is increasingly important to understand, as emphasis on individualized dosing regimens to achieve specific pharmacokinetic/pharmacodynamic (PK/PD) targets, coupled with therapeutic drug monitoring, come into play. The MIC is a fundamental value for which the proposed PK/PD targets for a given bacteria-antimicrobial combination are necessary to maximize the likelihood of a successful outcome (31). Recent data assessing the probability of target attainment (PTA) for ceftazidime-avibactam utilizing multiple Monte Carlo simulation methods demonstrated that PTAs can vary by ∼10% with just a single-dilution shift from an MIC of 8 μg/ml, the current susceptible breakpoint (32). While these variations may appear to be minor, they often play a significant role when setting breakpoints for new antibiotics, for which a desired PTA threshold is typically 90%. Data assessing piperacillin-tazobactam regimens and adequate exposures against P. aeruginosa isolates have also shown that for a “standard” dose of 4.5 g every 6 h over 30 min, the PTA is ∼80% at an MIC of 16 μg/ml and drops considerably to below 50% for an MIC of 32 μg/ml (33). These data evaluating doses and associated PTAs are the rationale behind modified dosing regimens such as prolonged or continuous infusion to help maximize drug exposures.
Clinical PK/PD studies are conducted to understand the relationship between antibiotic exposure and outcomes of infection. However, MIC data are often lacking for a portion of patients, and investigators are left to make assumptions about their MICs (34, 35). While there is no question about the practical challenges of obtaining MICs in the clinical laboratory, without MIC data, the target exposures carry even larger assumptions. Additionally, determining the appropriate target range is important for the purpose of both efficacy and toxicity (36).
Clinically, MICs are routinely performed once and reported as a single value. However, we know that the MIC is really an estimate of a population, and a more accurate measurement of the MIC is perhaps a range, rather than a single number. To accurately determine this, many replicates are likely needed to account for biological variation, which is impacted by factors such as growth rates, inoculum density, technical variation, and test conditions. The limitations of susceptibility testing are important to consider, particularly as they impact the measurements around clinical breakpoints and interpretations (37). In some regards, these limitations are known to laboratorians and some clinicians. It is at least part of the reason that there is an intermediate or susceptible, dose-dependent category (CLSI) or area of technical uncertainty (EUCAST) for the interpretation of MIC testing; these ranges allow for a buffer to avoid high rates of very major (false-susceptible) and major (false-resistant) errors. These ranges also signal to providers that the isolate is approaching an MIC that may not achieve acceptable PK/PD targets with standard doses. Furthermore, these data highlight some of the challenges a reference method (i.e., BMD) with inherent variability can have for the development of other susceptibility tests used in clinical laboratories. A baseline understanding of the extent of variability with the organisms and antimicrobial combinations of interest may enhance early development efforts of susceptibility tests and antimicrobial compounds.
MATERIALS AND METHODS
Intraassay precision (bacterial isolate set 1).
Thirty-five isolates of P. aeruginosa were evaluated. Three separate inocula were made for each isolate for BMD testing. Each inoculum was then tested in triplicate following the BMD procedures outlined below, resulting in a total of 9 MIC results per isolate. The same technician performed testing and reading for each isolate, eliminating interoperator variability. Colony counts were also performed.
Interassay precision (bacterial isolate set 2).
Sixty-two isolates of P. aeruginosa were evaluated. Each isolate evaluated had a total of 9 MICs, representing 3 separate inocula tested on 3 unique days. Multiple technicians performed testing and reading of the BMDs. The isolates used in set 2 were all unique from set 1, with the exception of 1 isolate.
Broth microdilution.
Antibiotics were tested using frozen panels prepared at Accelerate Diagnostics, Inc. according to CLSI methods (38) using cation-adjusted Mueller-Hinton broth (Difco, BD, Sparks, MD). BMD was performed in triplicate for each inoculum of bacteria. Bacterial inocula were prepared to 0.5 McFarland from isolates subcultured twice on sheep’s blood agar plates from −70°C frozen stocks. DensiChek was used to prepare inocula with additional saline or pure colony as needed for titration. BMD was performed as outlined in CLSI M07, and panels were incubated at 35°C ± 2°C for 16 to 20 h in an ambient air incubator (38). Piperacillin-tazobactam (TZP), aztreonam (ATM), meropenem (MEM), ceftazidime (CAZ), and cefepime (FEP) were evaluated for both isolate sets. Ceftazidime-avibactam (CZA) and ceftolozane-tazobactam (TZC) were only evaluated for bacterial isolate set 1, as these drugs were not tested at the time bacterial isolate set 2 data were generated. Antibiotic concentration ranges (μg/ml) on the BMD panels were 0.25 to 256 and 1 to 256 for TZP, 0.5 to 64 for ATM, 0.5 to 64 for FEP, 0.5 to 64 for CAZ, 0.0.3125 to 128 and 0.03125 to 32 for MEM, 0.25 to 64 for TZC, and 0.25 to 64 for CZA.
