JAC Advance Access originally published online on September 1, 2006
Journal of Antimicrobial Chemotherapy 2006 58(5):1082-1085; doi:10.1093/jac/dkl367
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Prior use of carbapenems may be a significant risk factor for extended-spectrum ß-lactamase-producing Escherichia coli or Klebsiella spp. in patients with bacteraemia
1 Department of Infectious Diseases, Hospital Clínic, IDIBAPS—University of Barcelona Barcelona, Spain 2 Microbiology Laboratory, Hospital Clínic, IDIBAPS—University of Barcelona Barcelona, Spain
*Corresponding author. Tel: +34-932272322; Fax: +34-934514438; E-mail: jamarti{at}clinic.ub.es
Received 7 June 2006; returned 5 July 2006; revised 14 August 2006; accepted 15 August 2006
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and methodsResultsDiscussionTransparency declarationsReferences |
|---|
Background: The increasing prevalence of extended-spectrum ß-lactamase (ESBL)-producing Enterobacteriaceae will probably trigger a rise in the use of carbapenems. The effect of these antibiotics on the risk of involvement of ESBL-producing organisms in serious infections is unclear.
Methods: Retrospective analysis of 2172 episodes of healthcare-associated bacteraemia diagnosed during a 3 year period in a teaching hospital. Putative risk factors included demographics, co-morbidities, previous isolation of an ESBL-producing organism and exposure to antibiotics. Univariate and multivariate analysis of the association of risk factors with ESBL-producing organisms was performed in the entire series of bacteraemic episodes and in those due to Escherichia coli or Klebsiella spp.
Results: In the entire series, prior isolation of an ESBL-producing organism [odds ratio (OR) 5.9 (3.02, 11.5)]; an ultimately/finally fatal co-morbidity [OR 2.8 (1.55, 4.95)]; renal transplantation [OR 4.3 (1.96, 9.63)]; a urinary source [OR 4.2 (2.22, 7.84)]; shock [OR 2.4 (1.35, 4.1)] and previous use of cephalosporins [OR 2.6 (1.54, 4.51)], carbapenems [OR 2.5 (1.24, 5.05)] and glycopeptides [OR 0.4 (0.13, 0.93)] were significantly associated with ESBL-producing E. coli or Klebsiella spp. by multivariate analysis. Prior isolation of an ESBL-producing organism, an ultimately/finally fatal co-morbidity, renal transplantation, and previous use of cephalosporins and carbapenems were also significant in the analysis restricted to episodes due to E. coli or Klebsiella spp.
Conclusions: In patients with healthcare-associated bacteraemia, prior use of carbapenems may be only second to cephalosporins as the most significant antibiotic exposure associated with the involvement of ESBL-producing organisms.
Keywords: imipenem , meropenem , hospital infections
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and methodsResultsDiscussionTransparency declarationsReferences |
|---|
Extended-spectrum ß-lactamase-producing Enterobacteriaceae (ESBL-E) are currently prevalent worldwide.1 As the frequency of severe infections due to ESBL-E increases, a rise in the use of carbapenems will probably ensue which may, in turn, have untoward consequences on the prevalence of resistance to these antibiotics in microorganisms such as Pseudomonas aeruginosa.2 In previous studies, several clinical characteristics have been noted as associated with ESBL-E colonization or infection, although in general they are rather non-specific.1 Concerning prior exposure to antibiotics, cephalosporins have the highest record as a predisposing factor for ESBL-E and, indeed, restriction of cephalosporins in the hospital, together with increased use of carbapenems or piperacillin/tazobactam, has been associated with the success in diminishing the prevalence of nosocomial ESBL-E.2 Other investigations have incriminated quinolones, aminoglycosides, trimethoprim/-sulfamethoxazole, metronidazole or total antibiotic use as a significant risk exposure.1,3–5
We have analysed the factors associated with ESBL-producing Escherichia coli or Klebsiella spp. in a cohort of consecutive patients with healthcare-associated bacteraemia and present evidence to suggest that recent carbapenem use may be only second to previous administration of cephalosporins as the most significant antibiotic exposure associated with the involvement of these microorganisms.
