Abstract
Background
Currently, third generation cephalosporin resistant Enterobacterales (3GCRE) are becoming more common in community-acquired infection, leading to increasing consumption of carbapenems. Because community-acquired 3GCRE infections are generally less severe and of lower pathogenicity, the impact of inappropriate empirical antibiotics among patients with community-acquired 3GCRE bacteremia remains unknown.
Materials and Methods
This prospective cohort study included adult patients with 3GCRE bacteremia from April 2018 to December 2021. Participants were followed for 30 days to measure the primary outcome of mortality. Propensity score analysis was performed to adjust for treatment selection bias.
Results
A total of 155 patients met the eligible criteria (42 participants in the appropriate antibiotics group, and 113 participants in the inappropriate antibiotics group). Eight participants in the inappropriate antibiotics group never received appropriate antibiotics, three of whom died before microbiological results were made available. The most common clinical syndromes were urinary tract infection (56.8%) and biliary tract infection (22.6%). The overall 30-day mortality rate was 12.9%, 14.3% in the appropriate empirical antibiotics group and 12.4% in the inappropriate empirical antibiotics group. After propensity score weighted adjustment, the 30-day mortality rate in the inappropriate group was non-inferior to the appropriate group (mean difference 1.9%; 95% confidence interval: -10.1 - 14.0). From the multivariate analysis, acute respiratory failure and primary bacteremia were associated with 30-day mortality.
Conclusion
Among patients with community-acquired 3GCRE bacteremia, inappropriate empirical treatment given within 24 hours after the onset of bacteremia was non-inferior to appropriate antibiotics. In the setting of a high prevalence of 3GCRE carriage in community, adjustment to carbapenem might be tolerable among patients with community-acquired infections.
Trial Registration
ClinicalTrials.gov Identifier: NCT03765749
Keywords: Third generation cephalosporin-resistant Enterobacterales, Community-acquired infection, Bacteremia, ESBL
Introduction
In the past decade, the incidence of third generation cephalosporin resistant Enterobacterales (3GCRE) infection has increased worldwide [1,2]. The 3GCRE commonly occurs when these bacteria express third-generation cephalosporin beta-lactamase enzymes. These enzymes include extended-spectrum beta-lactamase (ESBL), AmpC enzymes, and OXA-type beta-lactamase [3]. Traditionally, this form of resistance is commonly found among patients with specific risk factors, such as recent hospitalization or exposure to broad-spectrum antibiotics [4]. However, an increasing incidence of 3GCRE intestinal carriage in the community has been described [5]. Spread of 3GCRE outside hospitals has in turn led to the increased use of inappropriate empirical antibiotics in community-acquired infections [6]. Inappropriate antibiotics use has been shown to be a significant risk factor for mortality in 3GRCE associated infections; in a systematic review, delayed appropriate antibiotics use of greater than 24 hours was associated with a 1.6 fold increase in mortality for patients infected with Enterobacterales, mainly 3GCRE [7]. Inappropriate antibiotics use is also associated with increased hospital length of stay in Gram-negative sepsis [8]. However, most studies have been conducted in patients with hospital-acquired and hospital-associated infections, which were associated with higher clinical severity and mortality.
Two studies in patients with community-onset 3GCRE bacteremia did not discover any association between inappropriate antibiotics use and mortality [9,10]. These studies were conducted in the setting of low prevalence of community carriage at around 20%; and therefore, the majority of patients had a healthcare-associated infection [11]. Moreover, one of the studies analyzed only patients who received definitive treatment, excluding patients who might have died from inappropriate treatment before receiving definite treatment.
Because carbapenems are the treatment of choice for 3GCRE infections, community 3GCRE infection has led to an increased global carbapenem consumption and placed selection pressures for the emergence of carbapenem-resistant organisms [12,13]. In Thailand, the rate of intestinal carriage of 3GCRE is more than 50%, leading to an increasing trend of carbapenem use for the management of community-acquired infection [14]. Antimicrobial stewardship in this setting has become challenging [15]. However, there is support that community - acquired 3GCRE strains are associated with lower pathogenicity [16]. Whether early inappropriate antibiotics use in community-acquired infection is associated with worse outcomes remains unknown. As there are no well-controlled prospective studies to address this issue, this study aimed to investigate whether early inappropriate empirical antibiotics use was non-inferior to appropriate antibiotics with respect to clinical outcomes in patients with community-acquired 3GCRE bacteremia.
Materials and Methods
1. Study design
In this prospective cohort study, we enrolled adult patients (over 18 years) with the first onset of community-acquired monomicrobial 3GCRE bacteremia between April 2018 and December 2021 at Phramongkutklao Hospital, a 1,200-bed tertiary care university hospital in Bangkok, Thailand. Community-acquired infection was defined by the previously described criteria as those without: previous hospitalization within 90 days, previous antibiotics exposure within 90 days, history of hemodialysis or placement in a long-term health care facility [17]. Exclusion criteria included participants who were pregnant or transferred out to another hospital within the first seven days after onset of bacteremia.
A list of patients with monomicrobial 3GCRE bacteremia were notified to the investigators every month by the hospital microbiological laboratory. Participants were subsequently assessed for eligibility and classified in two groups according to the appropriateness of empirical antibiotics use. Appropriate antibiotics were defined as antibiotics with in vitro activity against etiologic organisms according to laboratory susceptibility data. Empirical antibiotics were defined as antibiotics administered within the first 24 hours. Definite antibiotics were defined as antibiotics administered after the availability of antimicrobial susceptibility test.
