INTRODUCTION
Sepsis is one of the important causes of morbidity and mortality, especially in immunocompromised patients. In the current era of modern medicine, an increasing number of patients receive chemotherapy or immunomodulatory drugs, including those who underwent organ transplantation. Modernization and advancement in the corresponding healthcare sector substantially impact the distribution and outcome of bloodstream infections (BSI).
BSI are associated with high mortality rates, especially in immunocompromised patients. Complications due to bacteremia are more common in neutropenic cancer patients, particularly when BSI are caused by gram-negative microorganisms [1, 2]. Early initiation of the appropriate antimicrobial therapy is the mainstay in the proper management of sepsis. Each hour of the delay in the treatment initiation is associated with a 7.6% reduction in the survival rate of a patient with septic shock [3]. Moreover, early identification of pathogens and selection of the right antimicrobials for the treatment of BSI plays an important role in the reduction of mortality rate [4]. The Surviving Sepsis Campaign guidelines also recommend starting antimicrobial therapy within an hour of the detection of septic shock [5].
To control unrecognized BSI, it can be beneficial to aseptically collect a set of blood cultures in adequate volumes before starting empirical treatment [6]. Blood culture analysis is considered the gold standard in the identification of microorganisms in septicemia, which involves aerobic incubation for a minimum of 14-16 h to yield microorganisms from the blood culture. Even with the advancement in automation in microbiology, blood culture still has a turnaround time (TAT) of approximately 48 h [7]. As per the standard guidelines, antimicrobial susceptibility testing (AST) with an automated system requires subculturing from positive blood culture bottles on Blood and MacConkey Agar plates and AST is performed after 16-18 h of subculture. Hence, an AST report takes around 48 h after the blood culture bottle is flagged as positive by the automated system [8].
There is a need to reduce the TAT of both pathogen identification and AST on blood cultures, which can be achieved by performing relevant identification tests and Direct Antimicrobial Susceptibility Testing (DAST) by the disk diffusion method using samples from the positive blood culture broth [9]. This method decreases the TAT by 24 h, and the early start of administration of the corresponding antimicrobial agents can improve the outcome of the disease as well as reduce the length of hospital stay and treatment costs [10].
The goal of this study was to compare the DAST by disk diffusion with Conventional Antimicrobial Susceptibility Testing (CAST), following the Clinical Laboratory Standards Institute (CLSI) AST breakpoints for interpretation of the results.
MATERIALS AND METHODS
Sample collection and screening
We performed a retrospective cross-sectional comparative study conducted for 18 months (from September 2018 to February 2020) at a tertiary care cancer center. The study included a total of 118 blood samples from different wards and the Intensive Care Unit (ICU) of the same hospital and was approved by the Institutional Ethics Committee, record number 11000048.
Adhering to aseptic precautions, blood samples were collected from patients with suspected bacteremia/septicemia, inoculated into blood culture bottles, and incubated in the BACTEC FX 120 blood culture system. When microorganisms are present in the blood sample inoculated into a BD BACTEC vial, they metabolize available nutrients in the culture medium and release CO2. A dye – a component of the sensor at the bottom of the culture bottles – reacts with CO2 and modulates the amount of light absorbed by a fluorescent material in the sensor. The instrument’s photodetectors measure the fluorescence level that corresponds to the amount of CO2 released by the microorganisms. Analysis of the rate and amount of released CO2 in the medium enables the BD BACTEC fluorescent series instrument to determine whether the culture in the bottle is positive, i.e., whether the test sample contains viable microorganisms. A minimum of 107–108 colony-forming units/ml (CFU/ml) of bacteria is required to obtain positive results [8]. Quality control for the BACTEC bottles was performed using the American Type Culture Collection (ATCC) strains E. coli ATCC 25922 and S. aureus ATCC 25923.
