Validation of quantitative real-time polymerase chain reaction for detection of Legionella pneumophila in hospital water networks.

BACKGROUND: Rapid monitoring of Legionella pneumophila (Lp) is essential to reduce the risk of Legionnaires' disease in healthcare facilities. However, culture results take at least eight days, delaying the implementation of corrective measures. Here, we assessed the performance of a qPCR method and determined qPCR action thresholds for the detection of Lp in hospital hot water networks (HWNs). METHODS: Hot water samples (N = 459) were collected from a hospital HWN. Lp were quantified using iQ-Check® Quanti real-time PCR Quantification kits (Bio-Rad) and the results were compared with those of culture. qPCR thresholds corresponding to the culture action thresholds of 10 and 1000 cfu/L were determined on a training dataset and validated on an independent dataset. RESULTS: Lp concentrations measured by culture and qPCR were correlated for both the training dataset (Spearman's correlation coefficient ρ = 0.687, P<0.0001) and the validation dataset (ρ = 0.661, P<0.0001). Lp qPCR positivity thresholds corresponding to culture action thresholds of 10 cfu/L was 91 genome units (gu) per litre (sensitivity, 86.4%; negative predictive value - NPV, 93.3%) and that corresponding to culture action thresholds of 1000 cfu/L was 1048 gu/L (sensitivity, 100%; NPV, 100%). CONCLUSION: Detection of Lp by qPCR could be implemented with confidence in hospitals as a complement to culture in the monitoring strategy to speed up the implementation of corrective measures.

Genome-based typing reveals rare events of patient contamination with Pseudomonas aeruginosa from other patients and sink traps in a medical intensive care unit.

AIM: We used genome-based typing data with the aim of identifying the routes of acquisition of Pseudomonas aeruginosa by patients hospitalized in a medical intensive care unit (MICU) over a long period in a non-epidemic context. METHODS: This monocentric prospective study took place over 10 months in 2019 in a 15-bed MICU that applies standard precautions of hygiene. Lockable sink traps installed at all water points of use were bleach disinfected twice a week. We sampled all sink traps weekly to collect 404 P. aeruginosa environmental isolates and collected all P. aeruginosa isolates (N = 115) colonizing or infecting patients (N = 65). All isolates had their phenotypic resistance profile determined and their genome sequenced, from which we identified resistance determinants and assessed the population structure of the collection at the nucleotide level to identify events of P. aeruginosa transmission. FINDINGS: All sink traps were positive for P. aeruginosa, each sink trap being colonized for several months by one or more clones. The combination of genomic and spatiotemporal data identified one potential event of P. aeruginosa transmission from a sink trap to a patient (1/65, 1.5%) and six events of patient cross-transmission, leading to the contamination of five patients (5/65, 7.7%). All transmitted isolates were fully susceptible to ß-lactams and aminoglycosides. CONCLUSIONS: Genome-based typing revealed the contamination of patients by P. aeruginosa originating from sink traps to be infrequent (1.5%) in an MICU with sink trap-bleaching measures, and that only 7.7% of the patients acquired P. aeruginosa originating from another patient.

Comparison of pulsed-field gel electrophoresis and whole-genome-sequencing-based typing confirms the accuracy of pulsed-field gel electrophoresis for the investigation of local Pseudomonas aeruginosa outbreaks.

AIM: To determine whether pulsed-field gel electrophoresis (PFGE) accurately recognizes isolates belonging to clusters defined by techniques based on whole-genome sequencing (WGS) using Pseudomonas aeruginosa as a model. METHODS: We selected 65 isolates of ST395 P. aeruginosa isolated in seven European hospitals between 1998 and 2012. Isolates were typed by PFGE and sequenced by WGS. A core genome multi-locus sequence typing (cgMLST) analysis based on 3831 genes was performed with a homemade pipeline. FINDINGS: PFGE identified eight pulsotypes and cgMLST differentiated nine clusters and nine singletons. Five cgMLST clusters and pulsotypes (31/65 isolates) coincided perfectly. Isolates without evident epidemiological links grouped by PFGE were separated by cgMLST (16/65 isolates) differentiating cities, suggesting that PFGE should be kept for the investigation of local outbreaks. Importantly, hypermutator isolates still shared the pulsotype with their parents (16/65 isolates), whereas they were not recognized by cgMLST. This shows that PFGE was less affected than WGS-based typing by the accelerated genetic drift that occurs in epidemic P. aeruginosa. CONCLUSIONS: although WGS-based typing has logically become the new reference standard, we show here that the PFGE can be used with confidence for the investigation of local outbreaks caused by P. aeruginosa.