Demographic data, preexisting medical conditions, source of bacteremia, source control measures, microbiological data, empirical treatment, definitive treatment, antibiotics de-escalation procedures, and outcome data for these eligible patients were retrieved from electronic medical records. Infectious diseases-trained physicians who provided patient care determined the source of infection and the appropriateness of source control measures. Mortality was assessed during the period of 30 days after the onset of bacteremia. Patients were routinely followed up by the hospital team after discharge and those who missed their appointment were assessed by telephone calls conducted by the study team. Those unable to be contacted were recorded as missing values after the time of the last known alive status. To minimize the effects of early death on length of stay, those who died during hospital admission were excluded from length of stay analysis.
Complications were defined within the first 24 hours after bacteremia. Septic shock was defined when norepinephrine was given. Acute kidney injury was defined as presence any of the following: increased serum creatinine ≥0.3 mg/dL within 48 hours, or increase in serum creatinine to ≥1.5 times baseline known or presumed to have occurred within the prior days [18]. Respiratory failure was defined when participants needed a mechanical ventilator or high flow oxygen. Disseminated intravascular coagulation (DIC) was defined when the DIC score endorsed by the International Society on Thrombosis and Haemostasis ≥5 points [19].
2. Ethics statement
The study protocol followed the guidelines of the Declaration of Helsinki and ethics approval was obtained from the Institutional Review Board Royal Thai Army Department. (IRB No R097h/61). This study was registered with the ClinicalTrials.gov, number NCT03765749. Informed consent was obtained in written forms from all participants or legal guardians in the study, either in person or by telephone call with the legal guardians in case participants had died before the enrolment.
3. Microbiological analysis
Bacterial isolates were collected in a clinical microbiology laboratory at the study hospital. Species identification was performed using Matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF MS, Bruker Daltonics, Germany). Antimicrobial susceptibility testing was performed using the automated broth microdilution method (Sensititre, TREK Diagnostic Systems Inc, Cleveland, OH, USA) and was interpreted according to the recommendations of the Clinical and Laboratory Standards Institute (CLSI) [20]. The 3GCRE was defined when Enterobacterales were resistant to at least one of the third-generation cephalosporins, i.e. ceftriaxone, ceftazidime, but had pan-susceptibility to carbapenems.
4. Statistical analysis
This study was designed as a non–inferiority study. A previous study found that the stabilized propensity score weighted mortality among patients with appropriate and inappropriate antibiotics at 7.7 and 3.7%, respectively [9]. With 80% power, a one-sided alpha level of 0.025, and non-inferiority margin of 12%, 33 participants were required in the appropriate antibiotics group. Because the rate of appropriate treatment received was 30.4% [10], a total of 108 participants would be required. Therefore, we planned the enrolment period to take place over four years between April 2018 and 2021.
For categorical variables, Chi-square or Fisher’s exact test was used, while Student’s t test or the Mann-Whitney U was used to compare continuous variables. To adjust for treatment selection bias, propensity score analysis using the inverse probability of treatment weights (IPTW) was performed. Propensity scores were calculated based on multivariable logistic regression modeling including age, pulse rate, consciousness status, white blood cell count, respiratory failure, septic shock, disseminated intravascular coagulation, intraabdominal infection status, treatment with surgical drainage, time to first antibiotics, Escherichia coli infection, Pitt bacteremia score, and Charlson comorbidity index score. Sensitivity analysis was conducted to investigate the impact of never receiving appropriate antibiotics by excluding participants in the inappropriate antibiotics group who never received definite appropriate antibiotics.
For factors associated with 30-day mortality, potentially clinical significant predictors in the univariate analyses (P-value <0.05) were subsequently included in a multiple logistic regression model. For all analyses, a two-sided P-value <0.05 was considered significant. Data were analyzed using SPSS (Version 28, SPSS Inc, Chicago, IL, USA).
Result
During the study, a total of 694 isolates of monomicrobial 3GCRE bacteremia were notified by the microbiological laboratory, 155 cases of which met the eligible criteria (42 participants in the appropriate antibiotics group, and 113 participants in the inappropriate antibiotics group) as described in Figure 1. The majority of participants were male (58.1%), with 39.4% having a diagnosis of diabetes. The most common clinical syndromes were urinary tract infection (56.8%), and biliary tract infection (22.6%). Participants in the appropriate empirical antibiotics group had a significantly higher Charlson comorbidity index (P = 0.002). A trend toward a higher proportion of surgical drainage or early intervention for patients with urinary tract or biliary infection was observed in the inappropriate empirical antibiotics group (31.9% vs. 16.7%, P = 0.06). Baseline characteristics and severity of participants between the two groups are shown in Table 1.
Figure 1. Participant enrolment.
