ABSTRACT
This study investigated the frequency of change of the antimicrobial susceptibility pattern when the same isolate was found in the same patient in various situations. We used laboratory data collected over a period of 8 years (January 2014 to December 2021) at the clinical microbiology laboratory of a tertiary hospital for Escherichia coli, Klebsiella pneumoniae, Enterobacter spp., Pseudomonas aeruginosa, and Staphylococcus aureus. Antimicrobial susceptibility tests (AST) were performed using Vitek 2 automated system. We determined essential agreement and categorical agreement, and introduced the new terms essential MIC increase and change from nonresistant to resistant to present changes in antimicrobial susceptibility over time. During the study period, 18,501 successive AST were included. The risk for S. aureus to be resistant to any antibiotic upon repeated culture was <10% during a follow-up of 30 days. For Enterobacterales, this risk was approximately 10% during a follow-up of 7 days. For P. aeruginosa, this risk was higher. The longer the follow-up period, the higher the risk that the bacteria would show phenotypic resistance. We also found that some drug-bug combinations were more likely to develop phenotypical resistance (i.e., E. coli/amoxicillin-clavulanic acid and E. coli/cefuroxime). A potential consequence of our finding is that if we regard a risk of resistance below 10% as acceptable, it may be feasible to omit follow-up AST within 7 days for the microorganisms investigated in this study. This approach saves money, time, and will reduce laboratory waste. Further studies are needed to determine whether these savings are in balance with the small possibility of treating patients with inadequate antibiotics.
KEYWORDS: antimicrobial susceptibility testing, VITEK, phenotypic resistance, clinical microbiology laboratory, costs, laboratory management
INTRODUCTION
Antimicrobial susceptibility tests (AST) are needed to guide clinicians in choosing the correct antimicrobial therapy. When the same isolate is found in samples from different sites (e.g., urine and blood) in one patient on the same day, the question arises whether AST should be performed on all isolates cultured from that patient on that day. If the same isolate is found on multiple days, the question arises whether referral is possible, and if so, what should be the maximum duration of the referral. For the clinicians, the change from susceptible (S) to resistant (R) when two isolates with the same identification are cultured from the same patient at two different times is relevant because they need to change the antibiotic that they have started based on earlier culture result.
Repeating AST is essential to ensure the accurate prescription of antibiotics, but performing it invariably is costly and time-consuming (1, 2). Moreover, performing AST also requires consumables that significantly contribute to medical waste (3). Clinical microbiology laboratories can play their role in curbing the rising health care cost and to reduce their ecological footprint by selectively performing repeat AST. The knowledge on how AST change over time is therefore paramount in routine clinical microbiology laboratory practice.
However, the number of studies on repeated AST is very limited and currently, no standards exist for determining the optimal frequency of repeat AST (4). In 1989, this topic was investigated by Thomson and coworkers (1). In that study, only 690 strains were investigated using disk diffusion. A more recent study showed that the median time until the antimicrobial susceptibility pattern changed was 12 days (5). However, this study did not analyze the change from non-resistant to resistant phenotype. This change is relevant for the clinicians to alter the prescription of antibiotics. A case series showed that repeating AST every 5 days was inadequate to identify resistant organisms (4). Clearly, more data are needed. Therefore, the aim of this study was to investigate the frequency and timing of the change of the susceptibility pattern between isolates cultured from the same patient. It was not the goal of this paper to investigate the clonal relatedness or molecular resistance development of the follow-up isolates.
MATERIALS AND METHODS
Study design.
This is an observational study using laboratory data collected at the clinical microbiology laboratory of Erasmus University Medical Center in Rotterdam, a tertiary hospital with 1,320 beds. Collected data were: sample type and anatomical origin, isolation date, isolates’ identification, and raw MIC results from Vitek 2 (bioMérieux SA, Marcy-l’Etoile, France) automated AST system (6) of Enterobacterales (in this study: Escherichia coli, Klebsiella pneumoniae, Enterobacter spp.), Pseudomonas aeruginosa, and Staphylococcus aureus. The data were collected over a period of 8 years, between January 2014 and January 2021. The most recent taxonomic nomenclature was used for all isolates.
Microbiological procedures.