Global emergence of the widespread Pseudomonas aeruginosa ST235 clone.

OBJECTIVES: Despite the non-clonal epidemic population structure of Pseudomonas aeruginosa, several multi-locus sequence types are distributed worldwide and are frequently associated with epidemics where multidrug resistance confounds treatment. ST235 is the most prevalent of these widespread clones. In this study we aimed to understand the origin of ST235 and the molecular basis for its success. METHODS: The genomes of 79 P. aeruginosa ST235 isolates collected worldwide over a 27-year period were examined. A phylogenetic network was built, using a Bayesian approach to find the Most Recent Common Ancestor, and we identified antibiotic resistance determinants and ST235-specific genes. RESULTS: Our data suggested that the ST235 sublineage emerged in Europe around 1984, coinciding with the introduction of fluoroquinolones as an antipseudomonal treatment. The ST235 sublineage seemingly spread from Europe via two independent clones. ST235 isolates then appeared to acquire resistance determinants to aminoglycosides, ß-lactams and carbapenems locally. Additionally, we found that all the ST235 genomes contained the exoU-encoded exotoxin and identified 22 ST235-specific genes clustering in blocks and implicated in transmembrane efflux, DNA processing and bacterial transformation. These unique combinations of genes may have contributed to the poor outcome associated with P. aeruginosa ST235 infections and increased the ability of this international clone to acquire mobile resistance elements. CONCLUSION: Our data suggest that P. aeruginosa ST235 (a) has become prevalent across the globe potentially due to the selective pressure of fluoroquinolones and (b) readily became resistant to aminoglycosides, ß-lactams and carbapenems through mutation and acquisition of resistance elements among local populations.

mcr-1 is borne by highly diverse Escherichia coli isolates since 2004 in food-producing animals in Europe.

OBJECTIVES: In November 2015, a plasmid-mediated colistin resistance, MCR-1, was described in animals, food and humans in China, and it was considered as a potential emerging threat to public health. Therefore, we screened for the mcr-1 gene a European collection of colistin-resistant Escherichia coli (n=218) and Salmonella spp. (n=74) isolated from diseased food-producing animals between 2004 and 2014 and characterized the mcr-1-positive clones. METHODS: Screening for mcr-1 gene was performed by PCR on isolates for which inhibition diameter was <15 mm around a 50 µg disk of colistin. Positive E. coli isolates were then characterized by phylogrouping, multilocus sequence typing and pulsed-field gel electrophoresis typing. Antibiotic susceptibility was determined by disk diffusion testing or by broth microdilution. RESULTS: Among the collection, 42 E. coli and three Salmonella spp. were positive for mcr-1, with continuous detection since 2004 mainly from bovine and swine digestive infections. Most of the mcr-1-positive strains were resistant to amoxicillin and cotrimoxazole but remained susceptible to cephalosporins, carbapenems and piperacillin/tazobactam. All but one isolate were resistant to colistin, with a minimum inhibitory concentration of >2 mg/L. Most of the mcr-1-positive E. coli belonged to the phylogroup A with two prevalent clonal complexes, CC10 and CC165, in which sequence type 10 and sequence type 100 were overrepresented and pulsed-field gel electrophoresis typing revealed a high diversity of pulsotypes. CONCLUSIONS: MCR-1 was detected yearly in European food-producing animal since 2004 with a high diversity of pulsotypes supporting the dissemination of mcr-1 via plasmids.

Appropriateness of aminoglycoside prescriptions in a French university hospital.

