Hostname: page-component-6898cfcb55-cbk24 Total loading time: 0 Render date: 2026-11-11T21:21:43.778Z Has data issue: false hasContentIssue false
  • English
  • Français

Increasing burden of Escherichia coli, Klebsiella pneumoniae, and Enterococcus faecium in hospital-acquired bloodstream infections (2000–2014): A national dynamic cohort study

Published online by Cambridge University Press:  23 April 2019

Koen BlotOpen the ORCID record for Koen Blot[Opens in a new window],
Naïma Hammami,
Stijn Blot,
Dirk Vogelaers and
Marie-Laurence Lambert
Koen Blot*
Affiliation:
Faculty of Medicine and Health Sciences, Ghent University, Ghent, Belgium
Naïma Hammami
Affiliation:
Healthcare-Associated Infections and Antimicrobial Resistance, Public Health and Surveillance Department, Scientific Institute of Public Health, Brussels, Belgium
Stijn Blot
Affiliation:
Faculty of Medicine and Health Sciences, Ghent University, Ghent, BelgiumBurns, Trauma and Critical Care Research Centre, University of Queensland Centre for Clinical Research, Faculty of Medicine, The University of Queensland, Brisbane, Queensland, Australia
Dirk Vogelaers
Affiliation:
Faculty of Medicine and Health Sciences, Ghent University, Ghent, BelgiumGeneral Internal Medicine, Ghent University Hospital, Ghent, Belgium
Marie-Laurence Lambert
Affiliation:
Healthcare-Associated Infections and Antimicrobial Resistance, Public Health and Surveillance Department, Scientific Institute of Public Health, Brussels, Belgium
*
Author for correspondence: Koen Blot, Email: koen.blot@ugent.be
Rights & Permissions [Opens in a new window]

Abstract

The epidemiology of hospital-acquired bloodstream infections (HABSIs) based on the Belgian national surveillance program was analyzed (2000–2014). Our mixed-effects regression analysis identified increased rates of Escherichia coli, Klebsiella pneumoniae, and Enterococcus faecium. HABSI incidence and resistance patterns should be further monitored because of their impact on proper empiric antibiotic therapy.

Information

Type
Concise Communication
Information
Infection Control & Hospital Epidemiology , Volume 40 , Issue 6 , June 2019, pp. 705 - 709
Copyright
© 2019 by The Society for Healthcare Epidemiology of America. All rights reserved. 

Hospital-acquired bloodstream infections (HABSIs) are an important cause of increased morbidity, mortality, length of stay, and costs.Reference Blot, Depuydt and Annemans 1 – Reference Pien, Sundaram and Raoof 3 Although surveillance of blood cultures may not fully represent the epidemiology of all underlying infections, namely those without bacteremia, the clinical relevance of these invasive isolates is indisputable. 4 The Belgian institution Sciensano, previously WIV-ISP, has headed the national “Surveillance of Bloodstream Infections in Hospitals” program. This surveillance program has encouraged hospitals to participate and to collect hospital-wide HABSI case-based data since 1992. Identifying incidence rates, origins, microorganisms, and their trends is essential to characterizing the burden of hospital-acquired infections. In this study, we aimed to analyze HABSI incidence trends (2000–2014) with subgroup analyses for microorganism, origin of infection, and according to setting (ie, intensive care unit (ICU) or university hospital).

Methods

Study design and setting

A national dynamic cohort study of HABSI epidemiology was performed based on the Belgian “Surveillance Data of BSI in Hospitals” program from January 2000 to December 2014. 5

Participants

Participation requires case-based recording of all HABSI for a minimum of 3 consecutive months annually. Participation became mandatory from 2014 onward; eligible hospitals were all acute- and chronic-care hospitals with >150 beds. Hospitals included in this cohort study reported both HABSI (numerator) and patient day (denominator) data for at least 1 trimester during at least 5 years.

