Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Mortality associated with carbapenem-susceptible and Verona Integron-encoded Metallo-β-lactamase-positive Pseudomonas aeruginosa bacteremia

Mortality associated with carbapenem-susceptible and Verona Integron-encoded... Background: Studies on various Gram-negative bacteria suggest that resistance to carbapenem antibioticsisresponsible for increased mortality in patients; however, results are not conclusive. We first assessed the 28-day in-hospital all-cause mortality in patients with Verona Integron-encoded Metallo-β-lactamase-positive Pseudomonas aeruginosa (VIM-PA) bacteremia compared to patients with VIM-negative, carbapenem-susceptible P. aeruginosa (CS-PA) bacteremia. Second, we identified determinants for mortality and survival. Methods: All patients with a positive blood culture with VIM-PA or CS-PA between January 2004 and January 2016 were included. Kaplan-Meier survival curves were constructed, and survivors and non-survivors were compared on relevant clinical parameters using univariate analyses, and multivariable analyses using a Cox-proportional hazard model. Results: In total, 249 patients were included, of which 58 (23.3%) died. Seventeen out of 40 (42.5%) patients with VIM-PA died, compared to 41 out of 209 (19.6%) patients with CS-PA (difference = 22.9%, P-value = 0.001). Assumed acquisition of the bacterium at the intensive care unit was significantly associated with mortality (HR = 3.32, 95%CI = 1.60–6.87), and having had adequate antibiotic therapy in days 1–14 after the positive blood culture was identified as a determinant for survival (HR = 0.03, 95%CI = 0.01–0.06). VIM-PA vs CS-PA was not identified as an independent risk factor for mortality. Conclusions: The crude mortality rate was significantly higher in patients with a VIM-PA bacteremia compared to patients with a CS-PA bacteremia; however, when analyzing the data in a multivariable model this difference was non-significant. Awareness of the presence of P. aeruginosa in the hospital environment that may be transmitted to patients and rapid microbiological diagnostics are essential for timely administration of appropriate antibiotics. Acquisition of P. aeruginosa should be prevented, independent of resistance profile. Keywords: Pseudomonas aeruginosa, Mortality, Bacteremia, Anti-bacterial agents, Carbapenemase, Intensive care units, Beta- lactamase VIM * Correspondence: j.severin@erasmusmc.nl Marjolein C. Persoon and Anne F. Voor in’t holt contributed equally to this work. Department of Medical Microbiology and Infectious Diseases, Erasmus MC University Medical Center, Rotterdam, The Netherlands Full list of author information is available at the end of the article © The Author(s). 2020 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Persoon et al. Antimicrobial Resistance and Infection Control (2020) 9:25 Page 2 of 8 Introduction organized into 48 different wards [14]. Of these wards, Studies suggest that infections with microorganisms re- three were high-level adult ICUs, and each had only sistant to carbapenem antibiotics may be responsible for single-occupancy rooms. Since 2008 ICU patients increased mortality in patients compared to infections received selective digestive tract decontamination if with susceptible microorganisms [1, 2]. However, results expected to be on mechanical ventilation for > 48 h or are not conclusive and evidence is limited; attributable anticipated to be admitted to the ICU for > 72 h [15]. mortality is therefore still the subject of ongoing investi- The number of clinical admissions and clinical admis- gations [3]. sion days for the study period are described in a previ- Carbapenem resistance in Pseudomonas aeruginosa iso- ous publication [5]. lates is an emerging problem, with nosocomial outbreaks of this microorganism occurring worldwide, also in the Patient inclusion and data collection Netherlands [4–6]. In P. aeruginosa isolates, carbapenem Patients were included if identified with a positive blood resistance may be caused by a variety of mechanisms. culture with P. aeruginosa between January 1st 2004 and However, of particular concern is the production of January 1st 2016. Only the first isolate of each patient metallo-beta-lactamase (MBL) enzymes catalyzing the hy- was included. We excluded the following patients: (i) pa- drolysis of all classes of beta-lactam antibiotics except tients < 18 years old, (ii) non-hospitalized patients, (iii) if monobactams [7]. Currently, the Verona Integron- the first positive blood culture contained more than one encoded MBL (VIM) is the most widespread MBL in P. pathogenic microorganism, and (iv) patients identified aeruginosa, with VIM-2 being the major source of with a carbapenem-resistant bla -negative P. aerugi- VIM worldwide outbreaks [8]. Infections (e.g. bacteremia) with nosa isolate. Positive blood cultures with both P. aerugi- VIM-positive P. aeruginosa (VIM-PA) are mainly seen in nosa and coagulase-negative staphylococci (CoNS) were patients with a suppressed immune system or otherwise included when the CoNS were considered to be contam- compromised, for example in patients admitted to the in- ination. These cultures were excluded if the CoNS was tensive care unit (ICU) [9]. Published mortality rates for cultured in a second blood culture and if the CoNS was bloodstream infections with P. aeruginosa are between 20 considered clinically relevant (e.g. antibiotic treatment and 50%, with the rate of carbapenem-resistant P. aerugi- was started or other actions were performed). nosa-attributable deaths among those patients being be- We collected the following patient data from the elec- tween 8 and 18% [2, 10]. However, the question still tronic patient records: (i) age at day of first positive remains if mortality in these patients is high because of 1) blood culture with P. aeruginosa, (ii) sex, (iii) date of first inadequate or delayed appropriate antibiotic therapy, 2) positive blood culture with P. aeruginosa and suscepti- severity of underlying disease, or 3) because of the VIM bility pattern, (iv) department of acquisition, defined as gene and the subsequent carbapenem resistance [1, 9, 11, the department at which the patient was admitted 48 h 12]. Moreover, clones with and without carbapenemase before the positive blood culture: if the patient was posi- genes can also differ in virulence characteristics, which tive at admittance this was scored as ‘at home’ or ‘at could lead to differences in mortality rates between pa- other healthcare facility’, (v) date of admission to and tients with different clones; independent of the resistant date of discharge from our hospital, (vi) death 28 days mechanism [13]. after first positive blood culture yes/no and date of First, we aimed first to compare the 28-day in-hospital all- death, (vii) nosocomial infection, defined as a positive cause mortality in patients with a VIM-PA bacteremia to pa- blood culture with P. aeruginosa > 48 h after admission, tients with a bacteremia with VIM-negative, carbapenem- (viii) neutropenia, defined as an absolute neutrophil susceptible P. aeruginosa (CS-PA) in a large tertiary hospital count of < 0.5 X10 cells/L during presentation of in the Netherlands. Second, we aimed to identify risk factors bacteremia; neutrophil count was assessed 2 days before for mortality by P. aeruginosa bacteremia in this setting. the first positive blood culture was taken to 7 days after- wards, (ix) use of corticosteroids 28 days before until 28 Methods days after the positive blood culture with P. aeruginosa, Ethics statement (x) use of other immunosuppressive agents than cortico- Written approval to conduct this study was received steroids in 28 days before until 28 days after positive from the medical ethical research committee from the blood culture with P. aeruginosa, and (xi) antibiotic use. Erasmus MC University Medical Center (Erasmus MC), During analyses, viii, ix and x were combined and re- Rotterdam, the Netherlands (MEC-2015-306). ported as being immunocompromised yes or no. Use of one or more of the following antibiotics was Setting checked from the day the first positive blood culture was This study was conducted at the Erasmus MC. During taken to 14 days thereafter: piperacillin/tazobactam, gen- the study period, this was a 1200-bed university hospital tamicin, tobramycin, ceftazidime, ciprofloxacin, colistin Persoon et al. Antimicrobial Resistance and Infection Control (2020) 9:25 Page 3 of 8 and carbapenems. These antibiotics have antipseudomo- Statistical analysis nal activity and are part of our local antibiotic policy. Pi- To calculate differences in mortality, crude mortality