Evaluation of Risk Factors for Invasive Carbapenem Resistant Enterobacteriaceae Infections and Resultant Mortality in Atlanta, 2011-2015 Open Access

Sexton, Mary Elizabeth (2017)

Permanent URL: https://etd.library.emory.edu/concern/etds/hd76s0880?locale=en



Carbapenem-resistant Enterobacteriaceae (CRE) have become an increasing public health concern, given limited treatment options and associated high mortality. Understanding risk factors for CRE infections of sterile sites (invasive infection) and for mortality may have important implications for prevention. A retrospective cohort study was performed using the Georgia Emerging Infections Program (EIP) CRE surveillance data to evaluate risk factors associated with invasive infection and mortality.


The study population comprised incident CRE cases from 8/2011 to 12/2015. A case required isolation, from urine or a normally-sterile site, of E. coli, Klebsiella spp., or Enterobacter spp. that was carbapenem-nonsusceptible and resistant to third-generation cephalosporins. Cases were incident if the patient resided in the surveillance area and had no prior positive cultures in 30 days, with only the first incident case for each patient included. Cases were considered invasive infection if the patient had a sterile site culture positive during the 30-day period. Mortality was defined as in-hospital mortality for admitted patients, or 30-day mortality in long-term care facility or dialysis patients. Demographic characteristics and CRE risk factor prevalence were compared using chi-square analysis between patients with sterile site versus urinary cultures positive, and patients with fatal versus non-fatal outcomes. Multivariable logistic regression was performed to evaluate whether particular risk factors were associated with invasive infection and mortality.


Of 567 CRE patients, 91 (16.0%) had an invasive infection and 476 (84.0%) had only urinary cultures positive. Central line presence, indwelling devices, and recent surgery were associated with invasive infection in multivariable analysis. The overall mortality rate was 9.0% (51/567), including 30 deaths in patients with urinary cultures positive (6.3%) and 21 deaths in patients with invasive infection (23.1%). In multivariable analysis, ICU stay, a central line, or invasive infection predicted mortality.


Device use was common and was associated with invasive infection. Patients with invasive infection and markers of severity of illness were more likely to die. Future research should focus on whether removal of unnecessary devices decreases risk of invasive infection with CRE, and whether early identification and initiation of appropriate antibiotic therapy in high-risk patients decreases mortality.

Table of Contents

Section Pages

Introduction 1-2

Background 3-10

Methods 11-17

Results 18-24

Discussion 25-34

References 35-37

Tables and Figures 38-44

Table 1 39

Figure 1 40

Table 2 41

Figure 2 42

Table 3 43

Table 4 44

About this Master's Thesis

Rights statement
  • Permission granted by the author to include this thesis or dissertation in this repository. All rights reserved by the author. Please contact the author for information regarding the reproduction and use of this thesis or dissertation.
  • English
Research Field
Committee Chair / Thesis Advisor
Committee Members
Last modified

Primary PDF

Supplemental Files