PRACTICAL GUIDANCE FOR

CLINICAL MICROBIOLOGY

crossm

A Comprehensive Update on the Problem of Blood Culture

Contamination and a Discussion of Methods for Addressing
the Problem
Gary V. Doern,a Karen C. Carroll,b Daniel J. Diekema,c Kevin W. Garey,d Mark E. Rupp,e Melvin P. Weinstein,f Daniel J. Sextong

a Department of Pathology, University of Iowa Carver College of Medicine, Iowa City, Iowa, USA
b Division of Medical Microbiology, Department of Pathology, John Hopkins University School of Medicine, Baltimore, Maryland, USA
c Division of Infectious Diseases, Department of Medicine and Department of Pathology, University of Iowa Carver College of Medicine, Iowa City, Iowa, USA
d Department of Pharmacy Practice and Translational Research, University of Houston College of Pharmacy, Houston, Texas, USA
e Division of Infectious Diseases, Department of Medicine, University of Nebraska Medical Center, Omaha, Nebraska, USA
f
Department of Pathology and Laboratory Medicine, Department of Medicine, Robert Wood Johnson Medical School, New Brunswick, New Jersey, USA
g Division of Infectious Diseases, Department of Medicine, Duke University School of Medicine, Durham, North Carolina, USA

SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
THE MICROBIOLOGY OF BLOOD CULTURE CONTAMINATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
Magnitude of the Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
Sources of Blood Culture Contamination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
Bacteria Associated with Contaminated Blood Cultures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

Downloaded from https://journals.asm.org/journal/cmr on 24 July 2024 by 2409:40f4:1:3c3a:456a:7f5a:e81e:35f9.

THE LABORATORY IMPACT OF BLOOD CULTURE CONTAMINATION . . . . . . . . . . . . . . . . . . . . . 3
THE CLINICAL CONSEQUENCES OF BLOOD CULTURE CONTAMINATION . . . . . . . . . . . . . . . . 5
THE ECONOMIC CONSEQUENCES OF BLOOD CULTURE CONTAMINATION . . . . . . . . . . . . . 7
PREVENTION OF BLOOD CULTURE CONTAMINATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Patient Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Skin Antisepsis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Blood Culture Bottle Disinfection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Blood Culture Collection Site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Single Needle versus Double Needle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Sterile Gloves and Hand Hygiene . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Blood Culture Kits and Standard Procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Sterile Drapes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Blood Sampling and Volume . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Phlebotomy Teams/Education . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Multidisciplinary/Multimodal Performance Improvement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Surveillance and Feedback . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Initial Specimen Diversion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
Unanswered Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
ACKNOWLEDGMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
AUTHOR BIOS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

Citation Doern GV, Carroll KC, Diekema DJ,

SUMMARY In this review, we present a comprehensive discussion of matters re- Garey KW, Rupp ME, Weinstein MP, Sexton DJ.
lated to the problem of blood culture contamination. Issues addressed include 2019. A comprehensive update on the
problem of blood culture contamination and a
the scope and magnitude of the problem, the bacteria most often recognized as discussion of methods for addressing the
contaminants, the impact of blood culture contamination on clinical microbiol- problem. Clin Microbiol Rev 33:e00009-19.
ogy laboratory function, the economic and clinical ramifications of contamina- https://doi.org/10.1128/CMR.00009-19.
Copyright © 2019 American Society for
tion, and, perhaps most importantly, a systematic discussion of solutions to the
Microbiology. All Rights Reserved.
problem. We conclude by providing a series of unanswered questions that per- Address correspondence to Gary V. Doern,
tain to this important issue. [email protected].
Published 30 October 2019
KEYWORDS blood culture contamination

January 2020 Volume 33 Issue 1 e00009-19 Clinical Microbiology Reviews cmr.asm.org 1