Case definitions and variables

HABSI is defined as BSI with onset 2 or more days after hospital admission, not present upon admission (Appendix 1 online). A BSI requires at least 2 separate cultures with clinical symptoms if the causal microorganism is a skin commensal or with 1 culture of a recognized pathogen. HABSI origins are classified as central-line associated, secondary, or unknown.Reference Garner, Jarvis, Emori, Horan and Hughes 6 CLABSI was defined as BSI with concomitant catheter culture or a central venous catheter (CVC) in place within 2 days of BSI onset, unrelated to another infectious site. CDC CLABSI surveillance definitions were applied post hoc to HABSI of unknown origin with a CVC in place in the previous 48 hours. 7 Collected data include HABSI onset date, probable infectious origin, causal microorganism(s), setting (hospital ward or ICU), and hospital type [(university-affiliated or acute-care versus chronic-care [ie, >14-day average length of stay]). Denominator data included number of hospital-wide and ICU patient days per trimester. HABSI incidence was reported as a mean rate per 10,000 patient days hospital-wide and mean rate per 1000 ICU patient days. Antibiotic resistance data collection was mandatory from 2013 to 2014 for Staphylococcus aureus (oxacillin, vancomycin), Enterococcus spp (vancomycin), Enterobacteriaceae (third-generation cephalosporin, carbapenem), Pseudomonas aeruginosa (carbapenem), and Acinetobacter spp (carbapenem).

Statistical methods

Mixed-effects negative binomial regression analysis was used to calculated adjusted incidence rate ratios (IRRs) with 95% confidence interval (CI) to determine the annual change in total HABSI rate (Appendix 2 online). Fixed confounding factors were university hospital status, chronic care facility, and risk exposure (patient days per trimester). Mixed effects were applied to adjust for varying hospital participation and characteristics. This analysis was applied for total, gram-negative, and gram-positive HABSI rates. Subgroup analyses were performed for ICUs, university hospitals, and the most common microorganisms. In the regression analysis, we calculated both relative (incidence rate ratio) and absolute infection rate changes (per 10,000 patient days). Antibiotic resistance data were classified as sensitive or resistant (intermediate or complete resistance).

In our sensitivity analysis, we evaluated selection bias by repeating the analyses of HABSI rates with 3 different hospital cohorts with varying levels of participation: ≥1 trimester for ≥3 years, ≥1 trimester for ≥10 years, or 4 trimesters for ≥3 years. Statistical analyses were performed using Stata version 14 software (StataCorp, College Station, TX). P ≤ .05 was considered statistically significant.

Results

In total, 110 hospitals participated for at least 1 trimester for 5 years. This cohort included 56 450 patients with 59,941 HABSIs from 66,610 microorganisms. Table 1 presents the included hospital characteristics. Missing quarterly patient day data led to the exclusion of 91 HABSI hospital-wide (0.2%) and 1,265 from the ICU (9.8%). We detected a 6-fold higher infection rate in ICUs (4.4 HABSIs per 1,000 patient days; IQR, 2.8–7.2) versus hospital-wide (6.8 HABSIs per 1,000 patient days; IQR, 4.6–9.4). The CLABSI rates were 1.6 HABSIs per 1,000 patient days (IQR 0.9–2.8) hospital-wide and 1.8 HABSIs per 1,000 patient days (IQR, 0.9–2.7) in the ICUs.

Table 1.

Hospital Characteristics and Hospital-Acquired Bloodstream Infection Incidence

Note. HABSI, hospital-acquired bloodstream infection; CLABSI, central line-associated bloodstream infection.

HABSIs were primarily central-line associated (24.2%) or were of urinary tract (14.9%) or pulmonary origin (9.7%). An important proportion of HABSIs (27.8%) were of unknown origin. The most common microorganisms were coagulase-negative staphylococci (CNS, 18.6%), E. coli (18.0%), and S. aureus (11.6%), followed by Enterococcus (7.8%), Klebsiella (7.1%), Enterobacter (6.1%), Candida (5.7%), and Pseudomonas spp (5.6%). Approximately 10% of HABSIs were polymicrobial (9.7%).

Our mixed-effects regression analysis identified a decrease in total HABSI rates (IRR, −0.5%; 95% CI, −0.8 to −0.1; P = .006). However, this effect size was small and the incidence trend demonstrated cyclical, fluctuating rates (Appendix 3 online). This trend was nonsignificant within the subgroup of university hospitals (IRR, −0.6%; 95% CI, −1.3 to −0.1; P = .08). CLABSI rates did not decrease either hospital-wide (IRR, +0.1%; 95% CI, −0.6 to 0.8; P = .73) or in intensive care (IRR, −0.6%; −1.7 to 0.6; P = .32). The rate of HABSIs of unknown origin decreased over the year (IRR, −5.8%; 95% CI, −6.6 to −5.1; P < .001) with concomitant increases in HABSIs of all secondary origins.

Compared with other hospitals, university hospitals demonstrated higher infection rates for hospital-wide total HABSIs (IRR, 2.3; 95% CI, 1.7–3.2; P < .001), CLABSIs (IRR, 3.4; 95% CI, 2.0–5.9; P < .001), and total ICU HABSIs (IRR, 1.4; 95% CI, 1.0–1.96; P = .04). ICU CLABSIs were not significantly higher among university hospitals (IRR, 1.6; 95% CI, 0.95–2.6; P = .08).