Blood culture aerobic and anaerobic bottles (BD Bactec FX Blood Culture System) were cultured for up to 5 days (except for possible endocarditis and prosthetic joint infections for 14 days). Positive blood cultures were inoculated onto Columbia agar with 5% sheep blood (BA), McConkey, Chocolate, and Brucella blood agar (BD Diagnostics, Breda, the Netherlands). BA and McConkey were used for urine cultures. To culture sputum samples, bronchus aspirates, bronchoalveolar lavage fluid, and swab materials of the skin and mucosa, BA, McConkey, and Chocolate agar were used. All agar plates were incubated for 48 h in ambient air supplemented with 5% CO2 at 35°C.
Species identification of the microorganisms was performed using matrix-assisted laser desorption ionization time-of-flight (MALDI-TOF) mass spectrometry (Bruker Daltonik, Bremen, Germany). AST was performed using Vitek 2 instrument with appropriate cards (AST P-567 for S. aureus, AST N-344 for Enterobacterales and P. aeruginosa).
Definitions.
We defined “paired” isolates as two microorganisms with similar identification that originated from different anatomical sites of one patient on the same day, and defined “successive” isolates as two microorganisms with similar identification that were cultured from the same patient at two different time points. A patient can have multiple “successive” isolates if the same microorganism was cultured at >2 time points. In this case, a comparison was performed for each of the intervals.
We calculated categorical agreement (CA) and essential agreement (EA) of “paired” and “successive” isolates according to the Clinical and Laboratory Standards Institute (CLSI) (7). CA was defined as “paired” or “successive” isolates that yielded the same categorical interpretation, and EA was defined as “paired” or “successive” isolates with the MIC result that is within one doubling-dilution.
Both CA and EA measure change in two directions, e.g., from resistant to susceptible and vice versa and higher MIC to lower MIC and vice versa. For “successive” isolates, only MIC increase and change from nonresistant to resistant are clinically relevant. Therefore, we introduced the terms “essential MIC increase” (EMI) and “change from nonresistant to resistant” (CNRR). The EMI was defined as an MIC increase of ≥2 doubling-dilution over time in “successive” isolates. CNRR was defined as a change from nonresistant to resistant over time in “successive” isolates. Further, we also introduced a more stricter category combining CNRR with EMI (CNRR + EMI).
For the analysis, no expert rules or interpretative reading were performed (e.g., deeming clindamycin as resistant for inducible resistance [8] or deeming a third-generation cephalosporin as resistant when extended-spectrum beta lactamases test was positive). Given that breakpoints might change during the study period, we categorized the raw MIC data according to a single EUCAST breakpoint table (version 12.0) (9).
For the calculation, we only used the antibiotics in the Vitek card that are routinely used to treat infections because not all antibiotics in the Vitek card are clinically relevant. For S. aureus, eight antibiotics were selected: erythromycin (MIC range 0.25 to 8 mg/L), clindamycin (0.125 to 4 mg/L), tetracycline (1 to 16 mg/L), trimethroprim-sulfmethoxazole (0.5 to 16 mg/L), vancomycin (0.5 to 2 mg/L), linezolid (0.5 to 4 mg/L), gentamicin (0.5 to 2 mg/L), and rifampicin (0.03 to 4 mg/L). For flucloxacillin, screening for methicillin-resistant S. aureus was performed using cefoxitin disk. For Enterobacterales, 12 antibiotics were selected: amoxicillin-clavulanic acid (range 1 to 32 mg/L), piperacillin-tazobactam (1 to32 mg/L), cefuroxime (1 to 64 mg/L), cefoxitin (4 to 64), cefotaxime (before 2016: 0.5 to 64 mg/L; after 2016: 0.25 to 64 mg/L), ceftazidime (before 2016: 0.5 to 64 mg/L. after 2016: 0.125 to 64 mg/L), meropenem (0.25 to 16 mg/L), imipenem (0.25 to 16 mg/L), gentamicin (1 to 16 mg/L), tobramycin (1 to 16 mg/L), trimethroprim-sulfmethoxazole (1 to 16 mg/L), ciprofloxacin (0.25 to 4 mg/L). Six of these 14 antibiotics were also used for P. aeruginosa: piperacillin-tazobactam, ceftazidime, meropenem, imipenem-cilastatin, tobramycin, and ciprofloxacin.
Statistical analysis.