The bottles were removed from the automated system once the sound signal of the instrument indicated the growth of bacteria, and 2 ml of broth was used for smear preparation and Gram staining, along with the microorganism identification test and AST. Only the samples containing gram-negative bacilli (GNB) were included in the study. Blood cultures with two or more morphological types, yeasts, or gram-positive organisms were excluded from the analysis.
DAST by disk diffusion method
Blood culture bottles that were flagged as positive were removed from the BACTEC FX 120 system, and 2 ml of broth was drawn with a syringe for plating and smear examination. Positive for GNB samples were centrifuged at 2000 rpm for 10 min. After discarding the supernatant, 100 μl of the precipitate was diluted in 2 ml of peptone water (Neogen). A sterile cotton wool swab was dipped into the mixture, and the excessive solution was removed by turning the swab against the walls of the container. The swab was used to spread the inoculum evenly over the surface of a susceptibility plate (Mueller Hinton Agar). The following panel of antimicrobial disks (HiMedia, India) were used: amikacin (30 μg), amoxicillin/clavulanic acid (20/10 μg), cefepime (30 μg), ceftriaxone (30 μg), ciprofloxacin (5 μg), cefoperazone/sulbactam (5 μg), ertapenem (10 μg), gentamicin (10 μg), imipenem (10 μg), meropenem (10 μg), piperacillin/tazobactum (10 μg), ceftazidime (30 μg), cefazolin (30 μg), cefotaxime (30 μg), tobramycin (10 μg), doripenem (10 μg). The AST results were interpreted as susceptible (S), intermediate (I), or resistant (R) as per the CLSI guidelines 2018, 2019, and 2020 [11, 12, 13].
CAST by disk diffusion method
Positive blood culture broth samples were plated on MacConkey Agar, Blood Agar, and Chocolate Agar and incubated at 35°C overnight to obtain isolated colonies. Routine biochemical tests were performed to identify microorganisms. The colonies were simultaneously inoculated in Mueller-Hinton Broth, making the suspension equivalent to a 0.5 McFarland standard, and then AST was performed by the Kirby-Bauer disc diffusion method. Interpretation of results was done according to the CLSI guidelines.
Interpretation of AST results and data analysis
Susceptibility results obtained by DAST were compared with CAST.
The following definitions were used (Table 1):
Minor errors (mE): the result obtained by the conventional method corresponds to susceptible (S) or resistant (R), while the result obtained by the direct method corresponds to intermediate (I); alternatively, the conventional method indicates intermediate (I), while the direct method suggests susceptible (S) or resistant (R).
Major errors (ME): the result obtained by the conventional method indicates susceptible (S), whereas the direct method suggests to resistant (R).
Very major errors (VME): the conventional method suggests to resistant (R), while the direct method indicates susceptible (S).
RESULTS
A total of 2414 blood culture samples were analyzed, of which 85.9 % (2073) were sterile, and contaminants/commensals grew in 3.5 % (85). Growth of pathogenic microorganisms was observed in 10.6 % (256) of cases. Growth of fungi was detected in 9% of samples. A total of 118 GNB positive blood samples were tested by DAST and CAST: direct smears from blood culture broth were tested by DAST, whereas the corresponding isolates obtained from the positive blood cultures grown on the corresponding agar were tested by CAST.
Table 2 shows the distribution of gram-negative bacteria isolated from the positive blood cultures for which both DAST and CAST were performed. From all gram-negative organisms isolated from blood cultures, Enterobacterales accounted for 72.0% (85) and nonfermenters for 22.9% (27). Escherichia coli was the most common isolate (41.5%), followed by Klebsiella spp. (22.0%) and Pseudomonas spp. (19.5%).