Table 1. Baseline characteristics, severity of participants in appropriate empirical antibiotics and inappropriate empirical antibiotics groups.
| Characteristics | Non-matching | Stabilized propensity score weighted | |||||
|---|---|---|---|---|---|---|---|
| Appropriate empirical antibiotics (n = 42) | Inappropriate empirical antibiotics (n = 113) | P-value | Appropriate empirical antibiotics (n = 33.4) | Inappropriate empirical antibiotics (n = 107.8) | P-value | ||
| Male | 24 (57.1) | 66 (58.4) | 0.887 | 19.8 (59.3) | 65.6 (60.9) | 0.812 | |
| Age (years) | 78 (64 - 85) | 72 (63 - 82) | 0.346 | 71 (61 - 84) | 73 (64 - 84) | 0.910 | |
| Underlying diseases | 42 (100) | 109 (96.5) | 0.575 | 33.4 (100) | 104.0 (96.4) | 0.573 | |
| Charlson comorbidity index score | 2 (2 - 3) | 1 (1 - 3) | 0.002 | 2 (1 - 3) | 2 (1 - 3) | 0.273 | |
| Diabetes mellitus | 18 (42.9) | 43 (38.1) | 0.586 | 13.1 (39.2) | 41.5 (38.5) | 0.912 | |
| Cirrhosis | 3 (7.1) | 4 (3.5) | 0.586 | 0.9 (2.7) | 4.2 (3.9) | 1.000 | |
| Chronic kidney disease | 10 (23.8) | 21 (18.6) | 0.389 | 8.8 (26.4) | 20.8 (19.3) | 0.381 | |
| Ischemic heart disease | 8 (19) | 10 (8.8) | 0.470 | 6 (18) | 11.7 (10.9) | 0.287 | |
| Autoimmune disease | 3 (7.1) | 1 (0.9) | 0.078 | 2.6 (7.8) | 1 (0.9) | 0.043 | |
| Cerebrovascular disease | 8 (19) | 18 (15.9) | 0.061 | 5.6 (16.8) | 17.7 (16.4) | 0.894 | |
| Dementia | 5 (11.9) | 4 (3.5) | 0.644 | 3.1 (9.3) | 4.1 (3.8) | 0.354 | |
| Solid malignancy | 11 (26.2) | 18 (15.9) | 0.061 | 6.2 (18.6) | 20 (18.6) | 1.000 | |
| Hematological malignancy | 0 (0) | 4 (3.5) | 0.145 | 0 (0) | 3.1 (2.9) | 1.000 | |
| Clinical parameters | |||||||
| Mean arterial pressure | 86.3 (75.7 - 97.3) | 87.7 (76.7 - 98.3) | 0.693 | 86 (78.7 - 97.3) | 88 (76.3 - 98.3) | 0.799 | |
| Body temperature | 38.15 (36.7 - 39) | 38 (37 - 38.7) | 0.875 | 38.4 (37 - 39) | 38 (37 - 38.8) | 0.379 | |
| Pulse rate | 108.5 (90 - 116) | 97 (82 - 112) | 0.167 | 106 (91 - 116) | 97 (82 - 114) | 0.492 | |
| Respiratory rate | 20 (18 - 22) | 20 (18 - 22) | 0.706 | 20 (18 - 22) | 20 (18 - 22) | 0.423 | |
| Glasgow coma score | 15 (15 - 15) | 15 (15 - 15) | 0.775 | 15 (15 - 15) | 15 (15 - 15) | 0.991 | |
| Oxygen saturation (%) | 98 (93 - 99) | 98 (96 - 99) | 0.856 | 98 (95 - 99) | 98 (96 - 99) | 0.776 | |
| Sodium | 136 (131 - 139) | 135 (132 - 138) | 0.955 | 136 (132 - 139) | 135 (132 - 139) | 0.973 | |
| Potassium | 3.9 (3.6 - 4) | 3.9 (3.43 - 4.08) | 0.796 | 3.9 (3.66 - 4) | 4 (3.43 - 4.08) | 0.777 | |
| Creatinine | 1.4 (0.8 - 2.1) | 1.2 (0.85 - 2.1) | 0.863 | 1.48 (0.8 - 2.77) | 1.28 (0.85 - 2.1) | 0.913 | |
| Hematocrit (%) | 33.8 (30.6 - 37.1) | 35 (29.6 - 39) | 0.708 | 34.6 (33.1 - 37.7) | 35 (29.6 - 39) | 0.884 | |
| White blood cell (cells ×109/L) | 11.2 (9 - 1.6) | 16 (10.7 - 19.4) | 0.051 | 11.9 (9 - 19) | 15.9 (10.7 - 19.2) | 0.463 | |
| Platelet (cells ×109/L) | 220.5 (168 - 260) | 196 (145 - 271) | 0.667 | 221 (160 - 273) | 196 (147 - 257) | 0.454 | |
| Lactate (mmol/L) | 2 (1.33 - 4) | 2 (1.2 - 3.59) | 0.808 | 2 (1.33 - 4) | 2.11 (1.2 - 4) | 0.903 | |
| Pitt bacteremia score | 3 (2 - 6) | 3 (2 - 4) | 0.095 | 3 (2 - 6) | 3 (2 - 4) | 0.193 | |
| APACHE II score | 11 (8 - 14) | 11 (8 - 14) | 0.906 | 12 (8 - 14) | 11 (9 - 14) | 0.808 | |
| Clinical syndrome | |||||||
| Urinary tract infection | 24 (57.1) | 64 (56.6) | 0.955 | 21.7 (65) | 63.6 (59) | 0.571 | |
| Biliary tract infection | 7 (16.7) | 28 (24.8) | 0.293 | 4.6 (13.8) | 23.3 (21.6) | 0.387 | |
| Other intraabdominal infections | 5 (11.9) | 3 (2.7) | 0.034 | 1.9 (5.7) | 3.8 (3.5) | 0.582 | |
| Skin and soft tissue infection | 1 (2.4) | 3 (2.7) | 1.000 | 1.1 (3.3) | 2.3 (2.1) | 0.554 | |
| Primary bacteremia | 3 (7.1) | 8 (7.1) | 1.000 | 2.4 (7.2) | 7.8 (7.2) | 1.000 | |
| Others | 3 (7.1) | 7 (6.2) | 1.000 | 2.3 (6.9) | 7.2 (6.7) | 1.000 | |
| Complications | 25 (59.5) | 66 (58.4) | 0.900 | 20 (59.9) | 62.8 (58.3) | 0.816 | |
| Septic shock | 15 (35.7) | 29 (25.7) | 0.217 | 12.6 (37.7) | 30 (27.8) | 0.247 | |
| Acute kidney injury | 16 (38.1) | 47 (41.6) | 0.694 | 14 (41.9) | 44.2 (41) | 0.863 | |
| Respiratory failure | 13 (31) | 24 (21.2) | 0.207 | 10.7 (32.1) | 25.3 (23.5) | 0.295 | |
| Disseminated Intravascular Coagulation | 11 (26.2) | 37 (32.7) | 0.433 | 10.4 (31.2) | 34.5 (32) | 0.873 | |
Data are n (%), or median (IQR).