To have enough statistical power, we only performed an analysis if there were >100 “paired” or “successive” isolates. For data analysis of “successive” isolates, data of Enterobacterales and P. aeruginosa isolated from blood, urine, and respiratory samples were used. For S. aureus, all sample types were included. There were >100 “paired” isolates only for S. aureus, E. coli, and K. pneumoniae.
We calculated the percentage EA and CA of “paired” (two isolates the same day) and “successive” isolates (two isolates from >1 day). The percentage (and its 95% confidence interval) of EMI, CNRR, and CNRR + EMI were calculated for “successive” isolates for different time periods: between 1 and 7 days, between 8 and 30 days, and between 31 and 365 days. For S. aureus, all sample types were included, whereas for E. coli, K. pneumoniae, Enterobacter spp., and P. aeruginosa, only urine, respiratory, and blood samples were included.
RESULTS
“Paired” isolates.
The percentage of EA and CA of two isolates with the same identification from the same patient isolated on the same day are shown in Table 1.
TABLE 1.
Essential and categorical agreement of any antibiotic of two isolates with the same identification from the same patient isolated from the same day
| Microorganisms (n antibiotics tested, no. of paired samples | % Essential agreementa | % Categorical agreementb |
|---|---|---|
| S. aureus (8, n = 183) | 98.9 | 98.3 |
| E. coli (12, n = 418) | 87.6 | 93.1 |
| K. pneumoniae (12, n = 119) | 86.6 | 92.4 |
Two isolates with the same identification from the same patient isolated on the same day with a MIC that is within 1 doubling-dilution in every tested antibiotic.
Two isolates with the same identification from the same patient isolated on the same day that yielded the same categorical interpretation in every tested antibiotic.
“Successive” isolates.
Included were 18,501 “successive” isolates, in which “successive” isolates were defined as two microorganisms with similar identification that were cultured from the same patient at two different time points. The percentage (and 95% confidence interval) of EA, CA, EMI, CNRR, and CNRR + EMI of these samples are presented in Table 2. The EMI, CNRR, and CNRR + EMI of S. aureus were all below 10% up and until 30 days. For E. coli, K. pneumoniae, and Enterobacter spp. the EMI, CNRR, and CNRR + EMI was approximately 10% up to 7 days. P. aeruginosa had a higher EMI, CNRR, and CNRR + EMI than the tested Enterobacterales at every tested interval. The longer the time period between two successive isolates of any identity, the lower the EA and CA, and the higher the EMI, CNRR, and CNRR + EMI.
TABLE 2.
Proportion (%) EA, CA, EMI, CNRR, and CNRR + EMI of any of the tested antibiotics of “successive” isolates (a successive isolate consists of two microorganisms with the same identification at two different time points from the same patient)a
| Microorganism (no. of tested antibiotics) | Changes in any one of the tested antibiotics | No. of days between “successive” isolates (total n = 18,501) |
||
|---|---|---|---|---|
| 1 to 7 days | 8 to 30 days | 31 to 365 days | ||
| S. aureus (8) | n = 2,819 (95% CI) | n = 1,629 (95% CI) | n = 3,521 (95% CI) | |
| Essential agreement (EA) | 94.9 (94.1 to 95.7) | 90.0 (88.5 to 91.5) | 82.9 (81.7 to 84.2) | |
| Categorical agreement (CA) | 92.1 (91.1 to 93.1) | 86.6 (85.0 to 88.3) | 80.4 (79.0 to 81.7) | |
| Essential MIC increase (EMI) | 3.2 (2.6 to 3.8) | 6.1 (5.0 to 7.1) | 9.2 (8.4 to 10.1) | |