Data analysis.
An assessment of reproducibility was performed by comparing the data in the following ways for each antibiotic: (i) to evaluate the effect of the inoculum, the mode of each isolate representing 9 replicates was compared to the mode of each individual inoculum (n = 3); (ii) intrainoculum reproducibility was determined by comparing each individual replicate (n = 1) to the mode of the inoculum (n = 3); and (iii) interinocula reproducibility was determined by comparing each individual replicate (n = 1) to the mode of the isolate (n = 9), with the mode of the isolate representing all 3 inocula. Comparisons were made and assigned a 0 if they were in absolute agreement (AA) and numbers (i.e., −1, −2, 1, 2) depending on how many doubling dilutions of difference existed between the two comparators. AA would represent identical values or no difference between the two MIC values being compared. A value of −1, 0, or 1 (i.e., ±1-log2 dilution) would represent essential agreement (EA). This same assessment was then also performed on a subset of isolates for which the modal MIC was intermediate or within 1 dilution range of intermediate (I ± 1 log2 dilution). For ceftazidime-avibactam, where an intermediate category does not exist, isolates at the susceptible and resistant breakpoints were included for this assessment (i.e., modal MIC of 8 μg/ml and 16 μg/ml). In situations where an isolate had no mode or an inoculum had no mode, the MICs were removed for that comparison. Off-scale isolates (MICs outside the range of concentrations on the BMD panel) were still included in the analysis and considered within AA if the same result was observed (i.e., >64 would be considered in AA with >64).
Next, all MIC data were log transformed (log2 MIC) such that MIC values become integers [e.g., log_2(0.25 μg/ml/μg/ml) = −2, log_2(0.5 μg/ml/μg/ml) = −1, etc.] to compute the standard deviation. For each beta-lactam, the standard deviation of the log2 MICs for each isolate, comprised of 9 replicates, was determined. The mean of the standard deviations for each set of isolates was calculated for each beta-lactam.
ACKNOWLEDGMENTS
We thank all the research associates for their immense efforts in technical assistance at Accelerate Diagnostics, Inc.
Footnotes
Supplemental material is available online only.
REFERENCES
- 1.Ibrahim D, Jabbour J-F, Kanj SS. 2020. Current choices of antibiotic treatment for Pseudomonas aeruginosa infections. Curr Opin Infect Dis 33:464–473. doi: 10.1097/QCO.0000000000000677. [DOI] [PubMed] [Google Scholar]
- 2.Karlowsky JA, Lob SH, Raddatz J, DePestel DD, Young K, Motyl MR, Sahm DF. 2020. In vitro activity of imipenem/relebactam and ceftolozane/tazobactam against clinical isolates of Gram-negative bacilli with difficult-to-treat resistance and multidrug-resistant phenotypes - Study for Monitoring Antimicrobial Resistance Trends United States 2015–2017. Clin Infect Dis doi: 10.1093/cid/ciaa381. [DOI] [PubMed] [Google Scholar]
- 3.Lister PD, Wolter DJ, Hanson ND. 2009. Antibacterial-resistant Pseudomonas aeruginosa: clinical impact and complex regulation of chromosomally encoded resistance mechanisms. Clin Microbiol Rev 22:582–610. doi: 10.1128/CMR.00040-09. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Humphries RM, Ambler J, Mitchell SL, Castanheira M, Dingle T, Hindler JA, Koeth L, Sei K, Hardy D, Zimmer B, Butler-Wu S, Dien BJ, Brasso B, Shawar R, Dingle T, Humphries R, Sei K, Koeth L, CLSI Methods Development and Standardization Working Group of the Subcommittee on Antimicrobial Susceptibility Testing . 