Almost all participants in the appropriate antibiotics group received carbapenems, mainly ertapenem, for empirical and definitive treatment (92.9 and 95.2%, respectively). On the other hand, most participants in the inappropriate antibiotics group received ceftriaxone as empirical treatment and carbapenems as definitive treatment (88.5 and 82.3% respectively). Participants in the inappropriate antibiotics group received appropriate antibiotics with a median delay of 55 hours after the onset of bacteremia. Eight participants in the inappropriate antibiotics group never received appropriate antibiotics, three of whom died before the report of microbiological results was available. Antibiotics treatment of participants between the two groups are shown in Table 2.
Table 2. Antibiotics treatment among participants in appropriate empirical antibiotics and inappropriate empirical antibiotics groups.
| Antibiotics | Appropriate empirical antibiotics (n = 42) | Inappropriate empirical antibiotics (n = 113) | |
|---|---|---|---|
| Antibiotics within 24 hours | |||
| Piperacillin/tazobactam | 2 (4.8) | 0 (0.0) | |
| Ertapenem | 35 (83.3) | 0 (0.0) | |
| Imipenem/cilastatin | 2 (4.8) | 0 (0.0) | |
| Meropenem | 2 (4.8) | 0 (0.0) | |
| Ciprofloxacin | 1 (2.4) | 1 (0.9) | |
| Ceftriaxone | 6 (14.3)a | 100 (88.5) | |
| Azithromycin | 0 (0.0) | 1 (0.9) | |
| Amoxicillin/clavulanate | 0 (0.0) | 1 (0.9) | |
| Ceftazidime | 0 (0.0) | 1 (0.9) | |
| None | 0 (0.0) | 9 (8.0) | |
| Definite antibiotics with activity against 3GCRE | |||
| Ertapenem | 36 (85.7) | 89 (78.8) | |
| Imipenem/cilastatin | 0 (0.0) | 1 (0.9) | |
| Meropenem | 4 (9.5) | 3 (2.7) | |
| Ciprofloxacin | 2 (4.8) | 8 (7.1) | |
| Trimethoprim/sulfamethoxazole | 0 (0.0) | 4 (3.5) | |
| None | 0 (0.0) | 8 (7.1) | |
Data are n (%), 3GCRE: third generation cephalosporin resistant Enterobacterales.
aCeftriaxone was not counted as an appropriate antibiotics but was given within 24 hours before the appropriate antibiotics.
E. coli was the most frequently isolated organism (85.8%). The overall 30-day mortality rate across both groups was 12.9%. The difference in mortality between the two groups was not statistically significant (14.3% vs. 12.4%, mean difference 1.9%; 95% confidence interval [CI]: -10.1 – 14.0%). The length of hospital stay and hospital mortality did not significantly differ between the two groups. After propensity score adjustment, the two groups were comparable regarding baseline characteristics and severity, except for underlying autoimmune diseases which were more common in the appropriate antibiotics group. The adjusted difference in mortality between the two groups was 1.6% (95% CI: -11.9 - 15.1%). The length of hospital stay and hospital mortality after adjusting were not statistically different between the two groups. The results of matching are shown in Table 3.
Table 3. Causative organisms and mortality of participants in appropriate empirical antibiotics and inappropriate empirical antibiotics groups.