| Categorical change from nonresistant to resistant (CNRR) | 4.2 (3.6 to 4.9) | 7.8 (6.6 to 9.0) | 10.3 (9.4 to 11.2) | |
| CNRR + EMI | 2.3 (1.8 to 2.8) | 5.0 (4.0 to 6.0) | 8.0 (7.2 to 8.8) | |
| E. coli (12) | n = 1,334 | n = 1,548 | n = 3,658 | |
| Essential agreement (EA) | 82.0 (79.9 to 84.1) | 66.9 (64.6 to 69.3) | 51.3 (49.7 to 53.0) | |
| Categorical agreement (CA) | 78.9 (76.7 to 81.1) | 63.0 (60.6 to 65.4) | 50.9 (49.3 to 52.5) | |
| Essential MIC increase (EMI) | 11.2 (9.5 to 12.9) | 22.9 (20.8 to 25.0) | 29.4 (27.9 to 30.9) | |
| Categorical change from nonresistant to resistant (CNRR) | 13.5 (11.7 to 15.3) | 25.1 (22.9 to 27.2) | 30.1 (28.6 to 31.6) | |
| CNRR + EMI | 8.8 (7.3 to 10.3) | 20.4 (18.4 to 22.4) | 26.1 (24.7 to 27.6) | |
| K. pneumoniae (12) | n = 472 | n = 428 | n = 703 | |
| Essential agreement (EA) | 81.6 (78.1 to 85.1) | 69.4 (65.0 to 73.8) | 55.0 (51.4 to 58.7) | |
| Categorical agreement (CA) | 84.5 (81.3 to 87.8) | 66.6 (62.1 to 71.1) | 58.3 (54.7 to 62.0) | |
| Essential MIC increase (EMI) | 10.6 (7.8 to 13.4) | 18.5 (14.8 to 22.1) | 22.9 (19.8 to 26.0) | |
| Categorical change from nonresistant to resistant (CNRR) | 9.3 (6.7 to 11.9) | 18.5 (14.8 to 22.1) | 20.9 (17.9 to 23.9) | |
| CNRR + EMI | 6.4 (4.2 to 8.6) | 14.0 (10.7 to 17.3) | 17.2 (14.4 to 20.0) | |
| Enterobacter spp. (7) | n = 381 | n = 117 | n = 122 | |
| Essential agreement (EA) | 81.9 (78.0 to 85.8) | 70.9 (62.7 to 79.2) | 56.6 (47.8 to 65.4) | |
| Categorical agreement (CA) | 90.0 (87.0 to 93.0) | 82.9 (76.1 to 89.7) | 67.2 (58.9 to 75.5) | |
| Essential MIC increase (EMI) | 9.7 (6.7 to 12.7) | 16.2 (9.6 to 22.9) | 25.4 (17.7 to 33.1) | |
| Categorical change from nonresistant to resistant (CNRR) | 5.5 (3.2 to 7.8) | 10.3 (4.8 to 15.8) | 22.1 (14.8 to 29.5) | |
| CNRR + EMI | 4.7 (2.6 to 6.9) | 9.4 (4.1 to 14.7) | 18.0 (11.2 to 24.9) | |
| P. aeruginosa (6) | n = 698 | n = 407 | n = 664 | |
| Essential agreement (EA) | 69.8 (66.4 to 73.2) | 57.0 (52.2 to 61.8) | 53.2 (49.4 to 57.0) | |
| Categorical agreement (CA) | 80.5 (77.6 to 83.5) | 64.6 (60.0 to 69.3) | 67.2 (63.6 to 70.7) | |
| Essential MIC increase (EMI) | 19.3 (16.4 to 22.2) | 32.4 (27.9 to 37.0) | 31.9 (28.3 to 35.5) | |
| Categorical change from nonresistant to resistant (CNRR) | 12.8 (10.3 to 15.2) | 26.0 (21.8 to 30.3) | 20.0 (17.0 to 23.1) | |
| CNRR + EMI | 10.6 (8.3 to 12.9) | 23.6 (19.5 to 27.7) | 16.4 (13.6 to 19.2) | |
For definitions, see Materials and Method section. Numbers between brackets are 95% confidence intervals. For the EMI, CNRR and EMI + CNRR: bold letters: < 10%; grey shade: 10% to 20%; darkgrey shade: > 20%.
Table 3 shows the percentage of “successive” isolates with categorical change from nonresistant to resistant when the MIC change leading to this category change was >1 doubling dilution (CNRR + EMI) specified for each antibiotic. Using this very strict criterion, we showed that the risk of a repeated culture with S. aureus that was resistant to flucloxacillin was 0.5% or less. The CNRR+EMI proportion for S. aureus was highest for erythromycin (less than 5% at 1 year). For E. coli and K. pneumoniae, the proportion was highest for amoxicillin-clavulanic acid, cefuroxime, ciprofloxacin, and trimethoprim-sulfamethoxazole (up to 9.8% at 1 year). For Enterobacter spp. and P. aeruginosa, the CNRR+EMI proportion was highest for piperacillin-tazobactam and ciprofloxacin.
TABLE 3.