Doern et al. Clinical Microbiology Reviews

INTRODUCTION

B lood cultures have long been recognized as one of the most important tests
performed in clinical microbiology laboratories. Unfortunately, blood cultures are
frequently contaminated (1–5). There is a substantial cost associated with contami-
nated blood cultures, a defined impact on clinical microbiology laboratory practice,
and, perhaps most importantly, the potential for negative outcomes among patients
from whom blood cultures have been obtained (2, 6–13).
The intent of this review is to systematically address five issues relevant to the
problem of contaminated blood cultures. These include the scope and magnitude of
the problem, its impact on laboratory practice, the clinical implications of contaminated
blood cultures, associated costs, and, finally and perhaps most importantly, potential
solutions to the problem. Inherent in this discussion is the long-held notion that a
certain percentage of blood cultures will be contaminated irrespective of what efforts
are made to prevent contamination (14). Currently, health care institutions in the
United States are held to a performance standard of ⬍3% rates of blood culture
contamination (15, 16). Clearly, as will be shown in this review, recent advances in
practice can lead to much lower rates of contamination. If this is true, in view of the
substantial negative consequences of contaminated blood cultures, the question arises,
should this arbitrary 3% contamination rate threshold be reconsidered?

THE MICROBIOLOGY OF BLOOD CULTURE CONTAMINATION

Magnitude of the Problem
There are two metrics that can be used to quantify the extent of blood culture
contamination: the percentage of all positive blood cultures that yield organisms
judged to be contaminants (i.e., overall blood culture contamination rate) and the
percentage of all blood cultures obtained that are contaminated. Using both of these
approaches, the reported frequency with which blood culture contamination occurs

Downloaded from https://journals.asm.org/journal/cmr on 24 July 2024 by 2409:40f4:1:3c3a:456a:7f5a:e81e:35f9.

has varied in published studies. In a series of large clinical studies examining blood
cultures and bacteremia over 4 recent decades, Weinstein and colleagues found that
one-third to one-half of all positive blood cultures were judged by infectious disease
physicians to represent contamination (2, 3, 17). Other studies have reported lower
rates. Story-Roller and Weinstein found that 26% of all positive blood cultures were
judged to contain contaminants (4). The overall contamination rate at the university
hospital where this study was done was 3.9% (4). Washer et al. found that 13% of all
positive blood cultures represented contamination and that overall contamination
rates were 0.8% when blood for culture was obtained peripherally by phlebotomists
who performed venipuncture (18). Rupp et al. reported that 23% of all positive blood
cultures represented contamination and that overall contamination rates were 1.8%
during a defined study period (9). Interestingly, the institutional contamination rate in
this study increased to 2.8% 6 months following conclusion of the study and reversion
to standard practice (9). Other studies have noted that 20 to 56% of all positive blood
cultures are found to be contaminated (19, 20), with overall contamination rates of 0.6
to 12.5% (21, 22). Clearly, a number of different variables impact the scope and
magnitude of the problem of blood culture contamination.

Sources of Blood Culture Contamination

The potential sources of blood culture contamination are numerous and varied (23,
24). Poor technique by individuals obtaining blood cultures is a major factor (7, 25–27).
So is insufficient disinfection of the skin, as the principal source of contaminants is
commensal bacteria that colonize the skin (18, 20–22, 28). Collection of blood through
indwelling vascular catheters is also problematic. Perhaps related to the difficulty in
obtaining adequate antisepsis in the port (i.e., access) area of the intravenous device at
the site where blood is obtained for culture, blood cultures that are obtained by
peripheral venipuncture have generally been reported to be associated with lower
contamination rates than those obtained from indwelling intravenous catheters
(29, 30).