| Microorganisms | No. of isolates, % (n) |
|---|---|
| Enterobacterales total | 72.0 (85) |
| Escherichia coli | 41.5 (49) |
| Klebsiella spp. | 22.0 (26) |
| Citrobacter spp. | 5.1 (6) |
| Enterobacter spp. | 0.9 (1) |
| Proteus spp. | 0.9 (1) |
| Edwardsiella spp. | 0.9 (1) |
| Salmonella spp. | 0.9 (1) |
| Non-fermenters (NF) total | 28.0 (33) |
| Pseudomonas spp. | 19.5 (23) |
| Acinetobacter spp. | 1.7 (2) |
| Stenotrophomonas spp. | 0.9 (1) |
| Other non-fermenter GNB | 5.9 (7) |
For further analysis of the categorical agreement (CA) between DAST and CAST, we excluded bacteria that were isolated from less than 20 samples (Citrobacter spp., Enterobacter spp., Proteeae tribe, Salmonella spp., Edwardisella spp., and non-fermenters except Pseudomonas species) since the results of the analysis in these cases would not be statistically significant. Hence, the analysis of the categorical agreement between DAST and CAST was based on the study of 98 isolates.
When the results of DAST were compared with the corresponding results of CAST for various groups of microorganisms, they showed a high categorical agreement: 98.9% for Enterobacterales and 99.6% for Pseudomonas spp. (Table 3). The relationship between the two methods was evaluated based on the Chi-square test. No significant difference in the antibiotic susceptibility results determined by DAST and CAST was found in our experiments (p=0.979).
| Microorganisms and antibiotics tested (n×Ab=N) | CA, n (%) | Categorical disagreement, n (%) | Chi-square value | P value | |||
|---|---|---|---|---|---|---|---|
| Minor Error | Major Error | Very Major Error | Total | ||||
| Enterobacteriaceae (75×14=1050) | 1038 (98.9) | 7 (0.7) | 5 (0.5) | 0 | 12 (1.1) | 0.00067 | 0.979 |
| Pseudomonas spp. (23×11=253) | 252 (99.6) | 1 (0.4) | 0 | 0 | 1 (0.4) | ||
| Total (1303) | 1290 (99) | 8 (0.6) | 5 (0.4) | 0 | 13 (1.0) | ||
CA – the categorical agreement between DAST and CAST
Tables 4, 5, and 6 illustrate the AST results obtained by the direct and conventional methods for three gramnegative bacteria.
| Antimicrobial drug used | Minor Error, n (%) | Major Error, n (%) | Very Major Error, n (%) | CA, n (%) |
|---|---|---|---|---|
| Amikacin (30 μg) | 1 (2.0) | 0 | 0 | 48 (98.0) |
| Amoxicillin/clavulanic acid (20/10 μg) | 1 (2.0) | 0 | 0 | 48 (98.0) |
| Cefepime (30 μg) | 0 | 1 (2.0) | 0 | 48 (98.0) |
| Ceftriaxone (30 μg) | 0 | 1 (2.0) | 0 | 48 (98.0) |
| Ciprofloxacin (5 μg) | 1 (2.0) | 0 | 0 | 48 (98.0) |
| Cefoperazone/sulbactam (75/10 μg) | 3 (6.2) | 1 (2.0) | 0 | 45 (91.8) |
| Ertapenem (10 μg) | 0 | 0 | 0 | 49 (100) |
| Gentamicin (10 μg) | 0 | 0 | 0 | 49 (100) |
| Imipenem (10 μg) | 0 | 0 | 0 | 49 (100) |
| Meropenem (10 μg) | 0 | 0 | 0 | 49 (100) |
| Piperacillin/tazobactum (100/10 μg) | 1 (2.0) | 1 (2.0) | 0 | 47 (96.0) |
| Ceftazidime (30 μg) | 0 | 0 | 0 | 49 (100) |
| Cefazolin (30 μg) | 0 | 0 | 0 | 49 (100) |
| Cefotaxime (30 μg) | 0 | 1 (2.0) | 0 | 48 (98.0) |
| Antimicrobial drug used | Minor Error, n (%) | Major Error, n (%) | Very Major Error, n (%) | CA, n (%) |