Proportion (%) of “successive” isolates with categorical change from nonresistant to resistant when the MIC change leading to this category change was >1 doubling dilution (CNRR + EMI) specified for each antibiotica
| Microorganisms | Antibiotics | No. of days between isolates |
||
|---|---|---|---|---|
| 1 to 7 | 8 to 30 | 31 to 365 | ||
| S. aureus (n = 10,986) | Flucloxacillinb | 0.2 | 0.5 | 0.5 |
| Vancomycin | 0.0 | 0.0 | 0.0 | |
| Clindamycin | 0.5 | 1.2 | 2.2 | |
| Erythromycin | 1.2 | 2.6 | 4.7 | |
| Tetracycline | 0.4 | 1.0 | 1.2 | |
| Linezolid | 0.1 | 0.2 | 0.4 | |
| Trimethoprim-sulfamethoxazole | 0.3 | 0.4 | 0.9 | |
| Gentamicin | 0.0 | 0.3 | 0.5 | |
| Rifampicin | 0.1 | 0.2 | 0.3 | |
|
E. coli (n = 7,936) |
Amoxicillin-clavulanic acid | 1.5 | 6.0 | 9.1 |
| Piperacillin-tazobactam | 3.8 | 5.4 | 4.8 | |
| Cefuroxime | 2.5 | 6.1 | 7.1 | |
| Cefotaxime | 1.0 | 2.9 | 3.7 | |
| Cefoxitin | 0.7 | 2.5 | 3.8 | |
| Ceftazidime | 1.0 | 2.4 | 2.8 | |
| Meropenem | 0.0 | 0.0 | 0.0 | |
| Imipenem | 0.1 | 0.2 | 0.0 | |
| Gentamicin | 1.4 | 2.7 | 3.5 | |
| Tobramycin | 1.5 | 3.0 | 3.7 | |
| Trimethoprim-sulfamethoxazole | 1.4 | 5.7 | 9.8 | |
| Ciprofloxacin | 0.7 | 5.4 | 7.1 | |
|
K. pneumoniae (n = 1,779) |
Amoxicillin-clavulanic acid | 0.4 | 2.8 | 5.8 |
| Piperacillin-tazobactam | 3.0 | 5.6 | 6.5 | |
| Cefuroxime | 2.8 | 6.3 | 8.3 | |
| Cefotaxime | 0.2 | 2.6 | 3.8 | |
| Cefoxitin | 2.3 | 3.5 | 4.4 | |
| Ceftazidime | 0.4 | 2.8 | 3.3 | |
| Meropenem | 0.0 | 0.0 | 0.0 | |
| Imipenem | 0.0 | 0.0 | 0.0 | |
| Gentamicin | 0.4 | 0.9 | 2.0 | |
| Tobramycin | 0.4 | 1.2 | 3.0 | |
| Trimethoprim-sulfamethoxazole | 0.6 | 2.6 | 4.8 | |
| Ciprofloxacin | 1.9 | 3.5 | 6.1 | |
|
Enterobacter spp. (n = 648) |
Piperacillin-tazobactam | 3.7 | 7.7 | 8.2 |
| Meropenem | 0.0 | 0.0 | 0.8 | |
| Imipenem | 0.0 | 0.0 | 0.8 | |
| Gentamicin | 0.5 | 0.0 | 1.6 | |
| Tobramycin | 0.3 | 0.0 | 2.5 | |
| Trimethoprim-sulfamethoxazole | 0.8 | 0.9 | 4.1 | |
| Ciprofloxacin | 0.5 | 0.9 | 6.6 | |
|
P. aeruginosa (n = 1,846) |
Piperacillin-tazobactam | 4.0 | 8.8 | 6.9 |
| Ceftazidime | 3.3 | 5.2 | 3.3 | |
| Meropenem | 1.0 | 1.7 | 1.4 | |
| Imipenem | 3.6 | 8.1 | 3.6 | |
| Tobramycin | 0.6 | 1.0 | 1.4 | |
| Ciprofloxacin | 3.2 | 9.1 | 8.3 | |
Each successive isolate consist of two microorganisms with the same identification from two different time points. No shade: < 3%; grey shade: 3% to 6%; darkgrey shade: >6%.
Flucloxacillin sensitivity was inferred from cefoxitin disk diffusion; therefore, no MIC data was available and only categorical regression could be determined.