Although the total rate exhibited minimal change, HABSI rates changed significantly in opposite directions for gram-positive and gram-negative bacteria. Gram-negative infections increased (IRR, +1.0%; 95% CI, 0.6–1.4; P < .001), whereas gram-positive HABSIs decreased (IRR, −1.8; 95% CI, −2.2 to −1.3; P < .001) (Appendix 4 online). This finding corresponds with an increase in the gram-negative proportion from 42.7% to 54.1%. The largest incidence rate changes were increases in E. coli (IRR, +2.8%; 95% CI, 2.2–3.3; P < .001) and decreases in CNS (IRR, −4.5%; 95% CI, −5.2 to −3.8; P < .001) (Fig. 1A). Although the absolute increase was less prominent, Klebsiella pneumoniae (IRR, +4.4%; 95% CI, 3.4–5.4; P < .001) and Enterococcus faecium (IRR, +10.9%; 95% CI, 9.0–12.8; P < .001) demonstrated relative rate increases that were large compared to their baseline rate. Notably, E. faecium displayed a nearly 10-fold increase among university hospitals (0.11–0.96 per 10,000 patient days). The low rates of Enterobacter aerogenes decreased further over the study period (IRR, −6.1%; 95% CI, −7.4 to −4.7; P < .001). Although statistically significant, the incidence rates of S. aureus (IRR, −0.8%; 95% CI, −1.5 to −0.2; P = .02) and Candida spp (IRR, −1.3%; 95% CI, −2.2 to −0.4; P = .004) did not exhibit a large absolute decrease. Hospital-wide trends of Klebsiella oxytoca, Enterobacter cloacae, P. aeruginosa, Bacteroides spp and Candida glabrata remained stable.

Fig. 1.

A. Hospital-wide hospital-acquired bloodstream infection incidence, per microorganism (2000–2014). Incidence of hospital-wide bloodstream infections for microorganisms that demonstrated significant annual incidence rate ratio changes over the years (see online appendix 4). Note. CNS, coagulase-negative staphylococci. B. Intensive care hospital-acquired bloodstream infection incidence, per microorganism (2000–2014). Incidence of intensive care bloodstream infections for microorganisms that demonstrated significant annual incidence rate ratio changes over the years (see online appendix 4). Note. CNS, coagulase-negative staphylococci.

Among ICUs, the CNS rate demonstrated a similar decrease, whereas gram-negative and Candida rates remained stable. However, E. coli, K. pneumoniae, E. cloacae, and E. faecium increased, whereas P. aeruginosa and Enterobacter aerogenes decreased (Fig. 1B).

Methicillin-resistant S. aureus accounted for 19.0% of isolates. Vancomycin-resistant Enterococcus was uncommon (2.7%). Third-generation cephalosporin resistance was common among E. coli (16.2%), Klebsiella spp (24.8%), Enterobacter spp (48.2%) and P. aeruginosa (14.6%). Carbapenem resistance was present among P. aeruginosa (18.2%) and Acinetobacter (8.1%) but was uncommon among Enterobacteriaceae (<2.5%).

Among total HABSI incidence, goodness of fit identified 2 severe regression model outliers. However, removal thereof did not meaningfully affect the results (Appendix 5 online). Sensitivity analysis was performed to assess selection bias associated with varying participation criteria (Appendix 6 online). The hospital-wide rise in gram-negative and decline in gram-positive and fungal rates remained significant among all cohorts, albeit with slightly different effect sizes. Hospitals participating ≥10 years showed attenuation of the increasing gram-negative rate, stronger rate reductions among gram-positive HABSI, and a nonsignificant trend toward lower CLABSI rates (IRR, −0.7%; 95% CI, −1.4 to 0.1; P = .07).

Discussion

In this study, we analyzed HABSI trends over 15 years of surveillance (2000–2014) in 110 hospitals. Although the total HABSI rate was slightly decreasing, there was a clear increase in gram-negative pathogen incidence represented by E. coli and K. pneumoniae. Enterococcus faecium demonstrated important rate increases, most markedly within ICUs and university-affiliated hospitals. Among ICUs the gram-negative HABSI rate remained stable yet the proportions of specific microorganisms changed with increases in E. coli, K. pneumoniae, and E. cloacae and decreases in P. aeruginosa and E. aerogenes. Notably, within the ICU, the incidence of E. faecium rose to the same level as K. pneumoniae and E. cloacae.