2018. CLSI Methods Development and Standardization Working Group best practices for evaluation of antimicrobial susceptibility tests. J Clin Microbiol 56:e01934-17. doi: 10.1128/JCM.01934-17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Campeau SA, Schuetz AN, Kohner P, Arias CA, Hemarajata P, Bard JD, Humphries RM. 2018. Variability of daptomycin MIC values for Enterococcus faecium when measured by reference broth microdilution and gradient diffusion tests. Antimicrob Agents Chemother 62:e00745-18. doi: 10.1128/AAC.00745-18. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Jorgensen JH, Ferraro MJ. 2009. Antimicrobial susceptibility testing: a review of general principles and contemporary practices. Clin Infect Dis 49:1749–1755. doi: 10.1086/647952. [DOI] [PubMed] [Google Scholar]
- 7.Abdul-Aziz MH, Alffenaar J-WC, Bassetti M, Bracht H, Dimopoulos G, Marriott D, Neely MN, Paiva J-A, Pea F, Sjovall F, Timsit JF, Udy AA, Wicha SG, Zeitlinger M, De Waele JJ, Roberts JA, Infection Section of European Society of Intensive Care Medicine (ESICM), Pharmacokinetic/pharmacodynamic and Critically Ill Patient Study Groups of European Society of Clinical Microbiology and Infectious Diseases (ESCMID), Infectious Diseases Group of International Association of Therapeutic Drug Monitoring and Clinical Toxicology (IATDMCT), Infections in the ICU and Sepsis Working Group of International Society of Antimicrobial Chemotherapy (ISAC) . 2020. Antimicrobial therapeutic drug monitoring in critically ill adult patients: a position paper. Intensive Care Med 46:1127–1153. doi: 10.1007/s00134-020-06050-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Tan TY, Ng LSY. 2006. Comparison of three standardized disc susceptibility testing methods for colistin. J Antimicrob Chemother 58:864–867. doi: 10.1093/jac/dkl330. [DOI] [PubMed] [Google Scholar]
- 9.Fuchs PC, Barry AL, Brown SD. 2001. Evaluation of daptomycin susceptibility testing by Etest and the effect of different batches of media. J Antimicrob Chemother 48:557–561. doi: 10.1093/jac/48.4.557. [DOI] [PubMed] [Google Scholar]
- 10.Humphries RM, Pollett S, Sakoulas G. 2013. A current perspective on daptomycin for the clinical microbiologist. Clin Microbiol Rev 26:759–780. doi: 10.1128/CMR.00030-13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Sader HS, Rhomberg PR, Jones RN. 2009. Nine-hospital study comparing broth microdilution and Etest method results for vancomycin and daptomycin against methicillin-resistant Staphylococcus aureus. Antimicrob Agents Chemother 53:3162–3165. doi: 10.1128/AAC.00093-09. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Mouton JW, Muller AE, Canton R, Giske CG, Kahlmeter G, Turnidge J. 2018. MIC-based dose adjustment: facts and fables. J Antimicrob Chemother 73:564–568. doi: 10.1093/jac/dkx427. [DOI] [PubMed] [Google Scholar]
- 13.Monogue ML, Kuti JL, Nicolau DP. 2016. Optimizing antibiotic dosing strategies for the treatment of Gram-negative infections in the era of resistance. Expert Rev Clin Pharmacol 9:459–476. doi: 10.1586/17512433.2016.1133286. [DOI] [PubMed] [Google Scholar]
- 14.Brenner VC, Sherris JC. 1972. Influence of different media and bloods on results of diffusion antibiotic susceptibility tests. Antimicrob Agents Chemother 1:116–122. doi: 10.1128/AAC.1.2.116. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Andrews J, Walker R, King A. 2002. Evaluation of media available for testing the susceptibility of Pseudomonas aeruginosa by BSAC methodology. J Antimicrob Chemother 50:479–486. doi: 10.1093/jac/dkf181. [DOI] [PubMed] [Google Scholar]
- 16.Reller LB, Schoenknecht FD, Kenny MA, Sherris JC. 1974. Antibiotic susceptibility testing of Pseudomonas aeruginosa: selection of a control strain and criteria for magnesium and calcium content in media. J Infect Dis 130:454–463. doi: 10.1093/infdis/130.5.454. [DOI] [PubMed] [Google Scholar]
