Baylor University Medical Center Proceedings October 2017 - 395

Clinical uptake of antimicrobial stewardship recommendations
following Nanosphere Verigene Blood Culture Gram-negative
reporting
Aaron Belknap, MD, Daniel S. Grosser, MD, Daniel A. Hale, MD, Benjamin J. Lang, MD, Peter Colley, PharmD,
Raul Benavides, MD, and Neelam Dhiman, PhD

We performed a retrospective chart review of patients to determine if
the Verigene Gram-negative blood culture (BC-GN) results would lead to
earlier deescalation of empiric therapy for inpatients with GN bacteremia
with Citrobacter spp., Enterobacter spp., Klebsiella spp., and Escherichia
coli to appropriate targeted coverage. A total of 899 records were reviewed from April 2014 to February 2016 from three institutions within
the Baylor Scott & White Health network. The cases were reviewed for
initial antibiotic coverage, timing of Verigene results, change in antibiotic coverage, and how these changes related to the timing of Verigene
results. The lab reported the BC-GN results and final conventional susceptibility results within 2.5 ± 1.3 and 73.6 ± 40.0 hours from the
Gram stain, respectively. Overall, 29.1% of patients were transitioned
from empiric to targeted therapy at 12.2 ± 13.5 hours in response to
BC-GN results, which was significantly earlier (P < 0.001) than results
by conventional methods. After accounting for patients already on targeted therapy, polymicrobial infections, and patients deceased or lost to
follow-up, we identified antibiotic stewardship opportunities in ∼28% of
GN infections. Further subanalysis demonstrated site-specific differences
in the uptake of stewardship recommendations, whereby 32.4%, 50.5%,
and 15.0% of cases at different hospitals demonstrated the expected
change in antibiotics. These results suggest that Verigene had the expected impact in a third of the cases and the results reporting algorithm
minimized the real-time involvement of the pharmacist while maintaining
optimal patient management. However, this impact varied substantially
by clinical site and was tempered by variable initial antibiotic coverage
and clinician response.

ﬀective and prompt antimicrobial therapy is crucial for
the survival of patients with sepsis (1-3). Rapid molecular technologies aimed at decreasing the time to identiﬁcation and susceptibility results have recently entered
the market (4, 5). The Verigene Gram-Negative Blood Culture
(BC-GN) assay is one such rapid panel that detects seven genera,
four species, and six resistance markers directly from positive
blood cultures (6). We adopted the BC-GN assay in our laboratory to serve the needs of the multisite Baylor Scott and White
Health network. For eﬀective implementation, we worked with
the Antimicrobial Stewardship Program (ASP) to provide recommended changes to antibiotic therapy based on the species
identiﬁcation and local pathogen susceptibility patterns in the
Proc (Bayl Univ Med Cent) 2017;30(4):395-399

result comments. The goal was to improve patient outcomes by
guiding transition of the empiric treatment (vancomycin and
piperacillin/tazobactam) to targeted therapy, while simultaneously preventing antibiotic overuse and the development of
antibiotic resistance. This study aimed to determine if decreasing the time to blood culture result using the BC-GN system
paired with well-deﬁned ASP-recommended therapy changes
would impact clinical outcomes. The primary objective was to
evaluate the clinical uptake and utilization of the BC-GN results
by measuring the reduction in time from positive blood culture
to the deescalation of empiric therapy and the switch to the ﬁrst
dose of appropriate antibiotics per the resulting algorithm. As
previously published, an algorithm designed using electronic
communications and minimum pharmacist intervention was
used (7, 8). We also determined if there were any site-speciﬁc
diﬀerences in the uptake of stewardship recommendations
within the network.
METHODS
This was a multisite, retrospective chart review of patients
admitted with Gram-negative (GN) bacteremia to Baylor
University Medical Center, a 1079-bed tertiary referral center in Dallas, Texas, and two smaller acute care hospitals also
within the Baylor Scott and White Health network, 574-bed
Baylor All Saints and 296-bed Baylor Irving. Blood cultures
were performed at the aﬃliated reference laboratory, med
fusion, in Lewisville, Texas. Molecular testing on the positive
blood cultures was performed using Verigene BC-GN panel
(Nanosphere, Inc., Northbrook, IL). This study was approved
by the institutional review board.
Medical records were reviewed from April 2014 to February
2016 from three sites for initial antibiotic coverage, timing of
the appearance of Verigene results in the electronic health record
(EHR), change in antibiotic therapy (if any), and how these
changes compared to the timing of Verigene results. Times were
From the Departments of Pathology and Laboratory Medicine (Belknap, Grosser,
Hale, Lang, Benavides) and Pharmacy (Colley), Baylor University Medical Center
at Dallas; and med fusion Laboratory, Lewisville, Texas (Dhiman).
Corresponding author: Neelam Dhiman, PhD, Scientific Director, Molecular
Infectious Diseases, med fusion, 2501 South State Highway 121, Suite 1100,
Lewisville, TX 75067 (e-mail: ndhiman@medfusionsvs.com).
395

Table of Contents for the Digital Edition of Baylor University Medical Center Proceedings October 2017