CLABSI rates did not decrease either hospital-wide or in ICUs. University hospitals displayed hospital-wide CLABSI rates triple that of nonuniversity hospitals, yet this difference was not significant among ICUs. This higher infection rate among university hospital wards may reflect an area for improvement through CLABSI prevention initiatives. The subgroup of hospitals participating at least 10 years exhibited a nonsignificant trend toward decreasing CLABSI rates, which aligns with evidence that CLABSI reduction can be achieved through long-term surveillance as part of a quality improvement initiative.Reference Blot, Bergs, Vogelaers, Blot and Vandijck 8

The strengths of this study include the long-term surveillance, hospital-wide data collection, nationwide participation, and mixed-effects modeling to account for confounding factors. Countrywide surveillance programs have been shown to estimate valid BSI incidences when data are collected during random trimesters of the year.Reference Fontela, Quach, Buckeridge, Pai and Platt 9 In this study, hospitals could choose which trimester to report from, but there was no statistically significant difference in reporting rates between trimesters. Furthermore, the hospital cohort reporting data year-round exhibited similar rate changes compared to cohorts that did not.

The limitations of this study include missing data on confounding factors such as catheter days, blood culturing frequency, and recent ICU admission. No distinction could be made between infections acquired in the ICU versus HABSIs that led to ICU admission. Finally, because it was not mandatory to perform surveillance consecutively across trimesters or years, time-series regression analysis with temporal autocorrelation was not possible without exclusion of most of the data, at the expense of its external validity. Unfortunately, because resistance data were only available during the final 2 surveillance years, resistance trends could not be analyzed in this cohort.

Overall, CNSs comprised the majority of CLABSI and demonstrated rate decreases both in intensive care and hospital-wide, yet CLABSI incidence appeared to remain stable. Although the definition requires at least 2 positive cultures with clinical symptoms, this CNS decline could partially represent an improved recognition of skin contaminants instead of a true incidence decline.

This study identified an increasing HABSI rate for E. coli, K. pneumoniae, and E. faecium, which represents an increased treatment burden and the underlying epidemiology of hospital-acquired infections. This is alarming because of these microorganisms’ resistance levels to fluoroquinolones, third-generation cephalosporins, and amoxicillin. Escherichia coli, K. pneumoniae, E. cloacae, and E. faecium increases were mirrored in the ICU, reflecting their clinical importance as either representing sepsis requiring critical care admission or infection of susceptible ICU patients.Reference Blot, Vandewoude and Colardyn 10 – Reference Blot, Vandewoude and Hoste 12 Other surveillance studies have likewise identified rises in E. coli and K. pneumoniae Reference Holmbom, Giske and Fredrikson 13 – Reference Søgaard, Nørgaard, Dethlefsen and Schønheyder 16 and increasing proportions of E. coli resistant to third-generation cephalosporins among 12 of 30 European countries. 4 Increasing trends of antimicrobial-resistant microorganisms threatens to complicate proper early and appropriate antibiotic treatment.Reference McGregor, Rich and Harris 17 Empiric antibiotic therapy should be based on local epidemiology, and these changing trends warrant further monitoring of microorganism incidence and resistance patterns.

Supplementary material

To view supplementary material for this article, please visit https://doi.org/10.1017/ice.2019.59

Author ORCIDs

Koen Blot, 0000-0002-0847-0133

Acknowledgments

Financial support

S.B. holds a research mandate of the Specific Research Fund at Ghent University.

Conflicts of interest

D.V. has received an institutional grant for work under consideration for publication from Pfizer and has been a consultant for Astellas, Pfizer, and Tibotec. All other authors report no potential conflicts.