January 2020 Volume 33 Issue 1 e00009-19 cmr.asm.org 2

Blood Culture Contamination Clinical Microbiology Reviews

Yet another contributing factor in contamination is the type of broth medium used
for blood cultures. Several decades ago, manufacturers of blood culture media and
detection systems developed media that contained antibiotic-binding resins. In clinical
studies comparing these so-called resin media to standard broth media without resins,
the resin-containing media provided an increased yield (i.e., growth) of both pathogens
and contaminants (31–36). This was particularly notable for both Staphylococcus aureus
and coagulase-negative staphylococci (CoNS), the latter most often representing con-
tamination.
Lastly, the standard of practice for obtaining blood cultures by venipuncture un-
derwent a substantial change as the importance of blood-borne pathogens (e.g., HIV)
and needlestick injuries to health care workers became widely recognized during the
1990s. Prior to the HIV era, blood cultures were often obtained by a 2-needle technique
wherein a sterile needle was used for venipuncture and then, prior to inoculation of the
blood culture vials, this needle was removed and a 2nd sterile needle was attached to
the syringe for inoculation of blood into the culture vials. To minimize exposure to
“sharps” that might result in injury, studies were conducted to determine whether a
single-needle technique could be used. Three such studies were published and showed
no statistically significant increase in contamination rates (37–39). However, a meta-
analysis of these studies reported that single-needle blood cultures were associated
with contamination rates of 3.7%, whereas contamination rates were only 2.0% when
a two-needle technique was used (40).

Bacteria Associated with Contaminated Blood Cultures

Over many years, a number of clinical laboratory investigations have helped inform
our current understanding of the microorganisms most likely to represent contamina-
tion when isolated from the blood (2, 3, 17, 41). These include most species of CoNS,
most species of Corynebacterium and related genera, Bacillus spp. other than Bacillus
anthracis, Micrococcus spp., and Cutibacterium acnes and related species. These organ-

Downloaded from https://journals.asm.org/journal/cmr on 24 July 2024 by 2409:40f4:1:3c3a:456a:7f5a:e81e:35f9.

isms may be considered contaminants unless recovered from multiple blood cultures
obtained in sequence, in which case, careful assessment of patients and additional labo-
ratory information is required in defining significance (or lack thereof). Enterococci,
viridans group streptococci, and Clostridium spp. are of variable clinical significance (i.e.,
may be either contaminants or true pathogens) when recovered from blood cultures.
In contrast, microorganisms such as Staphylococcus aureus, Streptococcus pneumoniae,
beta-hemolytic streptococci, Listeria monocytogenes, Escherichia coli and other mem-
bers of the Enterobacteriaceae, Pseudomonas aeruginosa, Neisseria meningitidis, Haemo-
philus influenzae, anaerobic Gram-negative rods (e.g., Bacteroides spp. and Fusobacte-
rium spp.), and Candida species rarely represent contamination.

THE LABORATORY IMPACT OF BLOOD CULTURE CONTAMINATION

According to the Clinical and Laboratory Improvements Act (CLIA), the clinical
microbiology laboratory is responsible for the preanalytical phase of testing related to
the diagnosis of infectious diseases (42). This includes the selection, collection, and
transport of specimens. Therefore, the clinical laboratory plays a central role in provid-
ing instructions for preventing contamination during blood culture procurement. Table
1 lists key elements for optimization of blood cultures that may impact contamination.
The last section of this review delineates in detail the various factors that must be
considered at the time of blood culture procurement with the aim of minimizing
contamination. Specific instructions on the proper collection of blood cultures should
be provided in laboratory directories or specimen collection guidelines.
In addition, the clinical microbiology laboratory plays an essential role in providing
information to care providers when blood culture isolates likely to represent contam-
ination are recovered. Several algorithms have been described that provide criteria
which can be used in the laboratory for determining whether isolates of bacteria
frequently found to contaminate blood cultures, such as those mentioned above, are
likely to be contaminants when recovered in blood cultures (1, 30, 43–45). In making

January 2020 Volume 33 Issue 1 e00009-19 cmr.asm.org 3

Doern et al. Clinical Microbiology Reviews

TABLE 1 Essential elements for reduction of contamination and optimization of blood culturesa
Element Comment
Preanalyticalb
Body site for specimen collection Peripheral venipuncture preferred
Preparation of the venipuncture site Use of alcoholic chlorhexidine gluconate and/or tincture of iodine with 70% alcohol;
iodophors (povidone iodine) are not recommended
Specimen procurement
Hand hygiene Use of trained phlebotomy teams
Sterile gloves Adherence to aseptic practices
Selective use of sterile drapes Avoid cultures obtained through lines
No. of bottles A minimum of two separate draws; usually one aerobic and anaerobic bottle per
draw
Vol of blood (in adults) 8–10 ml of blood per bottle; 20–30 ml per venipuncture
Transport to the laboratory Transport time should be as rapid as possible at room temperature; follow
manufacturers’ guidelines.