| Characteristics | Non-matching | Stabilized propensity score weighted | |||||
|---|---|---|---|---|---|---|---|
| Appropriate empirical antibiotics (n = 42) | Inappropriate empirical antibiotics (n = 113) | P-value | Appropriate empirical antibiotics (n = 33.4) | Inappropriate empirical antibiotics (n = 107.8) | P-value | ||
| Organisms | |||||||
| Escherichia coli | 39 (92.9) | 94 (83.2) | 0.125 | 29.7 (89) | 92.3 (85.6) | 1.000 | |
| Klebsiella pneumoniae | 3 (7.1) | 13 (11.5) | 0.560 | 3.69 (11.1) | 10.43 (9.7) | 0.743 | |
| Proteus mirabilis | 0 (0.0) | 3 (2.7) | 0.563 | 0 (0.0) | 2.54 (2.4) | 1.000 | |
| Salmonella spp. | 0 (0.0) | 1 (0.9) | 1.000 | 0 (0.0) | 0.8 (0.7) | 1.000 | |
| Citrobacter freundii | 0 (0.0) | 2 (1.8) | 1.000 | 0 (0.0) | 1.73 (1.6) | 1.000 | |
| Interventions | |||||||
| Time to empirical antibiotics (hours) | 1 (0.15 - 2) | 1 (0.25 - 3.25) | 0.139 | 0.9 (0.08 - 2.05) | 0.75 (0.25 - 3) | 0.275 | |
| Time to appropriate antibiotics (hours) | 1.1 (0.5 - 3) | 55 (46 - 72) | 0.0003 | 1.42 (0.5 - 3.1) | 55.2 (46.5 - 72) | <0.0001 | |
| Surgical drainage | 7 (16.7) | 36 (31.9) | 0.060 | 6.7 (20.1) | 30.7 (28.5) | 0.351 | |
| Clinical outcomes | |||||||
| 30-day mortality | 6 (14.3) | 14 (12.4) | 0.754 | 5 (14.9) | 14.3 (13.3) | 0.762 | |
| Length of hospital stay | 12 (9 - 18) | 11 (8 - 16) | 0.300 | 11 (7 - 15) | 12 (8 - 17) | 0.616 | |
| In-hospital mortality | 8 (19) | 17 (15) | 0.547 | 5.4 (16.0) | 17.3 (16.1) | 0.919 | |
Data are n (%), or median (interquartile range).
Sensitivity analysis which excluded 8 participants in the inappropriate antibiotics group never receiving definite appropriate antibiotics found the 30-day mortality of the appropriate group and inappropriate group was 14.9% and 10.1%, respectively. The adjusted difference in mortality between the two groups was 4.8% (95% CI: -7.8 - 17.4%).
Factors associated with 30-day mortality are shown in Table 4. In univariate analysis, cirrhosis, septic shock, acute respiratory failure, and primary bacteremia were significantly associated with mortality. In multivariate analyses, acute respiratory failure and diagnosis of primary bacteremia were associated with mortality. Inappropriate antibiotics use was not an independent risk factor for mortality.
Table 4. Results of the univariate and multivariate analysis of factors associated with 30-day mortality.
| Characteristics | Univariate analysis | Multivariate analysis | ||||
|---|---|---|---|---|---|---|
| Non-survivors (n = 20) | Survivors (n = 135) | P-value | Adjusted RR | P-value | ||
| Male | 11 (55) | 79 (58.5) | 0.766 | |||
| Age ≥75 years | 11 (55) | 63 (46.7) | 0.486 | |||
| Escherichia coli bacteremia | 18 (90) | 115 (85.2) | 0.741 | |||
| Underlying diseases | 20 (100) | 131 (97) | 1.000 | |||
| Charlson comorbidity index score ≥2 | 13 (65) | 73 (54.1) | 0.359 | |||
| Diabetes mellitus | 10 (50) | 51 (37.8) | 0.296 | |||
| Cirrhosis | 3 (15) | 4 (3) | 0.046 | 1.7 (0.3 - 10.3) | 0.578 | |
| Chronic kidney disease | 4 (20) | 27 (20) | 1.000 | |||
| Ischemic heart disease | 0 (0) | 18 (13.3) | 0.131 | |||
| Autoimmune disease | 0 (0) | 4 (3) | 1.000 | |||
| Cerebrovascular disease | 3 (15) | 23 (17) | 1.000 | |||
| Dementia | 1 (5) | 8 (5.9) | 1.000 | |||
| Solid malignancy | 6 (30) | 23 (17) | 0.165 | |||
| Hematological malignancy | 1 (5) | 3 (2.2) | 0.428 | |||
| Clinical syndromes | ||||||
| Urinary tract infection | 8 (40) | 80 (59.3) | 0.105 | |||
| Biliary tract infection | 2 (10) | 33 (24.4) | 0.149 | |||
| Other intraabdominal infections | 2 (10) | 6 (4.4) | 0.275 | |||
| Skin and soft tissue infection | 1 (5) | 4 (3) | 0.504 | |||
| Primary bacteremia | 6 (30) | 5 (3.7) | <0.0001 | 14.2 (3.1 - 65.3) | 0.0007 | |
| Others | 1 (5) | 7 (5.2) | 1.000 | |||
| Complications | ||||||
| Septic shock | 11 (55) | 33 (24.4) | 0.008 | 1.1 (0.3 - 4.4) | 0.881 | |
| Respiratory failure | 13 (65) | 24 (17.8) | <0.0001 | 9.9 (3.2 - 30.7) | <0.0001 | |
| Interventions | ||||||
| Appropriate empirical antibiotics | 6 (30) | 36 (26.7) | 0.754 | 0.95 (0.3 - 3.1) | 0.931 | |
| Surgical drainage | 3 (15) | 40 (29.6) | 0.283 | |||
Data are n (%).
Discussion
The results of this prospective cohort study indicated that inappropriate empirical antibiotics when given within 24 hours after the onset of bacteremia were non-inferior to appropriate antibiotics in terms of 30-day mortality, after adjusting for confounding factors using propensity score. Also, no differences were noted in length of hospital stay and hospital mortality between groups.