- 17.Harada Y, Morinaga Y, Kaku N, Nakamura S, Uno N, Hasegawa H, Izumikawa K, Kohno S, Yanagihara K. 2014. In vitro and in vivo activities of piperacillin-tazobactam and meropenem at different inoculum sizes of ESBL-producing Klebsiella pneumoniae. Clin Microbiol Infect 20:O831–839. doi: 10.1111/1469-0691.12677. [DOI] [PubMed] [Google Scholar]
- 18.Bedenić B, Beader N, Zagar Z. 2001. Effect of inoculum size on the antibacterial activity of cefpirome and cefepime against Klebsiella pneumoniae strains producing SHV extended-spectrum beta-lactamases. Clin Microbiol Infect 7:626–635. doi: 10.1046/j.1198-743x.2001.x. [DOI] [PubMed] [Google Scholar]
- 19.Smith KP, Kirby JE. 2018. The Inoculum effect in the era of multidrug resistance: minor differences in inoculum have dramatic effect on MIC determination. Antimicrob Agents Chemother 62:e00433-18. doi: 10.1128/AAC.00433-18. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Wexler HM, Lavin PT, Molitoris E, Finegold SM. 1990. Statistical analysis of the effects of trial, reader, and replicates on MIC determination for cefoxitin. Antimicrob Agents Chemother 34:2246–2249. doi: 10.1128/AAC.34.11.2246. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Mouton JW, Meletiadis J, Voss A, Turnidge J. 2018. Variation of MIC measurements: the contribution of strain and laboratory variability to measurement precision. J Antimicrob Chemother 73:2374–2379. doi: 10.1093/jac/dky232. [DOI] [PubMed] [Google Scholar]
- 22.van de Kassteele J, van Santen-Verheuvel MG, Koedijk FDH, van Dam AP, van der Sande MAB, de Neeling AJ. 2012. New statistical technique for analyzing MIC-based susceptibility data. Antimicrob Agents Chemother 56:1557–1563. doi: 10.1128/AAC.05777-11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Charlton CL, Hindler JA, Turnidge J, Humphries RM. 2014. Precision of vancomycin and daptomycin MICs for methicillin-resistant Staphylococcus aureus and effect of subculture and storage. J Clin Microbiol 52:3898–3905. doi: 10.1128/JCM.01571-14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Belley A, Huband MD, Fedler KA, Watters AA, Flamm RK, Shapiro S, Knechtle P. 2019. Development of broth microdilution MIC and disk diffusion antimicrobial susceptibility test quality control ranges for the combination of cefepime and the novel β-lactamase inhibitor enmetazobactam. J Clin Microbiol 57:e00607-19. doi: 10.1128/JCM.00607-19. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Clinical and Laboratory Standards Institute. 2018. Development of in vitro susceptibility testing criteria and quality control parameters , 4th ed.Approved Standard M23-A5. Clinical and Laboratory Standards Institute, Wayne, PA. [Google Scholar]
- 26.Clinical and Laboratory Standards Institute. 2020. Performance standards for antimicrobial susceptibility testing, 30th ed. CLSI M100-ED30. Clinical and Laboratory Standards Institute, Wayne, PA. [Google Scholar]
- 27.Lister PD, Prevan AM, Sanders CC. 1997. Importance of beta-lactamase inhibitor pharmacokinetics in the pharmacodynamics of inhibitor-drug combinations: studies with piperacillin-tazobactam and piperacillin-sulbactam. Antimicrob Agents Chemother 41:721–727. doi: 10.1128/AAC.41.4.721. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Nicasio AM, VanScoy BD, Mendes RE, Castanheira M, Bulik CC, Okusanya OO, Bhavnani SM, Forrest A, Jones RN, Friedrich LV, Steenbergen JN, Ambrose PG. 2016. Pharmacokinetics-pharmacodynamics of tazobactam in combination with piperacillin in an in vitro infection model. Antimicrob Agents Chemother 60:2075–2080. doi: 10.1128/AAC.02747-15. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Frank U, Mutter J, Schmidt-Eisenlohr E, Daschner FD. 2003. Comparative in vitro activity of piperacillin, piperacillin-sulbactam and piperacillin-tazobactam against nosocomial pathogens isolated from intensive care patients. Clin Microbiol Infect 9:1128–1132. doi: 10.1046/j.1469-0691.2003.00786.x. [DOI] [PubMed] [Google Scholar]