Analytical
Use of continuously monitored blood culture systems Reduces turnaround time
Timely Gram stain performance 24/7 processing of positives
Rapid diagnostics Provides faster identification of contaminants and true pathogens
Timely ASTc Resistance marker detection direct from bottle AST

Postanalytical (result interpretation)

Notify care providers when blood cultures are Establish institutional critical action value policy for reporting
determined to be positive
Define, minimize workup of contaminants Clearly report as contaminants organisms that meet criteria as such; do not perform
AST
Final report in electronic patient record Concise, pertinent report highlighting pathogen and resistance
aDerived from references 1, 4, 19, 20, 25–28, 40, 46, 53, 55, 58–63, 67, 68, and 71.
bLaboratories should have specimen collection guidelines available that address the preanalytical protocols for specimen procurement, training of phlebotomists or
other staff, and feedback on individual contamination rates.
cAST, antimicrobial susceptibility testing.

Downloaded from https://journals.asm.org/journal/cmr on 24 July 2024 by 2409:40f4:1:3c3a:456a:7f5a:e81e:35f9.

this assessment, in addition to the specific bacterium recovered, the number of blood
cultures obtained and the number positive are both important. (For purposes of this
review, a blood culture is defined as a culture obtained by one venipuncture or line
draw usually consisting of two bottles, one aerobic bottle and one anaerobic bottle.)
Consultation with care providers may be required in determining the significance of a
blood culture isolate when only a single blood culture set is submitted. When multiple
cultures yield the same organism, the positive predictive value for true bacteremia
increases (1, 41, 43, 46). Having made a determination that a blood culture isolate is
likely to be a contaminant, it is incumbent upon the laboratory to provide this
information to care providers both in written culture reports and, when warranted, in
direct communications. For example, the report might state: “Coagulase-negative
Staphylococcus, probable contaminant. No additional workup. Please call the laboratory
within 2 days if susceptibility testing is needed.”
A related consideration is the length of time blood culture bottles are incubated.
There exists at least anecdotal information which suggests that the length of time to
positivity is of some value in discerning the likelihood of an individual blood culture
isolate being a contaminant or of clinical significance (6, 23). The underlying premise is
that growth of pathogens, which are likely to be present in higher concentrations in
blood from patients with bacteremia, will be detected earlier than contaminants, in
which case the magnitude of bacteremia is generally lower (41). This concept may have
some validity, at least with blood culture systems that are characterized by prolonged
incubation periods. However, today, with the widespread use of continuous monitoring
blood culture systems and their attendant incubation periods of only 4 to 5 days, this
notion may not be valid (47). In addition, given the extensive overlap in detection times
between significant isolates and contaminants, length of time to detection cannot, by
itself, be relied upon to accurately predict the clinical significance of individual blood
culture isolates (48).
Continuously monitored blood culture systems and enhancements to culture media