The overall 30-day mortality in this study (12.9%) was lower than that of recent studies for 3GCRE bacteremia reporting the incidence at 20 -30% [21,22]. This may have been due to the lower virulence of community-acquired 3GCRE infection, as the two previous studies of community-onset 3GCRE bacteremia also found a lower incidence at 5-10% [9,10]. Although the participants in this study had fewer severity factors, we included participants who died before receiving definite antibiotics which had been excluded from the previous studies [9,10].
Inappropriate antibiotics use has been shown as a strong risk factor for mortality and increased length of hospital stay among patients mainly with hospital-acquired 3GRCE infections [7,8]. On the contrary, this study showed that for patients with community-acquired 3GCRE bacteremia, inappropriate empirical antibiotics was non-inferior to appropriate antibiotics use. This is consistent with two previous studies in patients with community-onset bacteremia, including those with healthcare-associated infections. This may be explained by three reasons. First, the strains and virulence factors between hospital- and community-onset 3GCRE may be different, especially in patients with intraabdominal infections. For example, biliary tract obstruction found in 23% of participants with cholangitis was commonly thought to be the cause of bacteremia, not bacterial virulence [23]. This is in accordance with the previous 3GCRE carriage study in 188 patients from Sweden traveling to four regions with high 3GCRE prevalence, including Thailand [16]. The study showed that most isolates belonged to phylogroup A which were rarely associated with extraintestinal infections. No clinical infections were observed after 26 months of follow-up. Therefore, the community-acquired 3GCRE in the previous cohort showed a seemingly lower pathogenicity. Secondly, although the majority of participants presented extraintestinal infections presumably as a result of organisms with a higher pathogenicity; most were urinary tract infections. Urinary tract infections are well-known to be associated with lower mortality rates than infections from other anatomical sites [24]. Lastly, non-medical interventions such as the removal of infected foci or decompression of obstructions are essential aspects to treat infections and this was carried out promptly in 28% of patients in our cohort [25,26].
Factors associated with mortality were consistent with those previously described [9,10,27]. After adjusting these factors, inappropriate empirical antibiotics use was not associated with increased mortality. The increased use of empirical carbapenems in response to increased prevalence of 3GCRE infections may be accompanied by rapid emergence of carbapenem resistant pathogens. Empirical carbapenem in community-acquired infection, even in the setting of high prevalence of 3GCRE carriage in community, should be discouraged.
We describe the prospective cohort study to demonstrate the non-inferiority of inappropriate empirical antibiotics on the outcome of community-acquired 3GCRE bacteremia. However, our study encountered certain limitations. Firstly, this was an observational study, as it is impossible to design and enroll patients for randomized clinical trials due to the long turnaround time in the reporting of microbiological results. Therefore, unaccounted biases might have underestimated the true relationship between inappropriate antimicrobial therapy and mortality. For example, the prescribing physicians may have been affected by patients at a higher risk of death resulting in increased use of broader spectrum antibiotics. However, propensity score weighting was specifically used to reduce this selection bias and baseline characteristics between the two groups were not significantly dissimilar. Secondly, we did not demonstrate the mechanisms of resistance, strain, or virulent factors associated with the causative pathogens. Although this approach closely reflects real-world clinical practice, these findings may not be generalizable to settings with a different epidemiology of 3GCRE. Thirdly, this study was conducted in a large university hospital with a microbiological result alert system. The culture result turnaround time was relatively short, allowing the adjustment to appropriate antibiotics at the median time of only 55 hours in the inappropriate group. Therefore, the study results may not be generalizable to hospitals with longer microbiological result turnaround time. Finally, most patients presented low-inoculum infections such as urinary tract infections or those in which a reduced inoculum could be achieved through surgical intervention rather than high risk infections such as primary bacteremia or pneumonia. Therefore, whether inappropriate treatment is associated with higher mortality in these high-risk infections could not be concluded.
In summary, among patients with community-acquired 3GCRE bacteremia, inappropriate empirical treatment when given within 24 hours after the onset of bacteremia was non-inferior to appropriate antibiotics use. In the setting of a high prevalence of community 3GCRE carriage, adjustment to carbapenem might be tolerable for acute community-acquired infections.
ACKNOWLEDGMENT
We thank the Division of Microbiology, Department of Pathology, Phramongkutklao Hospital for bacterial isolation and standard antimicrobial susceptibility test.
Footnotes
Funding: This study was supported by the Division of Infectious Diseases, Department of Internal Medicine, Phramongkutklao Hospital (Funding number IDPMK 01301-61). The funder supported budget regarding the patient data collection process and all research equipment expenses, e.g., office supplies.
Conflict of Interest: No conflict of interest.
- Conceptualization: WN, DC, VV.
- Data curation: WN, VV.
- Formal analysis: WN, VV.
- Funding acquisition: WN.
- Investigation: WN, DC, VV.
- Methodology: WN, DC, VV.
- Project administration: W.
- Resources: WN, DC, VV.
- Software: WN, VV.
- Supervision: WN, DC, VV.
- Validation: WN, VV.
- Visualization: WN, DC, VV.
- Writing - original draft: WN, VV.
- Writing - review & editing: WN, DC, VV.