|
Patients and methods |
|---|
|
TopAbstractIntroductionPatients and methodsResultsDiscussionTransparency declarationsReferences |
|---|
The study was carried out during a 3 year period (January 2003–December 2005) at Hospital Clínic, a 700 bed university hospital in Barcelona, Spain. At the study centre, a team of infectious diseases physicians, microbiologists and infection control nurses continuously and prospectively keep two databases, one in which all episodes of clinically significant bacteraemia are registered and another where all patients from whom a multiresistant microorganism is isolated are recorded. Data used in the present analysis were from these databases.
Clinical variables collected from patients with bacteraemia included the following: demographics (age and gender), place of acquisition (healthcare-associated or from the community), underlying diseases, McCabe classification of underlying diseases, neutropenia, previous treatment with corticosteroids, prior surgery (within the last month), prior systemic antibiotics, microbiological documentation of ESBL-producing E. coli or Klebsiella spp. in any clinical sample during the year before the index episode of bacteraemia, primary source of infection and shock on admission.
An infection was considered as healthcare-associated when it occurred 2 days after admission or within 30 days after discharge from an acute care hospital (hospital-acquired), or in patients attending a haemodialysis or other clinic or receiving any kind of intravenous therapy provided by a hospital-dependent facility within 30 days prior to the bloodstream infection. The source of infection was established according to common clinical, analytical, radiographic and microbiological criteria. Antimicrobial therapy given for at least 48 h in the last 30 days before bacteraemia diagnosis was considered as previous antibiotic treatment. Antibiotic susceptibility and ESBL production were determined following CLSI guidelines.6
Statistical analysis
Proportions were compared by the 
2 test or the Fisher's test when appropriate. Characteristics of patients with ESBL-producing E. coli or Klebsiella spp. were compared with those of patients with clinical significant bacteraemia not due to these microorganisms and with those with bacteraemia due to non-ESBL-producing E. coli or Klebsiella spp. Variables with a P value <0.2 in univariate analysis were subjected to further selection by a stepwise logistic regression procedure.
|
Results |
|---|
|
TopAbstractIntroductionPatients and methodsResultsDiscussionTransparency declarationsReferences |
|---|
During the study period, 3380 episodes of clinically significant bacteraemia were diagnosed, of which 2172 (64%) were considered to be healthcare-associated. Among these episodes, an ESBL-producing Enterobacteriaceae was involved in 83 (3.8%), of which 54 were due to E. coli (13% of bloodstream infections due to E. coli), and 31 to Klebsiella spp. (17% of those due to Klebsiella spp.). In two episodes, both E. coli and Klebsiella pneumoniae were involved. The species of Klebsiella were K. pneumoniae in 26 episodes (84%) and Klebsiella oxytoca in 5 (16%). The mean (±SD) age of the entire cohort was 61.6 (16.7) years, 1323 episodes (61%) occurred in males and 1413 (65%) were hospital-acquired. A total of 187 patients (9%) received a carbapenem within 30 days before the onset of bacteraemia. Imipenem was given to 104 cases (57%), meropenem to 82 (44%) and ertapenem to 1 (0.5%).