INTRODUCTION: Aminoglycosides are a major class of antibiotics. Their use is particularly interesting in the treatment of severe infections but their toxicity is well known. They are mostly prescribed combined with other agents and as first-line treatments. We aimed to assess the appropriateness of aminoglycoside prescriptions in a French university hospital on the basis of the latest French recommendations published in 2011. METHOD: We conducted a prospective study between January 17th and February 4th, 2014 to assess prescription modalities of aminoglycosides on the basis of the following criteria: indication, duration of treatment, dosing schedule, administration modalities, and drug level monitoring. Prescriptions were then compared to the 2011 national guidelines. RESULTS: A total of 68 consecutive prescriptions were analyzed and only 47.8% complied with guidelines. Most physicians complied with recommendations, particularly with the indication for severe infections (95.6%), the administration of a single daily dose (92.6%), and the slow intravenous infusion (30minutes) administration (84%). However, physicians tended to prescribe lower doses than recommended (40.3%), especially to patients presenting with renal insufficiency, and drug level monitoring was not optimal. CONCLUSION: Although new and accurate national recommendations were recently published, aminoglycoside prescription is still not optimal, in particular for dosing and plasma concentration monitoring.

What happens in hospitals does not stay in hospitals: antibiotic-resistant bacteria in hospital wastewater systems.

Hospitals are hotspots for antimicrobial-resistant bacteria (ARB) and play a major role in both their emergence and spread. Large numbers of these ARB will be ejected from hospitals via wastewater systems. In this review, we present quantitative and qualitative data of extended-spectrum ß-lactamase (ESBL)-producing Escherichia coli, vancomycin-resistant enterococci and Pseudomonas aeruginosa in hospital wastewaters compared to community wastewaters. We also discuss the fate of these ARB in wastewater treatment plants and in the downstream environment. Published studies have shown that hospital effluents contain ARB, the burden of these bacteria being dependent on their local prevalence. The large amounts of antimicrobials rejected in wastewater exert a continuous selective pressure. Only a few countries recommend the primary treatment of hospital effluents before their discharge into the main wastewater flow for treatment in municipal wastewater treatment plants. Despite the lack of conclusive data, some studies suggest that treatment could favour the ARB, notably ESBL-producing E. coli. Moreover, treatment plants are described as hotspots for the transfer of antibiotic resistance genes between bacterial species. Consequently, large amounts of ARB are released in the environment, but it is unclear whether this release contributes to the global epidemiology of these pathogens. It is reasonable, nevertheless, to postulate that it plays a role in the worldwide progression of antibiotic resistance. Antimicrobial resistance should now be seen as an 'environmental pollutant', and new wastewater treatment processes must be assessed for their capability in eliminating ARB, especially from hospital effluents.

Clonal complex 398 methicillin-susceptible Staphylococcus aureus bloodstream infections are associated with high mortality.

Within the last decade, methicillin-resistant Staphylococcus aureus belonging to clonal complex 398 (CC398) has become a worldwide threat associated with livestock. More recently, methicillin-susceptible S. aureus (MSSA) belonging to CC398 have been increasingly reported as a cause of invasive infections in patients without livestock contact. We investigated risk factors associated with CC398 bloodstream infections (BSIs) compared with non-CC398 BSIs with a case-control study in a French university Hospital. From January 2010 to December 2014, nonduplicate Staphylococcus aureus (SA) isolates responsible for BSIs in adult patient were typed to identify those belonging to CC398. Each adult patient with a CC398 SA BSI (cases) was matched with 2 non-CC398 SA BSI controls randomly selected on the basis of the time at risk, the unit of hospitalization and susceptibility to methicillin. We retrospectively extracted the clinical information from electronic medical records and used conditional logistic regression for univariate and multivariate analyses. We identified 67 CC398 isolates among the 770 SA responsible for BSI in adult patients. All CC398 isolates were susceptible to methicillin. The proportion of CC398 among MSSA increased steadily from 4.6% in 2010 to 15.1% in 2013 and then stabilized at 13.8% in 2014. Factors significantly associated with CC398 MSSA BSIs were healthcare-associated infection (odds ratio (OR) 3.02, 95% confidence interval (CI) 1.19-7.63), history of neurologic disease (OR 2.51, 95% CI 1.13-5.65) and 30-day mortality (OR 2.44, 95% CI 1.23-4.85).

Detection of Escherichia coli sequence type 131 by matrix-assisted laser desorption ionization time-of-flight mass spectrometry: implications for infection control policies?

