Infectious Diseases in Critical Care Medicine

(ff) #1

acquisition, it is important to determine the exact time of colonization or infection by VRE, to
use controls that are negative for VRE [as opposed to controls positive for vancomycin-
susceptible enterococci (VSE)], and to use multivariable statistics to identify independent risk
factors. Some studies of risk factors have included ICUs in addition to other areas of the
hospital (Table 4), and others have been limited to ICUs (Table 5).
Several of the studies included in Tables 4 and 5 have identified a significant relationship
between prior administration of an antimicrobial agent and acquisition of VRE. Drugs listed
included cephalosporins, metronidazole, vancomycin, carbapenems, ticarcillin–clavulanate,
and quinolones. The antibiotic most often identified as a risk factor was vancomycin. In an
extensive study of the effects of antimicrobial agents on fecal flora, it was found that
antianaerobic antibiotics promoted high-density colonization of stool with VRE (157).
Administration of vancomycin had no effect on the concentration of VRE in stool. Although
antianaerobic agents increased the concentration of VRE in stool, it is unclear whether these
agents or vancomycin predispose to acquisition of VRE.
Several case-control studies have shown that vancomycin is a risk factor for acquisition
of VRE. In an assessment of studies showing a relationship between vancomycin and
acquisition of VRE by meta-analysis, the authors concluded that the apparent relationship
between administration of vancomycin and colonization with VRE is due to selection of VSE as
the reference group, confounding by duration of hospitalization and publication bias (158).
However, several studies have shown a significant relationship between receipt of vancomycin
and colonization with VRE (150,151,159). In these studies the reference group was appropri-
ately selected (VRE-negative patients and not VSE-culture positive) and duration of
hospitalization was included to control for confounding due to longer exposure time. Thus,
the issue of whether vancomycin is a risk factor for acquisition of VRE is unsettled.
Risk factors from Tables 4 and 5 that appear multiple times are use of antacids and
enteral feedings. One study noted that a length of stay of less than or equal to five days in an
MICU was protective against VRE acquisition, whereas another study observed that
hospitalization for more than one week prior to MICU admission was a risk factor for
acquisition of VRE. In summary the most frequently identified risk factors for acquiring VRE
from these studies were administration of antibiotics and antacids and enteral feedings.


Neonatal ICUs.There are seven reports of outbreaks of VRE in NICUs (132–134,160–163).
Analytical epidemiology was used in two of the studies to identify risk factors for acquisition
of VRE (132,163). The first study examined a large number of variables by univariate analysis
and found many variables apparently related to VRE colonization. However, multivariable
analysis by logistic regression identified days of antimicrobial therapy (OR 1.21, 95% CI 1.045–
1.400,p= 0.01) and birth weight (OR 0.92, 95% CI 0.862–0.979, p= 0.009) as the only
independent associations with acquisition of VRE. The second study also examined a large
number of variables, but on multivariable analysis, no risk factors were identified for
colonization or infection by VRE. Additional studies are needed to further define the variables
associated with acquisition of VRE in this population.


Table 4 Risk Factors for Acquisition of VRE from Studies of Mixed Patient Populations


Publications Risk Factors


Adjusted Odds
Ratio (95% CI) pValue

Loeb et al. (148) Cephalosporin use 13.8 (2.5–76.3) 0.01
Byers et al. (149) Proximity to an unisolated patient 2.04 (1.32–3.14) 0.0014
History of major trauma 9.27 (1.43–60.3) 0.020
Therapy with metronidazole 3.04 (1.05–8.77) 0.040


Cetinkaya et al. (150) Vancomycin use 3.2 (1.7–6.0) 0.0003
Gastrointestinal bleedinga 0.26 (0.08–0.79) 0.02
Presence of central venous lines 2.2 (1.04–4.6) 0.04
Antacid use 2.9 (1.5–5.6) 0.002
Mean daily dose of Vicodin


Ra
0.0003

aProtective factors.


Abbreviation:VRE, vancomycin-resistant enterococci.


114 Mayhall

Free download pdf