References

Blot, SI, Depuydt, P, Annemans, L, et al. Clinical and economic outcomes in critically ill patients with nosocomial catheter-related bloodstream infections. Clin Infect Dis 2005;41:1591–1598.CrossRefGoogle ScholarPubMed
Vrijens, F, Hulstaert, F, Van de Sande, S, Devriese, S, Morales, I, Parmentier, Y. Hospital-acquired, laboratory-confirmed bloodstream infections: linking national surveillance data to clinical and financial hospital data to estimate increased length of stay and healthcare costs. J Hosp Infect 2010;75:158–162.CrossRefGoogle ScholarPubMed
Pien, BC, Sundaram, P, Raoof, N, et al. The clinical and prognostic importance of positive blood cultures in adults. Am J Med 2010;123:819–828.CrossRefGoogle ScholarPubMed
European Centre for Disease Prevention and Control. Antimicrobial resistance surveillance in Europe 2015. Annual Report of the European Antimicrobial Resistance Surveillance Network (EARS-Net). Stockholm: ECDC; 2017.Google Scholar
WIV-ISP. Surveillance nationale des septicémies dans les hôpitaux belges 2017. Sciensano website. http://www.nsih.be/surv_sep/beschrijving_fr.asp. Published 2017. Accessed March 8, 2019.Google Scholar
Garner, JS, Jarvis, WR, Emori, TG, Horan, TC, Hughes, JM. CDC definitions for nosocomial infections, 1988. Am J Infect Control 1988;16:128–140.CrossRefGoogle ScholarPubMed
Central line-associated bloodstream infection (CLABSI) event protocol. Center for Disease Control and Prevention website. http://www.cdc.gov/nhsn/PDFs/pscManual/4PSC_CLABScurrent.pdf. Updated January 2019. Accessed March 8, 2019.Google Scholar
Blot, K, Bergs, J, Vogelaers, D, Blot, S, Vandijck, D. Prevention of central line-associated bloodstream infections through quality improvement interventions: a systematic review and meta-analysis. Clin Infect Dis 2014;59:96–105.CrossRefGoogle ScholarPubMed
Fontela, PS, Quach, C, Buckeridge, D, Pai, M, Platt, RW. Surveillance length and validity of benchmarks for central line-associated bloodstream infection incidence rates in intensive care units. PLoS One 2012;7:e36582–6. doi: 10.1371/journal.pone.0036582.CrossRefGoogle ScholarPubMed
Blot, SI, Vandewoude, KH, Colardyn, FA. Evaluation of outcome in critically ill patients with nosocomial enterobacter bacteremia: results of a matched cohort study. Chest 2003;123:1208–13.CrossRefGoogle ScholarPubMed
Blot, SI, Vandewoude, KH, Colardyn, FA. Clinical impact of nosocomial Klebsiella bacteremia in critically ill patients. Eur J Clin Microbiol Infect Dis 2002;21:471–473.Google ScholarPubMed
Blot, S, Vandewoude, K, Hoste, E, et al. Absence of excess mortality in critically ill patients with nosocomial Escherichia coli bacteremia. Infect Control Hosp Epidemiol 2003;24:912–915.CrossRefGoogle ScholarPubMed
Holmbom, M, Giske, CG, Fredrikson, M, et al. 14-Year survey in a Swedish county reveals a pronounced increase in bloodstream infections (BSI). Comorbidity—an independent risk factor for both BSI and mortality. PLoS One 2016;11:e0166527. doi: 10.1371/journal.pone.0166527.CrossRefGoogle Scholar
Wilson, J, Elgohari, S, Livermore, DM, et al. Trends among pathogens reported as causing bacteraemia in England, 2004–2008. Clin Microbiol Infect 2011;17:451–458.CrossRefGoogle Scholar
Buetti, N, Marschall, J, Atkinson, A, Kronenberg, A, Swiss Centre for Antibiotic Resistance (ANRESIS). National bloodstream infection surveillance in Switzerland 2008–2014: different patterns and trends for university and community hospitals. Infect Control Hosp Epidemiol 2016;37:1060–1067.CrossRefGoogle ScholarPubMed
Søgaard, M, Nørgaard, M, Dethlefsen, C, Schønheyder, HC. Temporal changes in the incidence and 30-day mortality associated with bacteremia in hospitalized patients from 1992 through 2006: a population-based cohort study. Clin Infect Dis 2011;52:61–69.CrossRefGoogle ScholarPubMed
McGregor, JC, Rich, SE, Harris, AD, et al. A systematic review of the methods used to assess the association between appropriate antibiotic therapy and mortality in bacteremic patients. Clin Infect Dis 2007;45:329–337.CrossRefGoogle ScholarPubMed
Figure 0

Table 1.Hospital Characteristics and Hospital-Acquired Bloodstream Infection Incidence

Figure 1

Fig. 1.A. Hospital-wide hospital-acquired bloodstream infection incidence, per microorganism (2000–2014). Incidence of hospital-wide bloodstream infections for microorganisms that demonstrated significant annual incidence rate ratio changes over the years (see online appendix 4). Note. CNS, coagulase-negative staphylococci. B. Intensive care hospital-acquired bloodstream infection incidence, per microorganism (2000–2014). Incidence of intensive care bloodstream infections for microorganisms that demonstrated significant annual incidence rate ratio changes over the years (see online appendix 4). Note. CNS, coagulase-negative staphylococci.

Supplementary material: File

Blot et al. supplementary material

Blot et al. supplementary material 1

Download Blot et al. supplementary material(File)
File 510.5 KB