of peracillin without tazobactam is not used in our hospital. patients with CS-PA was subtracted from crude mortal- Three definitions of adequate antibiotic use were applied. ity of patients with VIM-PA. Kaplan-Meier survival Antibiotic use at the moment the first positive blood cul- curves were constructed for these 2 groups for 28-day ture was drawn was defined as adequate when administra- in-hospital survival. The log rank test was performed to tion of at least one antibiotic agent for which the P. statistically compare the two curves. Univariate analyses aeruginosa isolate is susceptible was given for at least 24 h in order to compare survivors to non-survivors were (adequate antibiotic use 1). In the 2 weeks after the blood conducted using a t test, the independent samples me- culture was drawn, antibiotic use was defined as adequate dian test or the Mann-Whitney U test when appropriate. when at least one administered antibiotic agent for which All analyses were performed using SPSS version 24 (IBM the P. aeruginosa isolate is susceptible was administered Corp., Armonk, New York, USA). Concerning the multi- for at least 24 h (adequate antibiotic use 2). Patients who variable analysis: because some patients were discharged deceased within 24 h after the first positive blood culture and readmitted within 28 days, a start-stop-event Cox- were excluded from analyses with variable adequate anti- proportional hazard model was fitted using the R project biotic use 2. As third definition, we combined adequate for statistical computing, version 3.4.3 (Vienna, Austria) antibiotic use 1 and adequate antibiotic use 2 (adequate to calculate hazard ratios (HRs) for 28-day in-hospital antibiotic use total). mortality. The selection of variables in the multivariable For all patients, we calculated the Charlson comorbid- model was based on clinical relevance and results of the ity score at admission and for patients admitted to the univariate analysis, with age and sex as standard parame- ICU we calculated the Acute Physiology And Chronic ters included. To test if the model improved when Health Evaluation (APACHE) score within 24 h after ad- relevant interaction terms were added, models were mission to the ICU [16, 17]. The Charlson comorbidity compared using the ANOVA statistic (analysis of devi- score is a method used to predict mortality by assigning ance). P-values of < 0.05 were considered statistically sig- different weights to comorbidities [17]. We used an up- nificant, and a 95% confidence interval (CI) was used. dated Charlson comorbidity index which was validated and published by Quan et al. in 2011 [17]. Results The primary endpoint of this study was 28-day in- Patient inclusion and characteristics hospital all-cause mortality. Patients were followed until Between 2004 and 2016, 249 patients with a P. aeruginosa 1) in-hospital death up until 28 days after the first posi- positive blood culture were included in this study (Fig. 1), tive blood culture, 2) hospital discharge within 28 days of which 40 (16.1%) were identified with VIM-PA, and after the first positive blood culture, or 3) until day 28 if 209 (83.9%) with CS-PA. In blood cultures of three pa- still admitted by that time. Readmissions within 28 days tients, a CoNS was present in a single blood culture, which after the first positive blood culture were also considered was considered to be contamination (i.e. n =1 Staphylo- and data was used for analyses. coccus epidermidis,n=1 S. hominis,and n =1 Staphylo- coccus species). In another patient, a CoNS was found in a Microbiological methods second blood culture together with a P. aeruginosa,how- Blood cultures taken on clinical indication were proc- ever, the two blood cultures taken on the same day were essed in the laboratory using standard microbiological negative, as well as a blood culture taken a few days later. methods (BACTEC system BD). Identification and sus- Furthermore, no antibiotics were started for this CoNS. ceptibility testing of Gram-negative aerobic bacilli were Therefore, this was also considered to be contamination. performed using Vitek2 (bioMérieux, Marcy l’Etoile, Overall, 159 male (63.9%), and 90 (36.1%) female patients France). Since January 2013, the MALDI-TOF (Bruker were included, with a mean age of 59.5 years old. Regard- Daltonics, Bremen, Germany) was used for identifica- ing age and sex, there were only small, non-significant dif- tion. Breakpoints were in accordance to Clinical and La- ferences between survivors and non-survivors (Table 1). boratory Standards Institute (CLSI) guidelines until Twenty-eight patients (11.2%) died within 48 h after the August 27th 2013, thereafter European Committee on positive blood culture, 11 with a VIM-PA bacteremia Antimicrobial Susceptibility Testing (EUCAST) guide- (4.4% total, 27.5% of patients identified with VIM-PA), lines were used. In case of suspected growth of and 17 with a CS-PA bacteremia (6.8% total, 8.1% of pa- carbapenemase-producing P. aeruginosa or multidrug- tients identified with CS-PA) (P-value < 0.001). Of these resistant P. aeruginosa, an in-house polymerase chain re- patients, 4 with a VIM-PA bacteremia (1.6% total, 10.0% action (PCR) for detection of bla on LightCycler 480 of patients identified with VIM-PA), and also 4 patients VIM (Roche Diagnostics, Almere, The Netherlands) was per- with a CS-PA bacteremia (1.6% total, 1.9% of patients formed as previously described [18, 19]. identified with CS-PA) died within 24 h (P-value = 0.008). Persoon et al. Antimicrobial Resistance and Infection Control (2020) 9:25 Page 4 of 8 Fig. 1 Flow diagram of patient inclusion. Abbreviations: CoNS, coagulase-negative staphylococci; VIM-PA, Verona Integron-encoded Metallo-β- lactamase-positive Pseudomonas aeruginosa; CS-PA, VIM-negative, carbapenem-susceptible Pseudomonas aeruginosa For these 8 patients, analysis on adequate antibiotic use 2 range 0–20). This was sooner compared to patients with could not be performed. CS-PA (median = 5 days, range 0–25). Univariate ana- lyses showed that mortality was associated with a noso- 28-day in-hospital all-cause mortality comial infection, having VIM-PA instead of CS-PA, Fifty-eight (23.3%) out of 249 patients died within 28 acquisition of the bacterium in the ICU, and inadequate days in the Erasmus MC, and 191 patients (76.7%) sur- antibiotic therapy in total and at days 1–15 after the vived during follow-up. Seventeen out of 40 (42.5%) pa- positive blood culture (adequate AB therapy 2) (Table tients with VIM-PA died within 28 days in the Erasmus 1). The Charlson score was not associated with mortal- MC, compared to 41 out of 209 (19.6%) patients with ity, as was the APACHE score on admission to the ICU CS-PA. Therefore, the difference in crude mortality rate (Table 1). Between patients identified with a VIM-PA or was 22.9%. Additionally, the Kaplan-Meier survival curve CS-PA there were no significant differences regarding visualizes the difference between patients identified with the Charlson or the APACHE score. VIM-PA compared to CS-PA over time (Log Rank P- In the multivariable model, the following variables value = 0.001) (Fig. 2). Patients with VIM-PA deceased were included: (i) sex, (ii) age, (iii) VIM-PA vs. CS-PA, shortly after the positive blood culture (median = 2 days, (iv) ICU as ward of acquisition, (v) nosocomial Persoon et al. Antimicrobial Resistance and Infection Control (2020) 9:25 Page 5 of 8 Table 1 Patient-related clinical variables of survivors and non-survivors Characteristic Non-survivors Survivors P-value (n = 58) (n = 191) Male gender (%) 38 (65.5) 121 (63.4) 0.764 Mean age ± SD at time of first blood culture with PA 60.5 ± 12.9 59.3 ± 14.1 0.539 Nosocomial infection (%) 45 (77.6) 113 (59.2) 0.011 VIM-PA bacteremia (%) 17 (29.3) 23 (12.0) 0.002 Nosocomial infection (%) 17 (100) 19 (82.6) 0.070 Adequate AB therapy 1 (%) 5 (31.3) 8 (42.1) 0.818 Adequate AB therapy 2 (%) 5 (38.5) 23 (100) < 0.001 c,d Median Charlson score(range) 2(0–6) 3 (0–10) 0.639 Ward of acquisition; ICU (%) 29 (50) 32 (16.8) < 0.001 Median APACHE score (range) 22 (10–43) 22.5 (7–41) 0.757 APACHE > 25 (%) 6 (11.8) 7 (35.0) 0.588 Immunocompromised 29 (50.0) 84 (44.0) 0.420 g,h Adequate AB therapy total (%) 31 (63.3) 157 (98.1) < 0.001 Adequate AB therapy 1 (%) 17 (30.9) 52 (27.8) 0.654 h j Adequate AB therapy 2 (%) 26 (52.0) 156 (97.5) < 0.001 Abbreviations: SD Standard deviation, PA Pseudomonas aeruginosa, VIM-PA Verona Integron-encoded Metallo-β-lactamase-positive Pseudomonas aeruginosa, APACHE Acute Physiology and Chronic Health Evaluation, ICU Intensive care unit, AB Antibiotic; adequate AB therapy 1, day 0 for at least 24 h adequate AB