BACKGROUND: Sequence type 131 (ST131) is a predominant lineage among extraintestinal pathogenic Escherichia coli. It plays a major role in the worldwide dissemination of extended-spectrum ß-lactamase (ESBL)-producing E. coli. The ST131 pandemic is mainly the result of clonal expansion of the single well-adapted subclone H30-Rx, which is acquired in hospitals more frequently than other ESBL-producing E. coli clones. AIM: To develop a rapid method using matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF MS) to identify ST131 for infection control purposes. METHODS: Peak biomarkers of ST131 were identified from the mass spectrum profiles of 109 E. coli isolates (including 50 ST131 isolates). FINDINGS: The models accurately identified ST131 isolates from mass spectrum profiles obtained with and without protein extraction. CONCLUSIONS: The rapid identification of ST131 isolates with MALDI-TOF MS can be easily implemented in the laboratory, and could help to target infection control measures in patients carrying multi-drug-resistant E. coli that are more likely to spread.

Molecular epidemiology of OXA-48-producing Klebsiella pneumoniae in France.

We characterized 53 OXA-48-producing Klebsiella pneumoniae (OXA-48-Kp) isolated between 2011 and 2013 in 21 French hospitals. All the isolates were genotyped using MLST and PFGE and the population structure of the species was determined by a nucleotide-based analysis of the entire K. pneumoniae MLST database. Most of the OXA-48-Kp isolates also produced CTX-M-15 and remained susceptible to imipenem and meropenem. The isolates were distributed into 20 STs, of which five were dominant (ST15, ST101, ST147, ST395 and ST405). All the OXA-48-Kp clustered in the major clade of K. pneumoniae KpI.

Pseudomonas aeruginosa in French hospitals between 2001 and 2011: back to susceptibility.

The European Antimicrobial Resistance Surveillance Network (EARS-Net) reported an increase in the rates of resistance of Pseudomonas aeruginosa to antimicrobials between 2008 and 2011 in France. This alarming report was based on data collected during the harmonisation of breakpoints by the European Committee on Antimicrobial Susceptibility Testing (EUCAST) committee. However, these data were not supported by the findings of other national surveillance networks. In this study, we assessed the trends in P. aeruginosa antimicrobial drug resistance at six French hospitals over a longer period of time (2001-2011) and with a constant definition of resistance. After the exclusion of incomplete data and duplicates, we sorted 34,065 isolates into the antimicrobial resistance patterns defined by the European Centre for Disease Prevention and Control (ECDC). The proportion of isolates with a resistant pattern (non-susceptible to one or two antimicrobial categories), a multidrug-resistant pattern (non-susceptible to three or four antimicrobial categories) or an extensively drug-resistant pattern (non-susceptible to five or six antimicrobial categories) decreased significantly over time. Logically, the proportion of isolates with a wild-type resistance pattern has increased significantly over the same period. No significant changes in the rates of resistance to cephalosporins and penicillins were observed, whereas carbapenem resistance rates increased. By contrast, the proportion of isolates resistant to fluoroquinolones, aminoglycosides and monobactams decreased significantly over time. In conclusion, our data do not confirm the EARS-net data, suggesting instead that antimicrobial drug resistance in P. aeruginosa might not have increased in French hospitals over the last decade.

Hospital cross-transmission of extended-spectrum ß-lactamase producing Escherichia coli and Klebsiella pneumoniae.

OBJECTIVES: We had for objective to measure the incidence and the clonal diversity of Escherichia coli and Klebsiella pneumoniae producing extended-spectrum ß-lactamases (ESBL) in order to assess the role of patient stay in amplification of the phenomenon, in our teaching hospital. MATERIAL AND METHODS: We measured the quarterly incidence rates of E. coli and K. pneumoniae producing or not producing ESBL in clinical samples between 1999 and 2010. The incidence of ESBL-producing isolates was season-adjusted. We determined the pulsotype of and identified the ESBL in all non-redundant strains isolated between 2009 and 2010. RESULTS: The incidence for 1000 hospitalization days increased from 0.00 to 0.44 for ESBL-producing E. coli, from 0.012 to 0.24 for ESBL-producing K. pneumoniae, from 1999 to 2010. Fifty-three different clones of E. coli were identified among the 61 genotyped isolates. The 28 K. pneumoniae isolates genotyped clustered into 11 different clones, among which one major epidemic clone that included 18 isolates. Respectively 66 and 75% of E. coli and K. pneumoniae isolates produced a CTX-M group 1 ESBL. CONCLUSION: The hospital seems to play a different role in the amplification of ESBL according to the producing species (K. pneumoniae or E. coli). ESBL-producing E. coli seem to have a limited cross-transmission within the hospital and seem to be added to non-producers. Conversely, ESBL-producing K. pneumoniae seem to be cross-transmitted within the hospital and to replace non-producers.