Univariate analysis of putative predictors of ESBL-producing E. coli or Klebsiella spp. involvement is shown in Table 1. In both the entire series of bacteraemic episodes and in those due to E. coli or Klebsiella spp., prior isolation of an ESBL-producing organism, previous administration of corticosteroids, an ultimately or finally fatal underlying disease, renal transplantation and previous use of cephalosporins and carbapenems were the antecedent factors significantly associated with the involvement of ESBL-E. Nosocomial acquisition of the infection and previous use of cephalosporins, carbapenems, aminoglycosides and ‘other antibiotics’ were more strongly linked to an ESBL-producing organism when the analysis was limited to episodes due to E. coli or Klebsiella spp. With regard to the variables assessed on admission, it is of note that in the E. coli or Klebsiella spp. subgroup of bacteraemic episodes, the association of a urinary tract source with ESBL-producing organisms diminished to the point of being no longer significant.
|
The results of multivariate logistic regression analysis performed in both the entire series of bacteraemic episodes and in those due to E. coli or Klebsiella spp. are shown in Table 2. In both cases, the following four variables remained as significantly associated with the involvement of an ESBL-producing E. coli or Klebsiella spp. strain: prior isolation of an ESBL-producing organism, an ultimately or finally fatal underlying disease, renal transplantation and previous use of cephalosporins and carbapenems. In addition, for the entire cohort of patients, shock and a urinary tract source were also significant predictors, whereas prior treatment with glycopeptides entered the model as a protective factor. In the analysis of the subgroup of episodes due to E. coli or Klebsiella spp., hospital-acquisition was added as an independent predictor.
|
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and methodsResultsDiscussionTransparency declarationsReferences |
|---|
In the present study, we have analysed the variables associated with ESBL-producing E. coli or Klebsiella spp. in patients with healthcare-associated bacteraemia and provide evidence to suggest that recent exposure to carbapenems may be a significant risk factor.
Numerous studies, most of them using a case–control design, have assessed the variables associated with ESBL-E colonization or infection and have been extensively reviewed.1 In addition to the already described risk factors, present data suggest that, in patients with bacteraemia, antecedent infection or colonization with ESBL-producing organisms, renal transplantation and prior use of carbapenems may also be relevant. The loss of significance of the source of infection and shock as independent predictors when the analysis was limited to the subgroup of episodes due to E. coli or Klebsiella spp. suggests that they are factors more strongly associated with these species in general than specifically with ESBL-producing organisms.
In previous studies, antibiotic administration has been a common risk factor for ESBL-E colonization or infection and, in this regard, the role of cephalosporins and quinolones is widely recognized.1,4 Interestingly, in at least two case–control studies,3,5 carbapenems were among the antimicrobials more likely to have been previously used in patients with ESBL-E, although these antibiotics did not remain as significant predictors in the multivariate analysis. The association of previous carbapenem use with ESBL-E infection is, however, counterintuitive and of uncertain meaning. Two explanatory scenarios are possible. In the first one, carbapenems could enrich the gut reservoir of ESBL-E in patients already colonized with a low number of microorganisms or promote their settlement after a transmission event. This could be due to the lower in vitro activity of carbapenems against ESBL-E versus non-ESBL-producing Enterobacteriaceae (a contention at least valid for imipenem and ertapenem)7 combined with a relative poor biliary excretion of the drugs. Although the failure of meropenem to avoid faecal colonization with an SHV-5-producing K. pneumoniae in neonates has been described,8 no indication of any promotion of ESBL-E acquisition was observed after administering ertapenem to adults with intra-abdominal infection.9 At the other extreme, association of prior use of carbapenems with ESBL-E infection may only represent an instance of the paradox predicted by deterministic mathematical models.10 According to these constructions, when two antibiotics are used in the hospital, exposure to a drug for which there is no resistance will be, at the individual level, a risk factor for carrying bacteria resistant to the other drug. If this were the case, increasing the use of carbapenems would in fact decrease the prevalence of ESBL-E carriers.
The present analysis has drawbacks typical of observational retrospective studies. Although we have adjusted for the variables most obviously linked to both the risk of developing ESBL-E bacteraemia and previous use of carbapenems (antecedent isolation of these organisms from a clinical sample and several underlying conditions), we cannot entirely dismiss the influence of other possibly important but unaccounted variables.