use; adequate AB therapy 2, days 1–14 adequate AB use, for at least 24 h. Bold P-values are significant a b c d non-survivors = 1 missing. Four patients were excluded because they deceased within 24 h. Quan et al. [17]. Charlson: survivors = 1 missing, non-survivors = 1 e f missing. APACHE: survivors = 12 missing, non-survivors = 7 missing. Combination of variables neutropenia; use of corticosteroids and use of immunosuppressive g h agents other than corticosteroids. Adequate AB total: survivors = 31 missing, non-survivors = 1 missing. Eight patients from the non-survivors were excluded i j because they deceased within 24 h. Adequate AB 1: survivors = 4 missing, non-survivors = 3 missing. Adequate AB 2: survivors = 31 missing Fig. 2 Kaplan Meier curve of patients with VIM-negative, carbapenem-susceptible Pseudomonas aeruginosa (dark-grey line) and patients with VIM- positive Pseudomonas aeruginosa (light grey line). Y-axis, cumulative survival, X-axis, days. Log Rank P- value = 0.001 Persoon et al. Antimicrobial Resistance and Infection Control (2020) 9:25 Page 6 of 8 acquisition of P. aeruginosa, (vi) adequate antibiotic use 1 analysis we did not identify any differences between sur- and (vii) adequate antibiotic use 2. Figure 3 shows the re- vivors and non-survivors, and between patients with sults of the final multivariable model, in which acquisition VIM-PA and CS-PA regarding the Charlson score at in the ICU was identified as a statistically significant risk hospital admission or APACHE score on admission to factor for mortality in all patients (HR = 3.32, 95%CI = the ICU. The Charlson comorbidity score is obtained 1.60 to 6.87). Having had adequate antibiotic therapy in with data from time of admission and may therefore not days 1–14 after the positive blood culture (adequate anti- represent the severity of disease during admission, espe- biotic use 2) was identified as a determinant for survival cially not in a tertiary care hospital where patients need (HR = 0.03, 95%CI = 0.01 to 0.06) (Fig. 3). Although and receive high level care; this includes broad-spectrum significantly associated in univariate analyses, VIM-PA vs. antibiotics, several medical devices (e.g. mechanical ven- CS-PA was not identified as an independent risk factor for tilation, central venous catheters) and close monitoring mortality, thus after correcting for all other variables by physicians and nurses. This makes these patients present in the model. more vulnerable for acquisition and infection by micro- organisms, which could lead to mortality. Further- Discussion more, acquisition of P. aeruginosa in the hospital The 28-day in-hospital all-cause mortality in patients could be responsible for deterioration of the patients’ with a bacteremia with VIM-PA was significantly higher clinical condition. compared to patients with a bacteremia with CS-PA, Patients with VIM-PA deceased shortly after the posi- with a difference in crude mortality rate of more than tive blood culture, which was sooner compared to pa- 22%. In the systematic review published by Zhang et al. tients with CS-PA. A possible explanation could be that the attributable mortality ranged from 8 to 18.4%, with 4 more than half of patients with VIM-PA acquired this studies included from Brazil, Spain, Korea and the U.S. bacterium in the ICU, compared to 17.2% of patients [2]. Thus, in our center in the Netherlands the rate is with CS-PA. Acquisition in the ICU may therefore be an high. Despite the difference in crude mortality between indication of increased severity of disease at the moment VIM-PA and CS-PA, in the multivariable model VIM- of acquisition. A second possible explanation could be PA was not identified to be significantly associated with differences in receiving timely adequate antibiotic ther- mortality. This was also reported by Peña et al. They apy. Adequate antibiotic use 2 (i.e. days 1–14 after the described that the effect of resistance on mortality de- positive blood culture for at least 24 h) was identified as creased with higher Charlson scores, i.e. the effect disap- a protective factor, which means that this decreased risk peared in the presence of comorbidities [1]. We did of mortality. Survivors were two times more likely to re- identify that acquisition of P. aeruginosa in the ICU was ceive correct antibiotic therapy than non-survivors significantly associated with mortality, which could be (97.5% vs. 52.0%, respectively). This is also described by related to disease severity. However, in the univariate González et al., Paulsson et al., DiMondi et al., Peña Fig. 3 Multivariable analysis to identify determinants significantly related to mortality after Pseudomonas aeruginosa bacteremia. Abbreviations: ICU, intensive care unit; AB, antibiotic; HR, hazard ratio; 95%CI, 95% confidence interval; VIM-PA, Verona Integron-encoded Metallo-β-lactamase- positive Pseudomonas aeruginosa; adequate AB therapy 1, day 0 for at least 24 h adequate AB use; adequate AB therapy 2, days 1–14 adequate AB use, for at least 24 h Persoon et al. Antimicrobial Resistance and Infection Control (2020) 9:25 Page 7 of 8 et al. and Raman et al. [20–24]. This big difference and ICU in a vulnerable ICU patient group, 51.7 and 31.3% very low percentage of correct antibiotic use in non- respectively. Additionally, patients deceased rapidly after survivors could be largely explained by the time of death: acquiring Pseudomonas aeruginosa. Despite the higher 78% of patients died between 24 and 48 h after the first crude mortality rate for VIM-PA, our study showed that positive blood culture. In some cases, treatment with for in our population bacteremia with CS-PA were equally example meropenem was started, however not given for important regarding clinical outcome compared to VIM- > 24 h before the patient died. In other cases, an incor- PA, since multivariable analyses showed no difference rect antibiotic was given e.g. for which the identified P. between these two groups. Infections with CS-PA should aeruginosa was resistant, or no antibiotics were started. therefore not be underestimated. Adequate antibiotic Since VIM-PA in our patient group was found to be therapy for VIM-PA as well as for CS-PA has shown to resistant for several antibiotic classes, it can be hypothe- be a determinant for surviving. Therefore, since VIM-PA sized that patients with a VIM-PA bacteremia have an are resistant to many antibiotic groups and consequently increased risk of receiving inadequate empirical anti- more difficult to treat, awareness of the presence of biotic therapy. However, in our patient group there was VIM-PA in the hospital environment and rapid micro- no difference in receiving adequate therapy 1, 2 or total biological diagnostics and sensitivity analysis are essen- between patients with VIM-PA and CS-PA. A possible tial for timely administration of adequate antibiotics. explanation is that most patients with VIM-PA were in Acquisition of P. aeruginosa should be avoided and pre- the ICU when having a VIM-PA bacteremia, and in 52% vented, in particular in the ICU, independent of resist- VIM-PA was acquired in the ICU. At the ICU, patients ance profile. To elucidate the sole role of VIM we are monitored very closely and screening cultures of propose to perform a multi-center study in different various sample-sites are taken regularly. Therefore, countries with a design matched on life expectancy. colonization of VIM-PA in a patient may be detected be- Abbreviations fore infection occurs, facilitating starting adequate ther- APACHE: The Acute Physiology And Chronic Health Evaluation score; apy when indicated. When comparing patients with CI: Confidence interval; CLSI: Clinical and Laboratory Standards Institute; CoNS: Coagulase-negative staphylococci; CS-PA: Carbapenem-susceptible P. acquisition of P. aeruginosa in the ICU to patients with aeruginosa; Erasmus MC: Erasmus MC University Medical Center; acquisition in non-ICU wards, there was a difference in EUCAST: European Committee on Antimicrobial Susceptibility Testing; receiving adequate antibiotic therapy 1 (ICU, 36.7%; ICU: Intensive care unit; MBL: Metallo-beta-lactamases; PCR: Polymerase chain reaction; VIM: Verona Integron-encoded MBL; VIM-PA: VIM-positive P. non-ICU, 25.8%), however this was not statistically sig- aeruginosa nificant (P- value = 0.107). Acknowledgements We hereby thank the infection prevention practitioners and the intensive Limitations and strengths care staff of the Erasmus MC Rotterdam, The Netherlands for their help in Our study has some limitations. First, patient groups the data collection process. with VIM-PA and CS-PA were not matched. Ideally, Authors’ contributions