January 2020 Volume 33 Issue 1 e00009-19 cmr.asm.org 4

Blood Culture Contamination Clinical Microbiology Reviews

have improved the recovery and shortened the time to detection of both true patho-
gens and contaminants, thus shortening the time to provision of blood culture results
to care providers (49, 50). Similarly, the availability of novel laboratory technologies to
rapidly identify organisms and, in some cases, their resistance determinants directly
from positive blood cultures also allows for more rapid detection of potential contam-
inants, in some cases by as much as 24 h (51–53). Methods include matrix-assisted laser
desorption ionization–time of flight mass spectrometry (MALDI-TOF MS), nucleic acid
detection, nucleic acid sequencing, and peptide nucleic acid fluorescence in situ
hybridization (PNA-FISH). As stated above, the positive clinical implications of use of
these methods are substantial (54–56). It should be noted, however, that MALDI-TOF
MS, a rapid and relatively inexpensive identification method, is labor-intensive when
applied directly to blood culture broth and, as a result, is most often applied to colonies
after short incubation of subcultures from positive blood cultures (57). In contrast,
nucleic acid-based methods can readily be applied directly to blood culture broth and
thus do represent a practical rapid means for identifying causes of bacteremia. Nucleic
acid-based methods are, however, typically costly ($70 to $250 per application). In
determining whether such methods are cost justifiable in the workup of positive blood
cultures, the accrued laboratory costs must be balanced against the potential cost
savings in patient care.
Future technologies that measure host response to infection could be useful in
determining whether an organism recovered in a blood culture is of clinical signifi-
cance. This would be of particular value when combined with ultrasensitive detection/
identification molecular methods.
Blood culture contamination directly affects analytical testing and laboratory effi-
ciency. Workup of contaminated blood cultures increases technologists’ workloads at a
time when many microbiology laboratories are experiencing staffing shortages. In
addition, contaminated cultures divert technologist efforts away from other critical

Downloaded from https://journals.asm.org/journal/cmr on 24 July 2024 by 2409:40f4:1:3c3a:456a:7f5a:e81e:35f9.

samples. There is also the issue of increased time spent in phoning false-positive blood
cultures as critical action values to care providers. This is disruptive not only to the
laboratory but also to recipients of phone calls, e.g., nurses, physicians, and other health
care providers. Finally, high blood culture contamination rates may result in a negative
impression of clinical laboratory services, as users of the laboratory often believe that
contamination is the result of poor laboratory practices (6–8, 19, 22, 25).
Contaminated blood cultures also result in financial consequences to the laboratory,
as they lead directly to unnecessary and costly additional laboratory testing. Examples
include repeat blood cultures, cultures of ancillary sites, and nonmicrobiologic studies
such as therapeutic drug monitoring for agents such as vancomycin, basic metabolic
panels, and complete blood count (CBCs) (22). Microbiology laboratories may use more
media, perform additional organism identification procedures, and conduct unneces-
sary antimicrobial susceptibility tests. Gander et al. reported a difference in median
patient charges between negative and false-positive episodes of $8,720 per contami-
nant, a 47% increase (7). In this study, laboratory costs were not separated from the
other expenses. Bates et al. were among the few investigators to specifically address
additional microbiology costs between contaminated and negative cultures (6). Using
charge analysis as their metric, they noted an 80% increase in total microbiology
charges related to contaminated blood cultures, including a 30% increase in routine
culture charges and a 40% increase in blood culture charges (6). In a small study of
CoNS contaminants, Surdulescu et al. found a 150% increase in charges compared to
negative blood cultures (25). Finally, Alahmadi and colleagues found a 4-fold increase
in microbiology charges per contaminated culture compared to a negative culture (19).
Clearly, contaminated blood cultures have a substantial negative financial impact on
laboratory practice.

THE CLINICAL CONSEQUENCES OF BLOOD CULTURE CONTAMINATION

There are several untoward clinical consequences of contaminated blood cultures,
the most obvious of which is increased antibiotic exposure. Bates et al. found that