|---|---|---|---|---|
| Amikacin (30 μg) | 0 | 0 | 0 | 26 (100) |
| Amoxicillin/clavulanic acid (20/10 μg) | 0 | 0 | 0 | 26 (100) |
| Cefepime (30 μg) | 0 | 1 (3.8) | 0 | 25 (96.2) |
| Ceftriaxone (30 μg) | 0 | 1 (3.8) | 0 | 25 (96.2) |
| Ciprofloxacin (5 μg) | 1 (3.8) | 0 | 0 | 25 (96.2) |
| Cefoperazone/sulbactam (75/10 μg) | 2 (7.7) | 0 | 0 | 24 (92.3) |
| Ertapenem (10 μg) | 0 | 0 | 0 | 26 (100) |
| Gentamicin (10 μg) | 0 | 0 | 0 | 26 (100) |
| Imipenem (10 μg) | 0 | 0 | 0 | 26 (100) |
| Meropenem (10 μg) | 0 | 0 | 0 | 26 (100) |
| Piperacillin/tazobactam (100/10 μg) | 0 | 0 | 0 | 26 (100) |
| Ceftazidime (30 μg) | 0 | 0 | 0 | 26 (100) |
| Cefazolin (30 μg) | 0 | 0 | 0 | 26 (100) |
| Cefotaxime (30 μg) | 0 | 1 (3.8) | 0 | 25 (96.2) |
| Antimicrobial drug used | Minor Error, n (%) | Major Error, n (%) | Very Major Error, n (%) | CA, n (%) |
|---|---|---|---|---|
| Amikacin (30 μg) | 0 | 0 | 0 | 23 (100) |
| Cefepime (30 μg) | 0 | 0 | 0 | 23 (100) |
| Ciprofloxacin (5 μg) | 0 | 0 | 0 | 23 (100) |
| Cefoperazone/sulbactam (75/10 μg) | 1 (4.3) | 0 | 0 | 22 (95.7) |
| Doripenem (10 μg) | 0 | 0 | 0 | 23 (100) |
| Gentamicin (10 μg) | 0 | 0 | 0 | 23 (100) |
| Imipenem (10 μg) | 0 | 0 | 0 | 23 (100) |
| Meropenem (10 μg) | 0 | 0 | 0 | 23 (100) |
| Piperacillin/tazobactum (100/10 μg) | 0 | 0 | 0 | 23 (100) |
| Ceftazidime (30 μg) | 0 | 0 | 0 | 23 (100) |
| Tobramycin (10 μg) | 0 | 0 | 0 | 23 (100) |
DISCUSSION
Blood culture analysis is an essential primary test for the diagnosis of bloodstream infection. Our study revealed that 71% of bloodstream infections had an identifiable source and 32% of them were culture-proven [14]. Assessment of the bacteriological profile of pathogens along with their antibiotic susceptibility patterns plays a crucial role in better management of sepsis cases. Gramnegative microorganisms (73%) were the most isolated pathogens in this study. These data correspond to the results published by Parajuli et al. [15] and Keihanian et al. [16], although some studies, such as Maharath et al. [17] and Sahoo et al. [18], reported gram-positive microorganisms as the major type of pathogens in hospital settings. On the other hand, the likelihood of polymicrobial growth in blood culture is exceedingly low.
Early detection of pathogens along with analysis of their antibiotic susceptibility patterns are always the main goals of any diagnostic microbiology laboratory. Compared to CAST, performing DAST on positive blood culture broth provides a clinical team with information on the identity of the pathogen and its antibiotic susceptibility one day earlier, which can accelerate switching from empirical therapy to definitive treatment of the disease. Some of the studies have also proposed methods for DAST on clinical specimens [19]. Blood culture tests are critical investigations for any microbiology department, and a delay in reporting the results can significantly affect morbidity and mortality in patients.