matching should have been done on life-expectancy Conceived and designed the study: MV, JS, AV, MP. Collecting data: MP, AV. and/or severity of illness on the day of positive blood Analyzed the data: MP, AV. Interpretation of the data: MP, AV, CW, DG, MV, culture. Second, this is a retrospective study conducted JS. Drafted the work: MP, AV, JS, MV. All authors read and approved the final manuscript. in a single tertiary care hospital in the Netherlands; therefore, the results may not be generalizable to other Funding institutions and/or countries. However, we feel that a No grants or external funding was received for this work. single center study design is preferred above a multi- Availability of data and materials center when studying the sole effect of VIM presence The datasets generated and analyzed during the current study are not publicly above other patient- and care-related risk factors, as the available due to privacy regulations but are available from the corresponding author on reasonable request. latter differ considerably between centers. Third, because of the low total number of deceased patients (n = 58), Ethics approval and consent to participate small but true differences may have been missed. Written approval to conduct this study was received from the medical ethical research committee from the Erasmus MC University Medical Center (Erasmus A strength of our study is that despite it being a single MC), Rotterdam, the Netherlands (MEC-2015-306). center study, a large group of patients could be included. Second, we focused on solely the VIM gene. Consent for publication Not applicable. Conclusions and implications Competing interests The crude mortality rate was significantly higher in pa- The authors declare that they have no competing interests. tients with a VIM-PA bacteremia compared to patients Author details with a CS-PA bacteremia in our university hospital. Ac- 1 Department of Medical Microbiology and Infectious Diseases, Erasmus MC quisition of VIM-PA and CS-PA mainly occurred in the University Medical Center, Rotterdam, The Netherlands. National Institute for Persoon et al. Antimicrobial Resistance and Infection Control (2020) 9:25 Page 8 of 8 Public Health and the Environment, Bilthoven, The Netherlands. Department aeruginosa in the Netherlands: the nationwide emergence of a single of Adult Intensive Care, Erasmus MC University Medical Center, Rotterdam, sequence type. Clin Microbiol Infect. 2012;18(9):E369–72. The Netherlands. 19. Van der Bij AK, Van Mansfeld R, Peirano G, Goessens WH, Severin JA, Pitout JD, et al. First outbreak of VIM-2 metallo-beta-lactamase-producing Received: 11 October 2019 Accepted: 21 January 2020 Pseudomonas aeruginosa in the Netherlands: microbiology, epidemiology and clinical outcomes. Int J Antimicrob Agents. 2011;37(6):513–8. 20. Raman G, Avendano E, Berger S, Menon V. Appropriate initial antibiotic therapy in hospitalized patients with gram-negative infections: systematic References review and meta-analysis. BMC Infect Dis. 2015;15:395. 1. Pena C, Suarez C, Gozalo M, Murillas J, Almirante B, Pomar V, et al. 21. DiMondi VP, Townsend ML, Drew RH. Risk factors associated with Prospective multicenter study of the impact of carbapenem resistance on unfavorable short-term treatment outcome in patients with documented mortality in Pseudomonas aeruginosa bloodstream infections. Antimicrob Pseudomonas aeruginosa infection. Int J Clin Pharmacol. 2015;37(2):348–54. Agents Chemother. 2012;56(3):1265–72. 22. Paulsson M, Granrot A, Ahl J, Tham J, Resman F, Riesbeck K, et al. 2. Zhang Y, Chen XL, Huang AW, Liu SL, Liu WJ, Zhang N, et al. Mortality Antimicrobial combination treatment including ciprofloxacin decreased the attributable to carbapenem-resistant Pseudomonas aeruginosa bacteremia: a mortality rate of Pseudomonas aeruginosa bacteraemia: a retrospective meta-analysis of cohort studies. Emerg Microbes Infect. 2016;5:e27. cohort study. Eur J Clin Microbiol Infect Dis. 2017;36(7):1187–96. 3. Suarez C, Pena C, Gavalda L, Tubau F, Manzur A, Dominguez MA, et al. 23. Pena C, Suarez C, Ocampo-Sosa A, Murillas J, Almirante B, Pomar V, et al. Influence of carbapenem resistance on mortality and the dynamics of Effect of adequate single-drug vs combination antimicrobial therapy on mortality in Pseudomonas aeruginosa bloodstream infection. Int J Infect Dis. mortality in Pseudomonas aeruginosa bloodstream infections: a post hoc 2010;14(Suppl 3):e73–8. analysis of a prospective cohort. Clin Infect Dis. 2013;57(2):208–16. 4. Voor in't holt AF, Severin JA, Lesaffre EM, Vos MC. A systematic review and 24. Gonzalez AL, Leal AL, Cortes JA, Sanchez R, Barrero LI, Castillo JS, et al. meta-analyses show that carbapenem use and medical devices are the [Effect of adequate initial antimicrobial therapy on mortality in critical leading risk factors for carbapenem-resistant Pseudomonas aeruginosa. patients with Pseudomonas aeruginosa bacteremia] Efecto del tratamiento Antimicrob Agents Chemother. 2014;58(5):2626–37. antibiotico inicial adecuado sobre la mortalidad en pacientes en estado 5. Voor in't holt AF, Severin JA, MBH H, de Goeij I, Gommers D, Vos MC. VIM- critico con bacteriemia por Pseudomonas aeruginosa. Biomedica. 2014; positive Pseudomonas aeruginosa in a large tertiary care hospital: matched 34(Suppl 1):58–66. case-control studies and a network analysis. Antimicrob Resist Infect Control. 2018;7:32. 6. Knoester M, de Boer MG, Maarleveld JJ, Claas EC, Bernards AT, de Jonge E, Publisher’sNote et al. An integrated approach to control a prolonged outbreak of Springer Nature remains neutral with regard to jurisdictional claims in multidrug-resistant Pseudomonas aeruginosa in an intensive care unit. Clin published maps and institutional affiliations. Microbiol Infect. 2014;20(4):O207–15. 7. Makena A, Duzgun AO, Brem J, McDonough MA, Rydzik AM, Abboud MI, et al. Comparison of Verona Integron-borne Metallo-beta-lactamase (VIM) variants reveals differences in stability and inhibition profiles. Antimicrob Agents Chemother. 2015;60(3):1377–84. 8. Hong DJ, Bae IK, Jang IH, Jeong SH, Kang HK, Lee K. Epidemiology and characteristics of Metallo-beta-lactamase-producing Pseudomonas aeruginosa. Infect Chemother. 2015;47(2):81–97. 9. Persoon MC, Voor In't Holt AF, van Meer MPA, Bokhoven KC, Gommers D, Vos MC, et al. Mortality related to Verona Integron-encoded Metallo-beta- lactamase-positive Pseudomonas aeruginosa: assessment by a novel clinical tool. Antimicrob Resist Infect Control. 2019;8:107. 10. Zavascki AP, Barth AL, Goncalves AL, Moro AL, Fernandes JF, Martins AF, et al. The influence of metallo-beta-lactamase production on mortality in nosocomial Pseudomonas aeruginosa infections. J Antimicrob Chemother. 2006;58(2):387–92. 11. Dantas RC, Ferreira ML, Gontijo-Filho PP, Ribas RM. Pseudomonas aeruginosa bacteraemia: independent risk factors for mortality and impact of resistance on outcome. J Med Microbiol. 2014;63(Pt 12):1679–87. 12. Villegas MV, Pallares CJ, Escandon-Vargas K, Hernandez-Gomez C, Correa A, Alvarez C, et al. Characterization and clinical impact of bloodstream infection caused by Carbapenemase-producing Enterobacteriaceae in seven Latin American countries. PLoS One. 2016;11(4):e0154092. 13. Recio R, Villa J, Viedma E, Orellana MA, Lora-Tamayo J, Chaves F. Bacteraemia due to extensively drug-resistant Pseudomonas aeruginosa sequence type 235 high-risk clone: facing the perfect storm. Int J Antimicrob Agents. 2018;52(2):172–9. 14. Ministry of health welfare and sport. Jaarverantwoording in de zorg - Annual reporting by healthcare institutions: Ministry of Health Welfare and Sport; [Available from: https://www.desan.nl/net/DoSearch/Search.aspx. 15. de Smet AM, Kluytmans JA, Cooper BS, Mascini EM, Benus RF, van der Werf TS, et al. Decontamination of the digestive tract and oropharynx in ICU patients. N Engl J Med. 2009;360(1):20–31. 16. MDCalc. APACHE II Score - estimates ICU mortality:MDCalc; [Available from: http://www.mdcalc.com/apache-ii-score/. 17. Quan H, Li B, Couris CM, Fushimi K, Graham P, Hider P, et al. Updating and validating the Charlson comorbidity index and score for risk adjustment in hospital discharge abstracts using data from 6 countries. Am J Epidemiol. 2011;173(6):676–82. 18. Van der Bij AK, Van der Zwan D, Peirano G, Severin JA, Pitout JD, Van Westreenen M, et al. Metallo-beta-lactamase-producing Pseudomonas http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Antimicrobial Resistance & Infection Control Springer Journals

Mortality associated with carbapenem-susceptible and Verona Integron-encoded Metallo-β-lactamase-positive Pseudomonas aeruginosa bacteremia

Download PDF
8 pages

Loading next page...