In conclusion, recent use of carbapenems may be a significant independent risk factor for ESBL-producing E. coli or Klebsiella spp. in patients with bacteraemia. Although the relationship of prior carbapenem administration with ESBL-E colonization or infection needs further assessment, present data suggest that there may be an additional reason for prudence before an eventual increasing tendency towards a more indiscriminate use of these antibiotics.
|
Transparency declarations |
|---|
|
TopAbstractIntroductionPatients and methodsResultsDiscussionTransparency declarationsReferences |
|---|
None to declare.
|
Acknowledgements |
|---|
Financial support: none.
|
References |
|---|
|
TopAbstractIntroductionPatients and methodsResultsDiscussionTransparency declarationsReferences |
|---|
1 Paterson DL and Bonomo RA. (2005) Extended-spectrum ß-lactamases: a clinical update. Clin Microbiol Rev 18:657–86.
2 Owens RC and Rice L. (2006) Hospital-based strategies for combating resistance. Clin Infect Dis 42:Suppl 4, S173–81.
3 Kanafani ZA, Mehio-Sibai A, Raaj GF, et al. (2005) Epidemiology and risk factors for extended-spectrum ß-lactamase-producing organisms: a case control study at a tertiary care center in Lebanon. Am J Infect Control 33:326–32.[CrossRef][ISI][Medline]
4 Rodriguez Baño J, Navarro MD, Romero L, et al. (2006) Clinical and molecular epidemiology of extended-spectrum ß-lactamase-producing Escherichia coli as a cause of nosocomial infection or colonization: implications for control. Clin Infect Dis 42:37–45.[CrossRef][ISI][Medline]
5
Graffunder EM, Preston KE, Evans AM, et al. (2005) Risk factors associated with extended-spectrum ß-lactamase-producing organisms at a tertiary care hospital. J Antimicrob Chemother 56:139–45.
6 Performance Standards for Antimicrobial Susceptibility Testing; Sixteenth Informational Supplement Clinical and Laboratory Standard Institute. (2006) Document M100-S16. CLSI, Wayne, PA, USA.
7
Hernández JR, Martínez-Martínez L, Cantón R, et al. (2005) Nationwide study of Escherichia coli and Klebsiella pneumoniae producing extended-spectrum ß-lactamases in Spain. Antimicrob Agents Chemother 49:2122–5.
8 Gundes S, Arisoy AE, Kolayli F, et al. (2005) An outbreak of SHV-5 producing Klebsiella pneumoniae in a neonatal intensive care unit; meropenem failed to avoid fecal colonization. New Microbiol 28:231–6.[Medline]
9 Dinubile MJ, Friedland I, Chan CY, et al. (2005) Bowel colonization with resistant gram-negative bacilli after antimicrobial therapy of intra-abdominal infections: observations from two randomized comparative clinical trials of ertapenem therapy. Eur J Clin Microbiol Infect Dis 24:443–9.[CrossRef][ISI][Medline]
10
Lipsitch M, Bergstrom CT, Levin BR. (2000) The epidemiology of antibiotic resistance in hospitals: paradoxes and prescriptions. Proc Natl Acad Sci USA 97:1938–43.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  P. M. Hawkey The growing burden of antimicrobial resistance J. Antimicrob. Chemother., September 1, 2008; 62(suppl_1): i1 - i9. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Patel, L.-A. McNutt, and T. P. Lodise Relationship between Various Definitions of Prior Antibiotic Exposure and Piperacillin-Tazobactam Resistance among Patients with Respiratory Tract Infections Caused by Pseudomonas aeruginosa Antimicrob. Agents Chemother., August 1, 2008; 52(8): 2933 - 2936. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Tacconelli, G. De Angelis, M. A. Cataldo, E. Pozzi, and R. Cauda Does antibiotic exposure increase the risk of methicillin-resistant Staphylococcus aureus (MRSA) isolation? A systematic review and meta-analysis J. Antimicrob. Chemother., January 1, 2008; 61(1): 26 - 38. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||