The present study was designed to determine the potential accuracy of DAST on GNB-positive blood culture specimens. A total of 118 blood samples containing GNB were evaluated by DAST and CAST. For the pathogenic microorganisms that were isolated from more than 20 samples, the results of DAST and CAST were further analyzed for categorical agreement. We obtained an extremely satisfactory categorical agreement of 98.86% and 99.60% for Enterobacterales and Pseudomonas species, respectively. Similar findings were reported by Desai et al. [20] and Rajshekar et al. [21] who found the categorical agreement to be 90.4% and 96%, respectively. Our study showed a marginally better agreement, which might be due to the exclusion of gram-positive microorganisms that were isolated in insignificant numbers. The overall high degree of agreement between DAST and CAST results could motivate the substitution of the CAST method for the less time-consuming DAST. That would allow a physician to save 24 hours of crucial time for finding the definitive antibiotic therapy.
Categorical disagreement was found in 1.0 % of cases, among which 0.6 % were minor errors and 0.4 % were major errors. Very major errors were not identified in this study. The error percentages observed in this study were much lower than the acceptable limit according to international standards (ME & VME≤3%, mE≤10%) [13]. In comparison, Desai et al. [20] reported 1.8% VME, 1.9% ME, and 5.8% mE when analyzing the DAST and CAST results for samples with gram-negative microorganisms.
When the performance of DAST and CAST was evaluated for Enterobacterales and Pseudomonas spp., the categorical agreement was found to be extremely satisfactory (>95 %) for most of the tested antibiotics. Most of the errors were observed when comparing DAST and CAST results for two antibiotic groups, namely beta-lactam– beta-lactamase inhibitors (BL-BLI) and cephalosporins. Among Enterobacterales, the lowest categorical agreement was found for cefoperazone/sulbactam (CFS) both in Escherichia coli (91.8% with 6.2 % of mE and 2.0 % of ME) and Klebsiella pneumoniae (92.3 % with 7.7 % of mE). These data agree with the findings of Desai et al. [20] who reported a categorical agreement of 71.7% for ampicillin–sulbactam (BL-BLI), and Rajshekar et al. [21] who reported a categorical agreement of 91.8% for CFS (mE 2.5%, ME 5.5%) and a categorical agreement of 92.4% for piperacillin/tazobactam 100/10 μg (mE 2.1%, ME 5.5%). Thus, we showed concordant results between DAST and CAST for Enterobacterales and Pseudomonas species.
A large separate study is required to confirm the significance of early reporting of blood culture analysis and its impact on the patient’s recovery from the sepsis and clearance of the bloodstream infection. Also, if according to Gram staining the GNB from the blood positive bottle does not belong neither to Enterobacteriaceae nor to Non-fermenter all antibiotics are required to be tested and selective reporting has to be done on 2nd day of hospitalization. Gram-positive cocci (GPC) were not considered in this study, as coagulase-negative staphylococcus is seen in most of the cases where GPC are identified on Gram staining of samples from the positivelyflagged bottles. Similarly, further studies are required to evaluate the efficacy of DAST for analysis of blood cultures that are positive for yeast. An additional assessment of antibiotics sensitivity based on the minimum inhibitory concentration (MIC) would enable clinical teams to initiate treatment with a corresponding antibiotic in the correct dosage according to its MIC value.
CONCLUSION
The goal of diagnostic stewardship is to optimize patient management through continuous improvement in diagnostic care. Timely efforts are required for the early initiation of therapeutic interventions in sepsis, especially in critical illness when prompt administration of active antimicrobial agents is the keystone of sepsis management. DAST by disk diffusion is a promising AST method that reduces TAT to 24 h after a blood culture is flagged positive. We also conclude that one must be vigilant while reporting DAST results for BL-BLI combinations and cephalosporins. Such rapid identification of microorganisms and their antibiotic susceptibility can help clinicians to start targeted antimicrobial therapy earlier, which is essential for critically ill immunocompromised patients and can reduce mortality and morbidity in patients with bloodstream infections.