 
/lp/springer-journals/mortality-associated-with-carbapenem-susceptible-and-verona-integron-034HgVJ64y

References (28)

Publisher
Springer Journals
Copyright
Copyright © The Author(s). 2020
eISSN
2047-2994
DOI
10.1186/s13756-020-0682-4
Publisher site
See Article on Publisher Site

Abstract

Background: Studies on various Gram-negative bacteria suggest that resistance to carbapenem antibioticsisresponsible for increased mortality in patients; however, results are not conclusive. We first assessed the 28-day in-hospital all-cause mortality in patients with Verona Integron-encoded Metallo-β-lactamase-positive Pseudomonas aeruginosa (VIM-PA) bacteremia compared to patients with VIM-negative, carbapenem-susceptible P. aeruginosa (CS-PA) bacteremia. Second, we identified determinants for mortality and survival. Methods: All patients with a positive blood culture with VIM-PA or CS-PA between January 2004 and January 2016 were included. Kaplan-Meier survival curves were constructed, and survivors and non-survivors were compared on relevant clinical parameters using univariate analyses, and multivariable analyses using a Cox-proportional hazard model. Results: In total, 249 patients were included, of which 58 (23.3%) died. Seventeen out of 40 (42.5%) patients with VIM-PA died, compared to 41 out of 209 (19.6%) patients with CS-PA (difference = 22.9%, P-value = 0.001). Assumed acquisition of the bacterium at the intensive care unit was significantly associated with mortality (HR = 3.32, 95%CI = 1.60–6.87), and having had adequate antibiotic therapy in days 1–14 after the positive blood culture was identified as a determinant for survival (HR = 0.03, 95%CI = 0.01–0.06). VIM-PA vs CS-PA was not identified as an independent risk factor for mortality. Conclusions: The crude mortality rate was significantly higher in patients with a VIM-PA bacteremia compared to patients with a CS-PA bacteremia; however, when analyzing the data in a multivariable model this difference was non-significant. Awareness of the presence of P. aeruginosa in the hospital environment that may be transmitted to patients and rapid microbiological diagnostics are essential for timely administration of appropriate antibiotics. Acquisition of P. aeruginosa should be prevented, independent of resistance profile. Keywords: Pseudomonas aeruginosa, Mortality, Bacteremia, Anti-bacterial agents, Carbapenemase, Intensive care units, Beta- lactamase VIM * Correspondence: j.severin@erasmusmc.nl Marjolein C. Persoon and Anne F. Voor in’t holt contributed equally to this work. Department of Medical Microbiology and Infectious Diseases, Erasmus MC University Medical Center, Rotterdam, The Netherlands Full list of author information is available at the end of the article © The Author(s). 2020 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Persoon et al. Antimicrobial Resistance and Infection Control (2020) 9:25 Page 2 of 8 Introduction organized into 48 different wards [14]. Of these wards, Studies suggest that infections with microorganisms re- three were high-level adult ICUs, and each had only sistant to carbapenem antibiotics may be responsible for single-occupancy rooms. Since 2008 ICU patients increased mortality in patients compared to infections received selective digestive tract decontamination if with susceptible microorganisms [1, 2]. However, results expected to be on mechanical ventilation for > 48 h or are not conclusive and evidence is limited; attributable anticipated to be admitted to the ICU for > 72 h [15]. mortality is therefore still the subject of ongoing investi- The number of clinical admissions and clinical admis- gations [3]. sion days for the study period are described in a previ- Carbapenem resistance in Pseudomonas aeruginosa iso- ous publication [5]. lates is an emerging problem, with nosocomial outbreaks of this microorganism occurring worldwide, also in the Patient inclusion and data collection Netherlands [4–6]. In P. aeruginosa isolates, carbapenem Patients were included if identified with a positive blood resistance may be caused by a variety of mechanisms. culture with P. aeruginosa between January 1st 2004 and However, of particular concern is the production of January 1st 2016. Only the first isolate of each patient metallo-beta-lactamase (MBL) enzymes catalyzing the hy- was included. We excluded the following patients: (i) pa- drolysis of all classes of beta-lactam antibiotics except tients < 18 years old, (ii) non-hospitalized patients, (iii) if monobactams [7]. Currently, the Verona Integron- the first positive blood culture contained more than one encoded MBL (VIM) is the most widespread MBL in P. pathogenic microorganism, and (iv) patients identified aeruginosa, with VIM-2 being the major source of with a carbapenem-resistant bla -negative P. aerugi- VIM worldwide outbreaks [8]. Infections (e.g. bacteremia) with nosa isolate. Positive blood cultures with both P. aerugi- VIM-positive P. aeruginosa (VIM-PA) are mainly seen in nosa and coagulase-negative staphylococci (CoNS) were patients with a suppressed immune system or otherwise included when the CoNS were considered to be contam- compromised, for example in patients admitted to the in- ination. These cultures were excluded if the CoNS was tensive care unit (ICU) [9]. Published mortality rates for cultured in a second blood culture and if the CoNS was bloodstream infections with P. aeruginosa are between 20 considered clinically relevant (e.g. antibiotic treatment and 50%, with the rate of carbapenem-resistant P. aerugi- was started or other actions were performed). nosa-attributable deaths among those patients being be- We collected the following patient data from the elec- tween 8 and 18% [2, 10]. However, the question still tronic patient records: (i) age at day of first positive remains if mortality in these patients is high because of 1) blood culture with P. aeruginosa, (ii) sex, (iii) date of first inadequate or delayed appropriate antibiotic therapy, 2) positive blood culture with P. aeruginosa and suscepti- severity of underlying disease, or 3) because of the VIM bility pattern, (iv) department of acquisition, defined as gene and the subsequent carbapenem resistance [1, 9, 11, the department at which the patient was admitted 48 h 12]. Moreover, clones with and without carbapenemase before the positive blood culture: if the patient was posi- genes can also differ in virulence characteristics, which tive at admittance this was scored as ‘at home’ or ‘at could lead to differences in mortality rates between pa- other healthcare facility’, (v) date of admission to and tients with different clones; independent of the resistant date of discharge from our hospital, (vi) death 28 days mechanism [13]. after first positive blood culture yes/no and date of First, we aimed first to compare the 28-day in-hospital all- death, (vii) nosocomial infection, defined as a positive cause mortality in patients with a VIM-PA bacteremia to pa- blood culture with P. aeruginosa > 48 h after admission, tients with a bacteremia with VIM-negative, carbapenem- (viii) neutropenia, defined as an absolute neutrophil susceptible P. aeruginosa (CS-PA) in a large tertiary hospital count of < 0.5 X10 cells/L during presentation of in the Netherlands. Second, we aimed to identify risk factors bacteremia; neutrophil count was assessed 2 days before for mortality by P. aeruginosa bacteremia in this setting. the first positive blood culture was taken to 7 days after- wards, (ix) use of corticosteroids 28 days before until 28 Methods days after the positive blood culture with P. aeruginosa, Ethics statement (x) use of other immunosuppressive agents than cortico- Written approval to conduct this study was received steroids in 28 days before until 28 days after positive from the medical ethical research committee from the blood culture with P. aeruginosa, and (xi) antibiotic use. Erasmus MC University Medical Center (Erasmus MC), During analyses, viii, ix and x were combined and re- Rotterdam, the Netherlands (MEC-2015-306). ported as being immunocompromised yes or no. Use of one or more of the following antibiotics was Setting checked from the day the first positive blood culture was This study was conducted at the Erasmus MC. During taken to 14 days thereafter: piperacillin/tazobactam, gen- the study period, this was a 1200-bed university hospital tamicin, tobramycin, ceftazidime, ciprofloxacin, colistin Persoon et al. Antimicrobial Resistance and Infection Control (2020) 9:25 Page 3 of 8 and carbapenems. These antibiotics have antipseudomo- Statistical analysis nal activity and are part of our local antibiotic policy. Pi- To