[Chronic wound care leads to the bacterial contamination of the environment]. / Les soins des plaies chroniques entraînent une contamination bactérienne de l'environnement.

AIM: The aim of this study was to determine levels of bacterial contamination of the environment during chronic wound dressing changes. METHODS: Sampling of chronic wounds and of the environment (air and surfaces) was performed during changing of chronic wound dressing. A series of samples was defined as the entire sample for a given day for a given patient. Five sequential samples of air and six samples of surfaces were taken for each series. Staphylococcus aureus, Pseudomonas aeruginosa and enterobacteria were specifically cultured. RESULTS: Thirty series of samples were taken for 26 different patients. Twenty-seven out of these 30 series were colonized with one or two of the target species. For 13 series of the latter samples (13/27, 48.1%), bacteria isolated from the wound were recovered in the environment, namely S. aureus and P. aeruginosa. The six enterobacteria isolated from wounds were not retrieved in the environment. Air samples were more often positive than surfaces samples. CONCLUSION: We demonstrated frequent contamination of the hospital environment with bacteria colonizing wounds during dressing changes. This indicates that wearing of masks and hand disinfection after contact with the environment constitute key measures in the control of bacterial cross-transmission.

Temporal effects of infection control practices and the use of antibiotics on the incidence of MRSA.

BACKGROUND: Meticillin-resistant Staphylococcus aureus (MRSA) has spread throughout the world and has become highly endemic worldwide. The need for implementing MRSA control strategies is no longer a matter of debate. AIM: To determine the temporal association between various infection control practices, the use of antibiotics and the incidence of MRSA in a 1200-bed French university hospital. METHODS: A multi-variate time-series analysis, based on monthly data from a nine-year period (January 2000-December 2008), was performed in a 1200-bed French university hospital to determine the temporal association between different variables and the incidence of MRSA. MRSA colonization pressure, infection control practices and use of antibiotics were considered in the analysis. FINDINGS: Time-series analysis showed a positive significant relationship between the incidence of hospital-acquired MRSA (HA-MRSA) and MRSA colonization pressure, the use of antibiotics (fluoroquinolones, macrolides and aminoglycosides) and the use of gloves. Conversely, a global negative correlation was observed between the incidence of HA-MRSA and the use of alcohol-based hand rub. Overall, the model explained 40.5% of the variance in the monthly incidence of MRSA. CONCLUSION: This study showed that admission of patients with MRSA, the use of antibiotics and infection control practices contributed to the incidence of HA-MRSA. This suggests that efforts should be focused on high compliance with hand disinfection. These results also raise concerns about the use of gloves when caring for patients with MRSA.

Bacterial contamination of the hospital environment during wound dressing change.

INTRODUCTION: The hospital environment plays a role in the cross-transmission of multidrug-resistant bacteria. The aim of this study was to evaluate the bacterial contamination of the hospital environment during chronic wound dressing change. PATIENTS AND METHODS: This study was performed from July 2010 to May 2011. Staphylococcus aureus, Pseudomonas aeruginosa, Acinetobacter baumannii and Enterobacteriaceae were counted in environmental samples (air and surfaces) that were obtained in the rooms of patients with wounds colonized (cases, n=9) or not (controls, n=15) during or not during wound dressing change. Bacterial contamination was compared to that found in the rooms of patients without colonized wounds. RESULTS: The environment was frequently contaminated during wound dressing change (38% of the sampled series were positive). In comparison, the contamination was less frequent in the environment of patients with colonized wounds when the wounds were not being dressed (14.3%) and in controls (3.8%). S. aureus was the most frequent species identified in positive samples. DISCUSSION: These results suggest that previously recommended measures such as hand hygiene after contact with the environment and wearing a mask are justified. Moreover, other measures should be suggested, in particular cleaning the room before and after dressing change of colonized wounds. LEVEL OF EVIDENCE: Level III: case control study.

Molecular epidemiology of multidrug-resistant Pseudomonas aeruginosa in a French university hospital.