References
- 1.Belley A, Morrissey I, Hawser S, Kothari N, Knechtle P. Third-generation cephalosporin resistance in clinical isolates of Enterobacterales collected between 2016-2018 from USA and Europe: genotypic analysis of β-lactamases and comparative in vitro activity of cefepime/enmetazobactam. J Glob Antimicrob Resist. 2021;25:93–101. doi: 10.1016/j.jgar.2021.02.031. [DOI] [PubMed] [Google Scholar]
- 2.Ling W, Furuya-Kanamori L, Ezure Y, Harris PNA, Paterson DL. Adverse clinical outcomes associated with infections by Enterobacterales producing ESBL (ESBL-E): a systematic review and meta-analysis. JAC Antimicrob Resist. 2021;3:dlab068. doi: 10.1093/jacamr/dlab068. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Kim B, Seo MR, Kim J, Kim Y, Wie SH, Ki M, Cho YK, Lim S, Lee JS, Kwon KT, Lee H, Cheong HJ, Park DW, Ryu SY, Chung MH, Pai H. Molecular epidemiology of ciprofloxacin-resistant Escherichia coli isolated from community-acquired urinary tract infections in Korea. Infect Chemother. 2020;52:194–203. doi: 10.3947/ic.2020.52.2.194. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Mohd Sazlly Lim S, Wong PL, Sulaiman H, Atiya N, Hisham Shunmugam R, Liew SM. Clinical prediction models for ESBL-Enterobacteriaceae colonization or infection: a systematic review. J Hosp Infect. 2019;102:8–16. doi: 10.1016/j.jhin.2019.01.012. [DOI] [PubMed] [Google Scholar]
- 5.Bezabih YM, Sabiiti W, Alamneh E, Bezabih A, Peterson GM, Bezabhe WM, Roujeinikova A. The global prevalence and trend of human intestinal carriage of ESBL-producing Escherichia coli in the community. J Antimicrob Chemother. 2021;76:22–29. doi: 10.1093/jac/dkaa399. [DOI] [PubMed] [Google Scholar]
- 6.Clemenceau M, Ahmed-Elie S, Vilfaillot A, Chocron R, Compain F, Lebeaux D, Grohs P. Appropriateness of empirical antibiotic prescription for bloodstream infections in an emergency department from 2006 to 2018: impact of the spread of ESBL-producing Enterobacterales. Eur J Clin Microbiol Infect Dis. 2022;41:71–77. doi: 10.1007/s10096-021-04351-x. [DOI] [PubMed] [Google Scholar]
- 7.Lodise TP, Zhao Q, Fahrbach K, Gillard PJ, Martin A. A systematic review of the association between delayed appropriate therapy and mortality among patients hospitalized with infections due to Klebsiella pneumoniae or Escherichia coli: how long is too long? BMC Infect Dis. 2018;18:625. doi: 10.1186/s12879-018-3524-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Battle SE, Bookstaver PB, Justo JA, Kohn J, Albrecht H, Al-Hasan MN. Association between inappropriate empirical antimicrobial therapy and hospital length of stay in Gram-negative bloodstream infections: stratification by prognosis. J Antimicrob Chemother. 2017;72:299–304. doi: 10.1093/jac/dkw402. [DOI] [PubMed] [Google Scholar]
- 9.Joo EJ, Park DA, Lee NR, Moon SY, Choi JK, Ko JH, Peck KR. Impact of appropriateness of empiric therapy on outcomes in community-onset bacteremia by extended-spectrum-β-lactamase producing Escherichia coli and Klebisella pneumoniae definitively treated with carbapenems. Eur J Clin Microbiol Infect Dis. 2017;36:2093–2100. doi: 10.1007/s10096-017-3031-7. [DOI] [PubMed] [Google Scholar]
- 10.Kang CI, Wi YM, Ko KS, Chung DR, Peck KR, Lee NY, Song JH. Outcomes and risk factors for mortality in community-onset bacteremia caused by extended-spectrum beta-lactamase-producing Escherichia coli, with a special emphasis on antimicrobial therapy. Scand J Infect Dis. 2013;45:519–525. doi: 10.3109/00365548.2013.775479. [DOI] [PubMed] [Google Scholar]
- 11.Joo EJ, Kim SJ, Baek M, Choi Y, Seo J, Yeom JS, Ko KS. Fecal Carriage of antimicrobial-resistant Enterobacteriaceae in healthy Korean adults. J Microbiol Biotechnol. 2018;28:1178–1184. doi: 10.4014/jmb.1801.12060. [DOI] [PubMed] [Google Scholar]
- 12.Gutiérrez-Gutiérrez B, Rodríguez-Baño J. Current options for the treatment of infections due to extended-spectrum beta-lactamase-producing Enterobacteriaceae in different groups of patients. Clin Microbiol Infect. 2019;25:932–942. doi: 10.1016/j.cmi.2019.03.030. [DOI] [PubMed] [Google Scholar]