calculate differences in mortality, crude mortality of peracillin without tazobactam is not used in our hospital. patients with CS-PA was subtracted from crude mortal- Three definitions of adequate antibiotic use were applied. ity of patients with VIM-PA. Kaplan-Meier survival Antibiotic use at the moment the first positive blood cul- curves were constructed for these 2 groups for 28-day ture was drawn was defined as adequate when administra- in-hospital survival. The log rank test was performed to tion of at least one antibiotic agent for which the P. statistically compare the two curves. Univariate analyses aeruginosa isolate is susceptible was given for at least 24 h in order to compare survivors to non-survivors were (adequate antibiotic use 1). In the 2 weeks after the blood conducted using a t test, the independent samples me- culture was drawn, antibiotic use was defined as adequate dian test or the Mann-Whitney U test when appropriate. when at least one administered antibiotic agent for which All analyses were performed using SPSS version 24 (IBM the P. aeruginosa isolate is susceptible was administered Corp., Armonk, New York, USA). Concerning the multi- for at least 24 h (adequate antibiotic use 2). Patients who variable analysis: because some patients were discharged deceased within 24 h after the first positive blood culture and readmitted within 28 days, a start-stop-event Cox- were excluded from analyses with variable adequate anti- proportional hazard model was fitted using the R project biotic use 2. As third definition, we combined adequate for statistical computing, version 3.4.3 (Vienna, Austria) antibiotic use 1 and adequate antibiotic use 2 (adequate to calculate hazard ratios (HRs) for 28-day in-hospital antibiotic use total). mortality. The selection of variables in the multivariable For all patients, we calculated the Charlson comorbid- model was based on clinical relevance and results of the ity score at admission and for patients admitted to the univariate analysis, with age and sex as standard parame- ICU we calculated the Acute Physiology And Chronic ters included. To test if the model improved when Health Evaluation (APACHE) score within 24 h after ad- relevant interaction terms were added, models were mission to the ICU [16, 17]. The Charlson comorbidity compared using the ANOVA statistic (analysis of devi- score is a method used to predict mortality by assigning ance). P-values of < 0.05 were considered statistically sig- different weights to comorbidities [17]. We used an up- nificant, and a 95% confidence interval (CI) was used. dated Charlson comorbidity index which was validated and published by Quan et al. in 2011 [17]. Results The primary endpoint of this study was 28-day in- Patient inclusion and characteristics hospital all-cause mortality. Patients were followed until Between 2004 and 2016, 249 patients with a P. aeruginosa 1) in-hospital death up until 28 days after the first posi- positive blood culture were included in this study (Fig. 1), tive blood culture, 2) hospital discharge within 28 days of which 40 (16.1%) were identified with VIM-PA, and after the first positive blood culture, or 3) until day 28 if 209 (83.9%) with CS-PA. In blood cultures of three pa- still admitted by that time. Readmissions within 28 days tients, a CoNS was present in a single blood culture, which after the first positive blood culture were also considered was considered to be contamination (i.e. n =1 Staphylo- and data was used for analyses. coccus epidermidis,n=1 S. hominis,and n =1 Staphylo- coccus species). In another patient, a CoNS was found in a Microbiological methods second blood culture together with a P. aeruginosa,how- Blood cultures taken on clinical indication were proc- ever, the two blood cultures taken on the same day were essed in the laboratory using standard microbiological negative, as well as a blood culture taken a few days later. methods (BACTEC system BD). Identification and sus- Furthermore, no antibiotics were started for this CoNS. ceptibility testing of Gram-negative aerobic bacilli were Therefore, this was also considered to be contamination. performed using Vitek2 (bioMérieux, Marcy l’Etoile, Overall, 159 male (63.9%), and 90 (36.1%) female patients France). Since January 2013, the MALDI-TOF (Bruker were included, with a mean age of 59.5 years old. Regard- Daltonics, Bremen, Germany) was used for identifica- ing age and sex, there were only small, non-significant dif- tion. Breakpoints were in accordance to Clinical and La- ferences between survivors and non-survivors (Table 1). boratory Standards Institute (CLSI) guidelines until Twenty-eight patients (11.2%) died within 48 h after the August 27th 2013, thereafter European Committee on positive blood culture, 11 with a VIM-PA bacteremia Antimicrobial Susceptibility Testing (EUCAST) guide- (4.4% total, 27.5% of patients identified with VIM-PA), lines were used. In case of suspected growth of and 17 with a CS-PA bacteremia (6.8% total, 8.1% of pa- carbapenemase-producing P. aeruginosa or multidrug- tients identified with CS-PA) (P-value < 0.001). Of these resistant P. aeruginosa, an in-house polymerase chain re- patients, 4 with a VIM-PA bacteremia (1.6% total, 10.0% action (PCR) for detection of bla on LightCycler 480 of patients identified with VIM-PA), and also 4 patients VIM (Roche Diagnostics, Almere, The Netherlands) was per- with a CS-PA bacteremia (1.6% total, 1.9% of patients formed as previously described [18, 19]. identified with CS-PA) died within 24 h (P-value = 0.008). Persoon et al. Antimicrobial Resistance and Infection Control (2020) 9:25 Page 4 of 8 Fig. 1 Flow diagram of patient inclusion. Abbreviations: CoNS, coagulase-negative staphylococci; VIM-PA, Verona Integron-encoded Metallo-β- lactamase-positive Pseudomonas aeruginosa; CS-PA, VIM-negative, carbapenem-susceptible Pseudomonas aeruginosa For these 8 patients, analysis on adequate antibiotic use 2 range 0–20). This was sooner compared to patients with could not be performed. CS-PA (median = 5 days, range 0–25). Univariate ana- lyses showed that mortality was associated with a noso- 28-day in-hospital all-cause mortality comial infection, having VIM-PA instead of CS-PA, Fifty-eight (23.3%) out of 249 patients died within 28 acquisition of the bacterium in the ICU, and inadequate days in the Erasmus MC, and 191 patients (76.7%) sur- antibiotic therapy in total and at days 1–15 after the vived during follow-up. Seventeen out of 40 (42.5%) pa- positive blood culture (adequate AB therapy 2) (Table tients with VIM-PA died within 28 days in the Erasmus 1). The Charlson score was not associated with mortal- MC, compared to 41 out of 209 (19.6%) patients with ity, as was the APACHE score on admission to the ICU CS-PA. Therefore, the difference in crude mortality rate (Table 1). Between patients identified with a VIM-PA or was 22.9%. Additionally, the Kaplan-Meier survival curve CS-PA there were no significant differences regarding visualizes the difference between patients identified with the Charlson or the APACHE score. VIM-PA compared to CS-PA over time (Log Rank P- In the multivariable model, the following variables value = 0.001) (Fig. 2). Patients with VIM-PA deceased were included: (i) sex, (ii) age, (iii) VIM-PA vs. CS-PA, shortly after the positive blood culture (median = 2 days, (iv) ICU as ward of acquisition, (v) nosocomial Persoon et al. Antimicrobial Resistance and Infection Control (2020) 9:25 Page 5 of 8 Table 1 Patient-related clinical variables of survivors and non-survivors Characteristic Non-survivors Survivors P-value (n = 58) (n = 191) Male gender (%) 38 (65.5) 121 (63.4) 0.764 Mean age ± SD at time of first blood culture with PA 60.5 ± 12.9 59.3 ± 14.1 0.539 Nosocomial infection (%) 45 (77.6) 113 (59.2) 0.011 VIM-PA bacteremia (%) 17 (29.3) 23 (12.0) 0.002 Nosocomial infection (%) 17 (100) 19 (82.6) 0.070 Adequate AB therapy 1 (%) 5 (31.3) 8 (42.1) 0.818 Adequate AB therapy 2 (%) 5 (38.5) 23 (100) < 0.001 c,d Median Charlson score(range) 2(0–6) 3 (0–10) 0.639 Ward of acquisition; ICU (%) 29 (50) 32 (16.8) < 0.001 Median APACHE score (range) 22 (10–43) 22.5 (7–41) 0.757 APACHE > 25 (%) 6 (11.8) 7 (35.0) 0.588 Immunocompromised 29 (50.0) 84 (44.0) 0.420 g,h Adequate AB therapy total (%) 31 (63.3) 157 (98.1) < 0.001 Adequate AB therapy 1 (%) 17 (30.9) 52 (27.8) 0.654 h j Adequate AB therapy 2 (%) 26 (52.0) 156 (97.5) < 0.001 Abbreviations: SD Standard deviation, PA Pseudomonas aeruginosa, VIM-PA Verona Integron-encoded Metallo-β-lactamase-positive Pseudomonas aeruginosa, APACHE Acute Physiology and Chronic Health Evaluation, ICU Intensive