- 30.Jones RN, Stilwell MG, Rhomberg PR, Sader HS. 2009. Antipseudomonal activity of piperacillin/tazobactam: more than a decade of experience from the SENTRY Antimicrobial Surveillance Program (1997–2007). Diagn Microbiol Infect Dis 65:331–334. doi: 10.1016/j.diagmicrobio.2009.06.022. [DOI] [PubMed] [Google Scholar]
- 31.Craig WA. 1998. Pharmacokinetic/pharmacodynamic parameters: rationale for antibacterial dosing of mice and men. Clin Infect Dis 26:1–10. doi: 10.1086/516284. [DOI] [PubMed] [Google Scholar]
- 32.Kidd JM, Stein GE, Nicolau DP, Kuti JL. 2020. Monte Carlo simulation methodologies for β-lactam/β-lactamase inhibitor combinations: effect on probability of target attainment assessments. J Clin Pharmacol 60:172–180. doi: 10.1002/jcph.1510. [DOI] [PubMed] [Google Scholar]
- 33.Kim A, Sutherland CA, Kuti JL, Nicolau DP. 2007. Optimal dosing of piperacillin-tazobactam for the treatment of Pseudomonas aeruginosa infections: prolonged or continuous infusion? Pharmacotherapy 27:1490–1497. doi: 10.1592/phco.27.11.1490. [DOI] [PubMed] [Google Scholar]
- 34.Roberts JA, Paul SK, Akova M, Bassetti M, De Waele JJ, Dimopoulos G, Kaukonen K-M, Koulenti D, Martin C, Montravers P, Rello J, Rhodes A, Starr T, Wallis SC, Lipman J, Roberts JA, Lipman J, Starr T, Wallis SC, Paul SK, Margarit Ribas A, De Waele JJ, De Crop L, Spapen H, Wauters J, Dugernier T, Jorens P, Dapper I, De Backer D, Taccone FS, Rello J, Ruano L, Afonso E, Alvarez-Lerma F, Gracia-Arnillas MP, Fernandez F, Feijoo N, Bardolet N, Rovira A, Garro P, Colon D, Castillo C, Fernado J, Lopez MJ, Fernandez JL, Arribas AM, Teja JL, Ots E, Carlos Montejo J, Catalan M, et al. 2014. DALI: defining antibiotic levels in intensive care unit patients: are current beta-lactam antibiotic doses sufficient for critically ill patients? Clinical Infectious Diseases 58:1072–1083. doi: 10.1093/cid/ciu027. [DOI] [PubMed] [Google Scholar]
- 35.Bauer KA, West JE, O'Brien JM, Goff DA. 2013. Extended-infusion cefepime reduces mortality in patients with Pseudomonas aeruginosa infections. Antimicrob Agents Chemother 57:2907–2912. doi: 10.1128/AAC.02365-12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Richter DC, Frey O, Röhr A, Roberts JA, Köberer A, Fuchs T, Papadimas N, Heinzel-Gutenbrunner M, Brenner T, Lichtenstern C, Weigand MA, Brinkmann A. 2019. Therapeutic drug monitoring-guided continuous infusion of piperacillin/tazobactam significantly improves pharmacokinetic target attainment in critically ill patients: a retrospective analysis of four years of clinical experience. Infection 47:1001–1011. doi: 10.1007/s15010-019-01352-z. [DOI] [PubMed] [Google Scholar]
- 37.Ballestero-Téllez M, Jiménez-Morgades E, Arjona-Camacho P, Blanco-Suárez A, Padilla-Esteba E, Pérez-Jové J. 2020. Inter-technique variability between antimicrobial susceptibility testing methods affects clinical classification of cefuroxime in strains close to breakpoint. Clin Microbiol Infect 26:648.e1-648–e3. doi: 10.1016/j.cmi.2019.12.024. [DOI] [PubMed] [Google Scholar]
- 38.Clinical and Laboratory Standards Institute. 2018. Methods for dilution antimicrobial susceptibility tests for bacteria that grow aerobically, 10th ed. Approved standard M07-A11. Clinical and Laboratory Standards Institute, Wayne, PA. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplemental Table S1 and Figures S1 and S2. Download AAC.00640-21-s0001.pdf, PDF file, 0.3 MB (288.1KB, pdf)