The aim of this study was to assess the incidence and molecular epidemiology of multidrug-resistant (MDR) Pseudomonas aeruginosa in our university hospital. Analysis included antimicrobial susceptibility profiling, bla gene identification and pulsed-field gel electrophoresis (PFGE). During the one-year study, 654 patients had at least one sample that tested positive for P. aeruginosa, of whom 38 (5.8%) were colonised or infected with an MDR isolate, giving an incidence of 0.1 patient per 1000 patient-days. The 38 non-duplicate isolates yielded 12 different PFGE patterns, three of which included isolates from four patients and one of which included isolates from 15 patients. Two isolates produced acquired extended-spectrum ß-lactamase (one OXA-14 and one OXA-28). Genotyping showed that cross-transmission was responsible for about 70% of MDR P. aeruginosa cases although spatio-temporal analysis failed to demonstrate when this might have occurred for most cases. The major epidemic and the three main micro-epidemic clones were already present in our hospital with a more susceptible phenotype. It is likely that some P. aeruginosa clones are endemic in our hospital and that, within these clones, MDR isolates emerge under antibiotic pressure. Our results indicate that cross-transmission plays a major role in the spread of MDR P. aeruginosa and suggest that priority should be given to the improvement of standard hygienic precautions.

Emergence of extensive-drug-resistant Pseudomonas aeruginosa in a French university hospital.

The aim of this study was to describe the molecular epidemiology and the mechanisms of resistance to beta-lactams of emerging extensive-drug-resistant Pseudomonas aeruginosa (XDRPA) in a tertiary-care university hospital over a three-year period. Analysis included antimicrobial susceptibility profiling and pulsed-field gel electrophoresis (PFGE). Resistance mechanisms to beta-lactams were identified: production of naturally occurring and acquired beta-lactamases, overproduction of MexAB-OprM and MexXY efflux systems and loss of porin OprD were assessed. Eighteen patients were colonised or infected with XDRPA which remained susceptible to colistin and, to a lesser extent, to rifampicin. beta-lactam resistance was, in most cases, due to the overproduction of AmpC, overproduction of the MexXY efflux system and loss of porin OprD. One isolate produced the class D extended-spectrum oxacillinase (OXA-ESBL) Oxa-28, but none produced metallo-beta-lactamase (MBL) or class A extended-spectrum beta-lactamase (ESBL). The XDRPA clustered in eight PFGE patterns and both the acquisition and loss of resistance determinants was observed within a single clone during its spread. The emergence of XDRPA isolates in our university hospital has been characterised by genotypic heterogeneity, variation of mechanisms of resistance to beta-lactams in a single clone and the predominance of chromosomally encoded resistance mechanisms.

Antibiotic susceptibility and mechanisms of beta-lactam resistance among clinical strains of Pseudomonas aeruginosa: first report in Algeria.

OBJECTIVE: Pseudomonas aeruginosa is a major causative agent of hospital infections. Studies on this subject being rare in Algeria, we determined the antibiotic susceptibility of P. aeruginosa and investigated the mechanisms of beta-lactam resistance and the spread of multidrug resistant strains in the university affiliated Hospital of Tlemcen (Algeria). DESIGN: One hundred and ninety-nine consecutive strains of P. aeruginosa were collected between November 2005 and February 2007. MICs of antibiotics were measured by the agar dilution method. The resistance mechanisms to beta-lactams were identified phenotypically or by molecular methods (isoelectrofocusing, PCR and sequencing). Strains expressing a secondary beta-lactamase were serotyped and genotyped (Random Amplified Polymorphic DNA). RESULTS: The proportion of susceptible isolates were: ticarcillin (56%), piperacillin-tazobactam (81%), ceftazidime (88%), cefepime (80%), aztreonam (64%), imipenem (65%), amikacin (83%), tobramycin (81%) and ciprofloxacin (97%) according to the French CASFM breakpoints. Resistance to beta-lactams was linked to the production of transferable beta-lactamases (16%), overproduction of cephalosporinase AmpC (12%) and/or non-enzymatic mechanisms such as the loss of porin OprD (35%) and overproduction of the active efflux system MexAB-OprM (24%). High level resistance to ticarcillin was due to the expression of beta- lactamase OXA-10 alone or associated with TEM-110. A genotypic analysis revealed the spread of a multidrug resistant epidemic clone expressing these two acquired beta-lactamases in the surgical ICU. CONCLUSIONS: This study shows that resistance to antibiotics, in particular to imipenem of P. aeruginosa, is becoming a cause of concern in the Hospital of Tlemcen.