care unit, AB Antibiotic; adequate AB therapy 1, day 0 for at least 24 h adequate AB use; adequate AB therapy 2, days 1–14 adequate AB use, for at least 24 h. Bold P-values are significant a b c d non-survivors = 1 missing. Four patients were excluded because they deceased within 24 h. Quan et al. [17]. Charlson: survivors = 1 missing, non-survivors = 1 e f missing. APACHE: survivors = 12 missing, non-survivors = 7 missing. Combination of variables neutropenia; use of corticosteroids and use of immunosuppressive g h agents other than corticosteroids. Adequate AB total: survivors = 31 missing, non-survivors = 1 missing. Eight patients from the non-survivors were excluded i j because they deceased within 24 h. Adequate AB 1: survivors = 4 missing, non-survivors = 3 missing. Adequate AB 2: survivors = 31 missing Fig. 2 Kaplan Meier curve of patients with VIM-negative, carbapenem-susceptible Pseudomonas aeruginosa (dark-grey line) and patients with VIM- positive Pseudomonas aeruginosa (light grey line). Y-axis, cumulative survival, X-axis, days. Log Rank P- value = 0.001 Persoon et al. Antimicrobial Resistance and Infection Control (2020) 9:25 Page 6 of 8 acquisition of P. aeruginosa, (vi) adequate antibiotic use 1 analysis we did not identify any differences between sur- and (vii) adequate antibiotic use 2. Figure 3 shows the re- vivors and non-survivors, and between patients with sults of the final multivariable model, in which acquisition VIM-PA and CS-PA regarding the Charlson score at in the ICU was identified as a statistically significant risk hospital admission or APACHE score on admission to factor for mortality in all patients (HR = 3.32, 95%CI = the ICU. The Charlson comorbidity score is obtained 1.60 to 6.87). Having had adequate antibiotic therapy in with data from time of admission and may therefore not days 1–14 after the positive blood culture (adequate anti- represent the severity of disease during admission, espe- biotic use 2) was identified as a determinant for survival cially not in a tertiary care hospital where patients need (HR = 0.03, 95%CI = 0.01 to 0.06) (Fig. 3). Although and receive high level care; this includes broad-spectrum significantly associated in univariate analyses, VIM-PA vs. antibiotics, several medical devices (e.g. mechanical ven- CS-PA was not identified as an independent risk factor for tilation, central venous catheters) and close monitoring mortality, thus after correcting for all other variables by physicians and nurses. This makes these patients present in the model. more vulnerable for acquisition and infection by micro- organisms, which could lead to mortality. Further- Discussion more, acquisition of P. aeruginosa in the hospital The 28-day in-hospital all-cause mortality in patients could be responsible for deterioration of the patients’ with a bacteremia with VIM-PA was significantly higher clinical condition. compared to patients with a bacteremia with CS-PA, Patients with VIM-PA deceased shortly after the posi- with a difference in crude mortality rate of more than tive blood culture, which was sooner compared to pa- 22%. In the systematic review published by Zhang et al. tients with CS-PA. A possible explanation could be that the attributable mortality ranged from 8 to 18.4%, with 4 more than half of patients with VIM-PA acquired this studies included from Brazil, Spain, Korea and the U.S. bacterium in the ICU, compared to 17.2% of patients [2]. Thus, in our center in the Netherlands the rate is with CS-PA. Acquisition in the ICU may therefore be an high. Despite the difference in crude mortality between indication of increased severity of disease at the moment VIM-PA and CS-PA, in the multivariable model VIM- of acquisition. A second possible explanation could be PA was not identified to be significantly associated with differences in receiving timely adequate antibiotic ther- mortality. This was also reported by Peña et al. They apy. Adequate antibiotic use 2 (i.e. days 1–14 after the described that the effect of resistance on mortality de- positive blood culture for at least 24 h) was identified as creased with higher Charlson scores, i.e. the effect disap- a protective factor, which means that this decreased risk peared in the presence of comorbidities [1]. We did of mortality. Survivors were two times more likely to re- identify that acquisition of P. aeruginosa in the ICU was ceive correct antibiotic therapy than non-survivors significantly associated with mortality, which could be (97.5% vs. 52.0%, respectively). This is also described by related to disease severity. However, in the univariate González et al., Paulsson et al., DiMondi et al., Peña Fig. 3 Multivariable analysis to identify determinants significantly related to mortality after Pseudomonas aeruginosa bacteremia. Abbreviations: ICU, intensive care unit; AB, antibiotic; HR, hazard ratio; 95%CI, 95% confidence interval; VIM-PA, Verona Integron-encoded Metallo-β-lactamase- positive Pseudomonas aeruginosa; adequate AB therapy 1, day 0 for at least 24 h adequate AB use; adequate AB therapy 2, days 1–14 adequate AB use, for at least 24 h Persoon et al. Antimicrobial Resistance and Infection Control (2020) 9:25 Page 7 of 8 et al. and Raman et al. [20–24]. This big difference and ICU in a vulnerable ICU patient group, 51.7 and 31.3% very low percentage of correct antibiotic use in non- respectively. Additionally, patients deceased rapidly after survivors could be largely explained by the time of death: acquiring Pseudomonas aeruginosa. Despite the higher 78% of patients died between 24 and 48 h after the first crude mortality rate for VIM-PA, our study showed that positive blood culture. In some cases, treatment with for in our population bacteremia with CS-PA were equally example meropenem was started, however not given for important regarding clinical outcome compared to VIM- > 24 h before the patient died. In other cases, an incor- PA, since multivariable analyses showed no difference rect antibiotic was given e.g. for which the identified P. between these two groups. Infections with CS-PA should aeruginosa was resistant, or no antibiotics were started. therefore not be underestimated. Adequate antibiotic Since VIM-PA in our patient group was found to be therapy for VIM-PA as well as for CS-PA has shown to resistant for several antibiotic classes, it can be hypothe- be a determinant for surviving. Therefore, since VIM-PA sized that patients with a VIM-PA bacteremia have an are resistant to many antibiotic groups and consequently increased risk of receiving inadequate empirical anti- more difficult to treat, awareness of the presence of biotic therapy. However, in our patient group there was VIM-PA in the hospital environment and rapid micro- no difference in receiving adequate therapy 1, 2 or total biological diagnostics and sensitivity analysis are essen- between patients with VIM-PA and CS-PA. A possible tial for timely administration of adequate antibiotics. explanation is that most patients with VIM-PA were in Acquisition of P. aeruginosa should be avoided and pre- the ICU when having a VIM-PA bacteremia, and in 52% vented, in particular in the ICU, independent of resist- VIM-PA was acquired in the ICU. At the ICU, patients ance profile. To elucidate the sole role of VIM we are monitored very closely and screening cultures of propose to perform a multi-center study in different various sample-sites are taken regularly. Therefore, countries with a design matched on life expectancy. colonization of VIM-PA in a patient may be detected be- Abbreviations fore infection occurs, facilitating starting adequate ther- APACHE: The Acute Physiology And Chronic Health Evaluation score; apy when indicated. When comparing patients with CI: Confidence interval; CLSI: Clinical and Laboratory Standards Institute; CoNS: Coagulase-negative staphylococci; CS-PA: Carbapenem-susceptible P. acquisition of P. aeruginosa in the ICU to patients with aeruginosa; Erasmus MC: Erasmus MC University Medical Center; acquisition in non-ICU wards, there was a difference in EUCAST: European Committee on Antimicrobial Susceptibility Testing; receiving adequate antibiotic therapy 1 (ICU, 36.7%; ICU: Intensive care unit; MBL: Metallo-beta-lactamases; PCR: Polymerase chain reaction; VIM: Verona Integron-encoded MBL; VIM-PA: VIM-positive P. non-ICU, 25.8%), however this was not statistically sig- aeruginosa nificant (P- value = 0.107). Acknowledgements We hereby thank the infection prevention practitioners and the intensive Limitations and strengths care staff of the Erasmus MC Rotterdam, The Netherlands for their help in Our study has some