Geographical variation in therapy for bloodstream infections due to multidrug-resistant Enterobacteriaceae: a post-hoc analysis of the INCREMENT study

https://doi.org/10.1016/j.ijantimicag.2017.08.005Get rights and content

Highlights

  • Regional variation exists in therapy for bloodstream infections caused by ESBL-E or CPE.

  • Location influenced the empirical use of β-lactam/β-lactamase inhibitors (BLBLIs) or carbapenems.

  • BLBLI use for ESBL-E or combination therapy for CPE also varied by location.

  • Variation by location remained after adjustment for clinical factors.

  • These data may help clinical trial design and antimicrobial stewardship efforts.

Abstract

We describe regional differences in therapy for bloodstream infection (BSI) caused by extended-spectrum β-lactamase-producing Enterobacteriaceae (ESBL-E) or carbapenemase-producing Enterobacteriaceae (CPE). Patients (n = 1482) in 12 countries from an observational study of BSI caused by ESBL-E or CPE were included. Multivariate logistic regression was used to calculate adjusted odds ratios (aORs) for the influence of country of recruitment on empirical use of β-lactam/β-lactamase inhibitors (BLBLIs) or carbapenems, targeted use of BLBLIs for ESBL-E and use of targeted combination therapy for CPE. Compared with Spain, BLBLI use for empirical therapy was least likely in sites from Israel (aOR 0.34, 95% CI 0.14–0.81), Greece (aOR 0.49, 95% CI 0.26–0.94) and Canada (aOR 0.31, 95% CI 0.11–0.88) but more likely in Italy (aOR 1.58, 95% CI 1.11–2.25) and Turkey (aOR 2.09, 95% CI 1.14–3.81). Empirical carbapenem use was more likely in sites from Taiwan (aOR 1.73, 95% CI 1.03–2.92) and USA (aOR 1.89, 95% CI 1.05–3.39) and less likely in Italy (aOR 0.44, 95% CI 0.28–0.69) and Canada (aOR 0.10, 95% CI 0.01–0.74). Targeted BLBLIs for ESBL-E was more likely in Italian sites. Treatment at sites within Israel, Taiwan, Turkey and Brazil was associated with less combination therapy for CPE. Although this study does not provide precise data on the relative prevalence of ESBL-E or CPE, significant variation in therapy exists across countries even after adjustment for patient factors. Better understanding of what influences therapeutic choices for these infections will aid antimicrobial stewardship efforts.

Introduction

Bloodstream infections (BSIs) are an important cause of morbidity and mortality worldwide. Differences in population demographics, risk factor distribution and microbiology influence the incidence of BSI within different countries. Enterobacteriaceae are a major cause of BSI, with Escherichia coli and Klebsiella pneumoniae being the two most common Gram-negative species isolated from blood cultures both in the community and in healthcare settings [1], [2]. Extended-spectrum β-lactamase (ESBL) enzymes confer resistance to oxyimino-cephalosporins and monobactams in addition to penicillins and have become widespread among Enterobacteriaceae [3], [4], with rising trends even in low-prevalence countries [5], [6]. ESBL-producing organisms often carry other resistance genes, thus limiting choices for effective antimicrobial therapy [7]. Owing to their stability to ESBLs, carbapenems have been considered the preferred agent for the treatment of serious infections caused by ESBL-producers [3], but overuse of carbapenems may provide selection pressure for carbapenem resistance [8]. Carbapenem-resistant Enterobacteriaceae, often resulting from the acquisition of carbapenemase genes, are now an emerging global public health threat [9], [10]. Although geographical variation in the prevalence of ESBL-producing Enterobacteriaceae (ESBL-E) or carbapenemase-producing Enterobacteriaceae (CPE) causing BSI is well known, it is less clear how this variation influences clinical practice in terms of selecting empirical or targeted treatment regimens.

The objectives of this study were to investigate variation across countries in antibiotic regimens used as empirical or targeted therapy for resistant Gram-negative BSIs, with the following hypotheses: (i) regional variation exists in the choice of empirical or targeted therapy for BSI caused by ESBL-E or CPE; (ii) regional variation exists in the use of β-lactam/β-lactamase inhibitor (BLBLI) agents as targeted therapy for bacteraemia caused by ESBL-E; and (iii) regional variation exists in the use of targeted combination therapy for bacteraemia caused by CPE.

Section snippets

Study design and participants

This was a substudy of a retrospective international cohort study (INCREMENT project; ClinicalTrials.gov identifier NCT01764490) investigating the outcome impact of different antimicrobial regimens in the empirical and targeted therapy of BSI caused by ESBL-E or CPE from January 2004 to December 2013 [11]. Thirty-seven hospitals from 12 countries (Spain, Italy, Greece, Taiwan, Turkey, Israel, USA, Argentina, Canada, Germany, Brazil and South Africa) participated in the INCREMENT project.