DISCUSSION
To the best of our knowledge, only two studies on the changes in the AST pattern have been published (1, 5). One study from 1989 using disk diffusion test (1) and a more recent study that used Vitek results (5). The latter used different approach in analyzing the results, and it showed that 14.9% of the cases (each case was defined as one patient in one admission) had any change in the AST. This study showed that the change in the AST pattern occurred during the median time of 12 days (range 1 to 365 days) (5).
When AST for consecutive isolates of the same species from one patient is not performed, there is a risk that therapy that was started based on previous AST result may no longer be effective. Also, invariable performing AST in consecutive isolates is costly. We performed this study because no standard exists on the optimal frequency of repeat AST. Testing every 2 to 5 days is thought to be sufficient (4), but this rule of thumb is not supported by data.
We found the combined essential agreement of >98% for two S. aureus isolates found in the same patient on the same day. The essential agreement was lower for E. coli and K. pneumoniae (>85%), possibly due to the fact that more antibiotics were tested for these gram negatives. We also found that within 1 week between cultures, the risk of a consecutive S. aureus to show any resistant phenotype compared to the previous isolate was very low (<5%). For E. coli, K. pneumoniae, and Enterobacter, this percentage was below 10% if the combination of change from nonresistant to resistant and essential MIC increase (CNRR + EMI) was evaluated. When specific antibiotics were evaluated, all were below 5% in the category CNRR + EMI. The longer the period between two AST’s, the higher the risk that the bacteria will show a different susceptibility pattern.
A resistance percentage up to 10% is often considered as an acceptable threshold to use an antibiotic for empirical treatment (10), and a modeling study showed that when this threshold is increased to 25%, it is associated with excessive morbidity and mortality (11). If we choose 10% as an acceptable threshold, we can argue that AST should not be repeated within 1 week for the microorganisms and antibiotics investigated in this study because the risk of change from nonresistant to a resistant phenotype combined with an essential MIC increase for any antibiotic in the AST is less than 10% within 1 week (except for P. aeruginosa). Another possible consequence of our findings is that for those laboratories where multiple types of specimens taken on the same day are processed on the same bench, it can be considered to perform AST on one isolate from one specimen only. These approaches will save money, time, and will reduce laboratory waste.
Our results can guide clinicians to choose empirical therapy in the follow-up encounter, given that the risk for change in AST pattern from susceptible (or intermediate) to resistant can be estimated. As expected, the more time between the repeated AST, the higher the risk for the bug to show phenotypic resistance to antibiotics. The risk also varies for individual antibiotic–bacteria combination, but they are all less than 10% within 1 year. Caution is needed that a 10% risk of missing resistance may be acceptable for certain conditions or specimen types, but the threshold for acceptable risk might be lower for others (e.g., blood or cerebrospinal fluid). Also, it is obvious that the clinical judgement is important. A repeat AST should be ordered when the clinical response is less than expected after the use of an antibiotic.
We also observed that some drug–bug combinations are more likely to develop phenotypical resistance (i.e.), E. coli–amoxicillin-clavulanic acid, and E. coli–cefuroxime. There are several possible explanations for this observation. The detection of resistant phenotypes during the follow-up may reflect the rate of development of specific antimicrobial resistance, but it may also reflect technical difficulties in AST of certain antimicrobial–antibiotic combinations, and they can be used as trigger for a more in-depth research comparing various AST methods (12).
There are several limitations of this study. First, we do not investigate the development of resistance or selection of pre-existing mutants. Such a study needs larger resources such as whole-genome sequencing for consecutive isolates to show that they are clonal. Second, we have not included data regarding antibiotic treatment, which could potentially explain the development of resistance between successive isolates. If included, it would also need various data sources because a patient may also receive antibiotics from medical doctors outside of the hospital. Third, we use routine Vitek data. It is possible that the observed changes in AST patterns reflect measurement errors, natural variability, or variability in analysis conditions (13). Still, our study reflects a real-life situation in which the advice of antimicrobial use is based on the Vitek results. Also, we use multiple categories to define “real” change from S or I category to R. Finally, while it is important from a fundamental research point of view, this study was not designed to investigate the development of molecular resistance or clonally related isolates found at two different time points.
In conclusion, if we accept that a change in susceptibility pattern from nonresistant to resistant occurs in less than 10% for any routinely tested antibiotic (for whatever reasons, either technical or real resistance development), and a culture shows the same organism as before (in the case of E. coli, K. pneumoniae, Enterobacter spp., and S. aureus) with 7 days or less between the two cultures, repeat AST could be omitted on the second culture.