limitations. First, patient groups the data collection process. with VIM-PA and CS-PA were not matched. Ideally, Authors’ contributions matching should have been done on life-expectancy Conceived and designed the study: MV, JS, AV, MP. Collecting data: MP, AV. and/or severity of illness on the day of positive blood Analyzed the data: MP, AV. Interpretation of the data: MP, AV, CW, DG, MV, culture. Second, this is a retrospective study conducted JS. Drafted the work: MP, AV, JS, MV. All authors read and approved the final manuscript. in a single tertiary care hospital in the Netherlands; therefore, the results may not be generalizable to other Funding institutions and/or countries. However, we feel that a No grants or external funding was received for this work. single center study design is preferred above a multi- Availability of data and materials center when studying the sole effect of VIM presence The datasets generated and analyzed during the current study are not publicly above other patient- and care-related risk factors, as the available due to privacy regulations but are available from the corresponding author on reasonable request. latter differ considerably between centers. Third, because of the low total number of deceased patients (n = 58), Ethics approval and consent to participate small but true differences may have been missed. Written approval to conduct this study was received from the medical ethical research committee from the Erasmus MC University Medical Center (Erasmus A strength of our study is that despite it being a single MC), Rotterdam, the Netherlands (MEC-2015-306). center study, a large group of patients could be included. Second, we focused on solely the VIM gene. Consent for publication Not applicable. Conclusions and implications Competing interests The crude mortality rate was significantly higher in pa- The authors declare that they have no competing interests. tients with a VIM-PA bacteremia compared to patients Author details with a CS-PA bacteremia in our university hospital. Ac- 1 Department of Medical Microbiology and Infectious Diseases, Erasmus MC quisition of VIM-PA and CS-PA mainly occurred in the University Medical Center, Rotterdam, The Netherlands. National Institute for Persoon et al. Antimicrobial Resistance and Infection Control (2020) 9:25 Page 8 of 8 Public Health and the Environment, Bilthoven, The Netherlands. Department aeruginosa in the Netherlands: the nationwide emergence of a single of Adult Intensive Care, Erasmus MC University Medical Center, Rotterdam, sequence type. Clin Microbiol Infect. 2012;18(9):E369–72. The Netherlands. 19. Van der Bij AK, Van Mansfeld R, Peirano G, Goessens WH, Severin JA, Pitout JD, et al. First outbreak of VIM-2 metallo-beta-lactamase-producing Received: 11 October 2019 Accepted: 21 January 2020 Pseudomonas aeruginosa in the Netherlands: microbiology, epidemiology and clinical outcomes. Int J Antimicrob Agents. 2011;37(6):513–8. 20. Raman G, Avendano E, Berger S, Menon V. Appropriate initial antibiotic therapy in hospitalized patients with gram-negative infections: systematic References review and meta-analysis. BMC Infect Dis. 2015;15:395. 1. Pena C, Suarez C, Gozalo M, Murillas J, Almirante B, Pomar V, et al. 21. DiMondi VP, Townsend ML, Drew RH. Risk factors associated with Prospective multicenter study of the impact of carbapenem resistance on unfavorable short-term treatment outcome in patients with documented mortality in Pseudomonas aeruginosa bloodstream infections. Antimicrob Pseudomonas aeruginosa infection. Int J Clin Pharmacol. 2015;37(2):348–54. Agents Chemother. 2012;56(3):1265–72. 22. Paulsson M, Granrot A, Ahl J, Tham J, Resman F, Riesbeck K, et al. 2. Zhang Y, Chen XL, Huang AW, Liu SL, Liu WJ, Zhang N, et al. Mortality Antimicrobial combination treatment including ciprofloxacin decreased the attributable to carbapenem-resistant Pseudomonas aeruginosa bacteremia: a mortality rate of Pseudomonas aeruginosa bacteraemia: a retrospective meta-analysis of cohort studies. Emerg Microbes Infect. 2016;5:e27. cohort study. Eur J Clin Microbiol Infect Dis. 2017;36(7):1187–96. 3. Suarez C, Pena C, Gavalda L, Tubau F, Manzur A, Dominguez MA, et al. 23. Pena C, Suarez C, Ocampo-Sosa A, Murillas J, Almirante B, Pomar V, et al. Influence of carbapenem resistance on mortality and the dynamics of Effect of adequate single-drug vs combination antimicrobial therapy on mortality in Pseudomonas aeruginosa bloodstream infection. Int J Infect Dis. mortality in Pseudomonas aeruginosa bloodstream infections: a post hoc 2010;14(Suppl 3):e73–8. analysis of a prospective cohort. Clin Infect Dis. 2013;57(2):208–16. 4. Voor in't holt AF, Severin JA, Lesaffre EM, Vos MC. A systematic review and 24. Gonzalez AL, Leal AL, Cortes JA, Sanchez R, Barrero LI, Castillo JS, et al. meta-analyses show that carbapenem use and medical devices are the [Effect of adequate initial antimicrobial therapy on mortality in critical leading risk factors for carbapenem-resistant Pseudomonas aeruginosa. patients with Pseudomonas aeruginosa bacteremia] Efecto del tratamiento Antimicrob Agents Chemother. 2014;58(5):2626–37. antibiotico inicial adecuado sobre la mortalidad en pacientes en estado 5. Voor in't holt AF, Severin JA, MBH H, de Goeij I, Gommers D, Vos MC. VIM- critico con bacteriemia por Pseudomonas aeruginosa. Biomedica. 2014; positive Pseudomonas aeruginosa in a large tertiary care hospital: matched 34(Suppl 1):58–66. case-control studies and a network analysis. Antimicrob Resist Infect Control. 2018;7:32. 6. Knoester M, de Boer MG, Maarleveld JJ, Claas EC, Bernards AT, de Jonge E, Publisher’sNote et al. An integrated approach to control a prolonged outbreak of Springer Nature remains neutral with regard to jurisdictional claims in multidrug-resistant Pseudomonas aeruginosa in an intensive care unit. Clin published maps and institutional affiliations. Microbiol Infect. 2014;20(4):O207–15. 7. Makena A, Duzgun AO, Brem J, McDonough MA, Rydzik AM, Abboud MI, et al. Comparison of Verona Integron-borne Metallo-beta-lactamase (VIM) variants reveals differences in stability and inhibition profiles. Antimicrob Agents Chemother. 2015;60(3):1377–84. 8. Hong DJ, Bae IK, Jang IH, Jeong SH, Kang HK, Lee K. Epidemiology and characteristics of Metallo-beta-lactamase-producing Pseudomonas aeruginosa. Infect Chemother. 2015;47(2):81–97. 9. Persoon MC, Voor In't Holt AF, van Meer MPA, Bokhoven KC, Gommers D, Vos MC, et al. Mortality related to Verona Integron-encoded Metallo-beta- lactamase-positive Pseudomonas aeruginosa: assessment by a novel clinical tool. Antimicrob Resist Infect Control. 2019;8:107. 10. Zavascki AP, Barth AL, Goncalves AL, Moro AL, Fernandes JF, Martins AF, et al. The influence of metallo-beta-lactamase production on mortality in nosocomial Pseudomonas aeruginosa infections. J Antimicrob Chemother. 2006;58(2):387–92. 11. Dantas RC, Ferreira ML, Gontijo-Filho PP, Ribas RM. Pseudomonas aeruginosa bacteraemia: independent risk factors for mortality and impact of resistance on outcome. J Med Microbiol. 2014;63(Pt 12):1679–87. 12. Villegas MV, Pallares CJ, Escandon-Vargas K, Hernandez-Gomez C, Correa A, Alvarez C, et al. Characterization and clinical impact of bloodstream infection caused by Carbapenemase-producing Enterobacteriaceae in seven Latin American countries. PLoS One. 2016;11(4):e0154092. 13. Recio R, Villa J, Viedma E, Orellana MA, Lora-Tamayo J, Chaves F. Bacteraemia due to extensively drug-resistant Pseudomonas aeruginosa sequence type 235 high-risk clone: facing the perfect storm. Int J Antimicrob Agents. 2018;52(2):172–9. 14. Ministry of health welfare and sport. Jaarverantwoording in de zorg - Annual reporting by healthcare institutions: Ministry of Health Welfare and Sport; [Available from: https://www.desan.nl/net/DoSearch/Search.aspx. 15. de Smet AM, Kluytmans JA, Cooper BS, Mascini EM, Benus RF, van der Werf TS, et al. Decontamination of the digestive tract and oropharynx in ICU patients. N Engl J Med. 2009;360(1):20–31. 16. MDCalc. APACHE II Score - estimates ICU mortality:MDCalc; [Available from: http://www.mdcalc.com/apache-ii-score/. 17. Quan H, Li B, Couris CM, Fushimi K, Graham P, Hider P, et al. Updating and validating the Charlson comorbidity index and score for risk adjustment in hospital discharge abstracts using data from 6 countries. Am J Epidemiol. 2011;173(6):676–82. 18. Van der Bij AK, Van der Zwan D, Peirano G, Severin JA, Pitout JD, Van Westreenen M, et al. Metallo-beta-lactamase-producing Pseudomonas

Journal

Antimicrobial Resistance & Infection Control – Springer Journals

Published: Feb 3, 2020

There are no references for this article.