Results

A total of 1482 patients (1003 with ESBL-E and 479 with CPE) were enrolled from 12 countries, with most cases recruited from sites in Spain (47.2%) (Fig. 1). The baseline patient characteristics are presented in Table 1. Overall, CPE accounted for 32.3% (479/1482) of cases and were most frequently submitted from Italy (n = 115), Spain (n = 99), Greece (n = 89) and Taiwan (n = 60), whereas Canada and Germany contributed no CPE cases (Fig. 1). It should be noted that these proportions reflect

Discussion

In the present study, we sought to understand the different therapeutic approaches to BSI caused by multidrug-resistant (MDR) Enterobacteriaceae across participant sites according to the country of recruitment. Considerable geographical variation was seen in the choice of therapy, either when selected empirically or targeted against a known pathogen. Whilst much of this might be explained by the background prevalence of resistance, this may not account for all of the variation seen.

Historical

Conclusions

In this international observational cohort of patients with BSIs caused by MDR Enterobacteriaceae, a preference to treat ESBL-E BSI with carbapenems and CPE BSI with alternatives to carbapenems or combination therapy was observed. In some countries, such as Italy and Turkey, the likelihood of using empirical BLBLIs for ESBL-E are significantly higher than in recruiting sites in other countries such as Israel, Greece and Canada. Being treated in participant sites from the USA or Taiwan was

Acknowledgments

The authors would like to acknowledge Stuart Paynter and Anita Pelecanos for assistance with the manuscript.

Funding: PNAH is supported by an Australian Postgraduate Award from the University of Queensland (Brisbane, QLD, Australia). The study was funded by Planes Nacionales de I+D+i 2008-2011 / 2013-2016 and Instituto de Salud Carlos III, Subdirección General de Redes y Centros de Investigación Cooperativa, Ministerio de Economía, Industria y Competitividad,Spanish Network for Research in

References (33)

  • R. Tal Jasper et al.

    The complex epidemiology of extended-spectrum β-lactamase-producing Enterobacteriaceae

    Future Microbiol

    (2015)
  • J.D. Turnidge et al.

    Community-onset gram-negative surveillance program annual report, 2012

    Commun Dis Intell Q Rep

    (2014)
  • P.R. Lagace-Wiens et al.

    Trends in antibiotic resistance over time among pathogens from Canadian hospitals: results of the CANWARD study 2007–11

    J Antimicrob Chemother

    (2013)
  • A.J. Mathers et al.

    The role of epidemic resistance plasmids and international high-risk clones in the spread of multidrug-resistant Enterobacteriaceae

    Clin Microbiol Rev

    (2015)
  • US Centers for Disease Control and Prevention

    Antibiotic resistance threats in the United States

    (2013)
  • B. Gutierrez-Gutierrez et al.

    A multinational, preregistered cohort study of β-lactam/β-lactamase inhibitor combinations for treatment of bloodstream infections due to extended-spectrum-β-lactamase-producing Enterobacteriaceae

    Antimicrob Agents Chemother

    (2016)
  • Cited by (8)

    • Healthcare-associated carbapenem-resistant Klebsiella pneumoniae bloodstream infections: Risk factors, mortality, and antimicrobial susceptibility, 2017–2019

      2021, Journal of the Formosan Medical Association
      Citation Excerpt :

      In this study, the length of hospital stay and exposure to carbapenems before the onset of K. pneumoniae mBSI were predictive for healthcare-associated CRKP infection. The reason why these issues were particularly important may be due to the easy access to medical services in Taiwan with the high hospitalization rate21,22 and overuse of antibiotics, including carbapenem agents.23 Additional recognized risk factors for healthcare-associated CRKP bacteremia in this study were the presence of liver cirrhosis, and lymphoma or multiple myeloma.

    • Predictors of outcome in patients with severe sepsis or septic shock due to extended-spectrum β-lactamase-producing Enterobacteriaceae

      2018, International Journal of Antimicrobial Agents
      Citation Excerpt :

      BLBLI has been evaluated for the treatment of these infections [18,31], but no definitive data were reported about the efficacy of BLBLI in the treatment of severe infections compared with carbapenems [14,32]. The use of carbapenems should be optimized, with a role as empirical therapy only for patients who have high risk factors for colonization/infection, based also on geographical patterns of resistance for ESBL strains not susceptible to BLBLI, and on patients who are affected by non-urinary source of bacteremia [33]. In the remaining subjects, in whom clinical conditions or risk factors may not require immediate ESBL antimicrobial coverage, a guided therapy once culture results are obtained may be sufficient [34].

    View all citing articles on Scopus

    Present address: University of Queensland, UQ Centre for Clinical Research (UQCCR), Infection & Immunity Theme, Level 5, Building 71/918 Royal Brisbane and Women's Hospital, Herston, QLD 4029, Australia.

    Participants of the ESGBIS/REIPI/INCREMENT Group are given inAppendix 1.

    View full text