Contributor Information
Erlangga Yusuf, Email: angga.yusuf@gmail.com.
Nathan A. Ledeboer, Medical College of Wisconsin
REFERENCES
- 1.Thomson JB, File TM, Burgoon RA. 1989. Repeat antimicrobial susceptibility testing of identical isolates. J Clin Microbiol 27:1108–1111. doi: 10.1128/jcm.27.5.1108-1111.1989. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Benkova M, Soukup O, Marek J. 2020. Antimicrobial susceptibility testing: currently used methods and devices and the near future in clinical practice. J Appl Microbiol 129:806–822. doi: 10.1111/jam.14704. [DOI] [PubMed] [Google Scholar]
- 3.Yusuf E, Luijendijk A, Roo-Brand G, Friedrich AW. 2022. The unintended contribution of clinical microbiology laboratories to climate change and mitigation strategies: a combination of descriptive study, short survey, literature review and opinion. Clin Microbiol Infect 28:1245–1250. doi: 10.1016/j.cmi.2022.03.034. [DOI] [PubMed] [Google Scholar]
- 4.Giltner CL, Kelesidis T, Hindler JA, Bobenchik AM, Humphries RM. 2014. Frequency of susceptibility testing for patients with persistent methicillin-resistant Staphylococcus aureus bacteremia. J Clin Microbiol 52:357–361. doi: 10.1128/JCM.02081-13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Köck R, Schuler F, Idelevich EA, Schaumburg F. 2021. Variability of antibiograms: how often do changes in the antimicrobial susceptibility pattern occur in isolates from one patient? Clin Microbiol Infect 27:1638–1643. doi: 10.1016/j.cmi.2021.02.012. [DOI] [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.Humphries RM, Ambler J, Mitchell SL, Castanheira M, Dingle T, Hindler JA, Koeth L, Sei K, Hardy D, Zimmer B, Butler-Wu S, Dien Bard J, Brasso B, Shawar R, Dingle T, Humphries R, Sei K, Koeth L. 2018. CLSI methods development and standardization working group best practices for evaluation of antimicrobial susceptibility tests. J Clin Microbiol 56. doi: 10.1128/JCM.01934-17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Yusuf E, de Bel A, Bouasse J, Piérard D. 2014. D-Zone test for detection of inducible clindamycin resistance using SirScan paper disks and Rosco Neo-Sensitabs at 25 and 15 mm distances. J Med Microbiol 63:1052–1054. doi: 10.1099/jmm.0.074641-0. [DOI] [PubMed] [Google Scholar]
- 9.EUCAST: Clinical breakpoints and dosing of antibiotics [Internet]. [cited July 5, 2022]. Available from https://www.eucast.org/clinical_breakpoints/.
- 10.Cressman AM, Macfadden DR, Verma AA, Razak F, Daneman N. 2019. Empiric antibiotic treatment thresholds for serious bacterial infections: a scenario-based survey study. Clin Infect Dis 69:930–937. doi: 10.1093/cid/ciy1031. [DOI] [PubMed] [Google Scholar]
- 11.Daneman N, Low DE, Mcgeer A, Green KA, Fisman DN. 2008. At the threshold: defining clinically meaningful resistance thresholds for antibiotic choice in community-acquired pneumonia. Clin Infect Dis 46:1131–1138. doi: 10.1086/529440. [DOI] [PubMed] [Google Scholar]
- 12.Sundsfjord A, Giske C. 2022. The clinical impact of errors in semiautomated antimicrobial susceptibility testing of multidrug-resistant Enterobacterales causing bloodstream infections. Clin Microbiol Infect 28:1179–1181. doi: 10.1016/j.cmi.2022.05.021. [DOI] [PubMed] [Google Scholar]
- 13.Hatsuda Y, Ishizaka T, Koizumi N, Yasui Y, Saito T, Maki S, Omotani S, Mukai J, Tachi T, Teramachi H, Myotoku M. 2020. Influence of analysis conditions for antimicrobial susceptibility test data on susceptibility rates. PLoS One [Internet] 15:e0235059. doi: 10.1371/journal.pone.0235059. [DOI] [PMC free article] [PubMed] [Google Scholar]