- 13.McLaughlin M, Advincula MR, Malczynski M, Qi C, Bolon M, Scheetz MH. Correlations of antibiotic use and carbapenem resistance in enterobacteriaceae. Antimicrob Agents Chemother. 2013;57:5131–5133. doi: 10.1128/AAC.00607-13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Thamlikitkul V, Tangkoskul T, Seenama C. Fecal carriage rate of extended-spectrum beta-lactamase-producing Enterobacteriaceae as a proxy composite indicator of antimicrobial resistance in a community in Thailand. Open Forum Infect Dis. 2019;6:ofz425. doi: 10.1093/ofid/ofz425. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Yoon YK, Kwon KT, Jeong SJ, Moon C, Kim B, Kiem S, Kim HS, Heo E, Kim SW Korean Society for Antimicrobial Therapy; Korean Society of Infectious Diseases; Korean Society of Health-System Pharmacist. Guidelines on implementing antimicrobial stewardship programs in Korea. Infect Chemother. 2021;53:617–659. doi: 10.3947/ic.2021.0098. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Ny S, Löfmark S, Börjesson S, Englund S, Ringman M, Bergström J, Nauclér P, Giske CG, Byfors S. Community carriage of ESBL-producing Escherichia coli is associated with strains of low pathogenicity: a Swedish nationwide study. J Antimicrob Chemother. 2017;72:582–588. doi: 10.1093/jac/dkw419. [DOI] [PubMed] [Google Scholar]
- 17.Friedman ND, Kaye KS, Stout JE, McGarry SA, Trivette SL, Briggs JP, Lamm W, Clark C, MacFarquhar J, Walton AL, Reller LB, Sexton DJ. Health care--associated bloodstream infections in adults: a reason to change the accepted definition of community-acquired infections. Ann Intern Med. 2002;137:791–797. doi: 10.7326/0003-4819-137-10-200211190-00007. [DOI] [PubMed] [Google Scholar]
- 18.Khwaja A. KDIGO clinical practice guidelines for acute kidney injury. Nephron Clin Pract. 2012;120:c179–c184. doi: 10.1159/000339789. [DOI] [PubMed] [Google Scholar]
- 19.Taylor FB, Jr, Toh CH, Hoots WK, Wada H, Levi M Scientific Subcommittee on Disseminated Intravascular Coagulation (DIC) of the International Society on Thrombosis and Haemostasis (ISTH) Towards definition, clinical and laboratory criteria, and a scoring system for disseminated intravascular coagulation. Thromb Haemost. 2001;86:1327–1330. [PubMed] [Google Scholar]
- 20.Clinical and Laboratory Standards Institute (CLSI) CLSI supplement M100. 28th ed. Wayne, PA: CLSI; 2018. Performance standards for antimicrobial susceptibility testing. [Google Scholar]
- 21.Lim CL, Spelman D. Mortality impact of empirical antimicrobial therapy in ESBL- and AmpC-producing Enterobacteriaceae bacteremia in an Australian tertiary hospital. Infect Dis Health. 2019;24:124–133. doi: 10.1016/j.idh.2019.02.001. [DOI] [PubMed] [Google Scholar]
- 22.Sianipar O, Asmara W, Dwiprahasto I, Mulyono B. Mortality risk of bloodstream infection caused by either Escherichia coli or Klebsiella pneumoniae producing extended-spectrum β-lactamase: a prospective cohort study. BMC Res Notes. 2019;12:719. doi: 10.1186/s13104-019-4751-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Melzer M, Toner R, Lacey S, Bettany E, Rait G. Biliary tract infection and bacteraemia: presentation, structural abnormalities, causative organisms and clinical outcomes. Postgrad Med J. 2007;83:773–776. doi: 10.1136/pgmj.2007.064683. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Abe T, Ogura H, Kushimoto S, Shiraishi A, Sugiyama T, Deshpande GA, Uchida M, Nagata I, Saitoh D, Fujishima S, Mayumi T, Hifumi T, Shiino Y, Nakada TA, Tarui T, Otomo Y, Okamoto K, Umemura Y, Kotani J, Sakamoto Y, Sasaki J, Shiraishi SI, Takuma K, Tsuruta R, Hagiwara A, Yamakawa K, Masuno T, Takeyama N, Yamashita N, Ikeda H, Ueyama M, Fujimi S, Gando S JAAM FORECAST group. Variations in infection sites and mortality rates among patients in intensive care units with severe sepsis and septic shock in Japan. J Intensive Care. 2019;7:28. doi: 10.1186/s40560-019-0383-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Kang CI, Chung DR, Ko KS, Peck KR, Song JH Korean Network for the Study of Infectious Diseases (KONSID) Risk factors for mortality and impact of broad-spectrum cephalosporin resistance on outcome in bacteraemic intra-abdominal infections caused by Gram-negative bacilli. Scand J Infect Dis. 2011;43:202–208. doi: 10.3109/00365548.2010.539257. [DOI] [PubMed] [Google Scholar]
- 26.Kang CI, Kim SH, Park WB, Lee KD, Kim HB, Kim EC, Oh MD, Choe KW. Bloodstream infections due to extended-spectrum beta-lactamase-producing Escherichia coli and Klebsiella pneumoniae: risk factors for mortality and treatment outcome, with special emphasis on antimicrobial therapy. Antimicrob Agents Chemother. 2004;48:4574–4581. doi: 10.1128/AAC.48.12.4574-4581.2004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Tumbarello M, Sanguinetti M, Montuori E, Trecarichi EM, Posteraro B, Fiori B, Citton R, D’Inzeo T, Fadda G, Cauda R, Spanu T. Predictors of mortality in patients with bloodstream infections caused by extended-spectrum-beta-lactamase-producing Enterobacteriaceae: importance of inadequate initial antimicrobial treatment. Antimicrob Agents Chemother. 2007;51:1987–1994. doi: 10.1128/AAC.01509-06. [DOI] [PMC free article] [PubMed] [Google Scholar]