January 2020 Volume 33 Issue 1 e00009-19 cmr.asm.org 5

Doern et al. Clinical Microbiology Reviews

intravenous antibiotic charges were 39% higher for contaminant blood culture epi-
sodes than among culture-negative patients (6). Souvenir et al. demonstrated that 41%
of blood culture contaminant episodes due to CoNS were treated with antibiotics (with
34% receiving vancomycin unnecessarily) (22). Similarly, Lee et al. showed that 41% of
178 patients with contaminants received unneeded intravenous antibiotics (58). Many
patients who are started on antibiotics for contamination events receive prolonged
therapy. van der Heijden and colleagues found that the median antibiotic course
among patients receiving antibiotics after contamination events was 7 days (59), while
Souvenir et al. found a mean duration of 6.5 days of vancomycin for CoNS contamina-
tion episodes (22).
This increased antibiotic exposure is associated with several potential adverse
events, including allergic reactions, drug-drug interactions, antibiotic resistance emer-
gence, and disruption of the host microbiome that can lead to Clostridioides difficile
infection as well as other adverse consequences. Unfortunately, limited data exist to
quantify the burden of the adverse events that are specifically associated with con-
taminated blood cultures. While one study revealed that patients with contaminated
blood cultures who received antibiotics had higher 1- and 2-week crude mortality rates
than those who did not, this finding was confounded by the fact that patients given
antibiotics were sicker and had more comorbidities (e.g., malignancy), and their deaths
were not felt to be directly related to the consequences of unnecessary antibiotic
therapy (58).
Several other clinical consequences may follow blood culture contamination events.
Venous access must be maintained in order to deliver parenteral antibiotics and can
result in mechanical complications, thromboembolic disease, and infection (60). Un-
needed consultation requests may be generated (e.g., infectious diseases consulta-
tions), along with additional laboratory or other diagnostic testing (e.g., repeat blood
cultures and echocardiography). The search for a source of the bacteremia can lead to

Downloaded from https://journals.asm.org/journal/cmr on 24 July 2024 by 2409:40f4:1:3c3a:456a:7f5a:e81e:35f9.

unwarranted concern over, or even removal of, indwelling devices such as pacemakers
or implantable cardioverter defibrillators. Finally, the initial focus on the blood culture
result as the etiology of the patient’s presenting clinical syndrome may result in
“anchoring bias” (a form of cognitive bias in which one leans too heavily on an initial
piece of information when making subsequent decisions). This can lead to a delay in
obtaining the correct diagnosis and a delay in initiating appropriate therapy.
To quantify the frequency of some of these consequences, investigators at Vander-
bilt University Medical Center selected a random sample of 100 consecutive patients
with blood cultures positive for CoNS that were judged to be contaminants. Extrapo-
lating from this cohort, they estimated an annual burden of almost 900 additional
blood cultures obtained, over 350 additional antibiotic courses, and over 30 each of
catheter removals, echocardiograms, and subspecialty consultations at their hospital
(59). Moreover, an estimated 15 operative procedures would be postponed annually
due to contamination events. All of the above may then lead to increased hospital
lengths of stay.
Observational studies have described the magnitude of such increases in length of
hospitalization. For example, blood culture contamination resulted in an increase of
approximately 1 day of hospitalization (from 4 to 5 days) for patients with blood
cultures obtained in the emergency department (ED) prior to admission (7). Among
hospitalized patients, contaminated blood cultures were associated with a 5.4-day
increase in hospital stay compared with that of hospitalized controls matched for age,
comorbidity score, and admission month (19). Each additional hospital day due to
blood culture contamination increases the chances for a hospital-acquired adverse
event, including health care-associated infection, medication errors, falls, decubitus
ulcers, and thromboembolic events. While the risks for these adverse events vary
substantially by patient-level and facility-level variables, Hauck and Zhao estimated that
each additional night in a hospital increases a patient’s risk for an adverse drug reaction
by 0.5%, for hospital-acquired infections by 1.6%, and for pressure ulcers by 0.5% (61).
The negative clinical impact of blood culture contamination can be ameliorated in

January 2020 Volume 33 Issue 1 e00009-19 cmr.asm.org 6

Blood Culture Contamination Clinical Microbiology Reviews

part by the use of rapid diagnostics combined with active antibiotic stewardship efforts.
In a prospective, randomized controlled trial, Banerjee and colleagues at Mayo Clinic
introduced a rapid multiplex PCR for detection of pathogens directly from positive
blood cultures (62). They compared treatment with ⬎24 h of antibiotic therapy of CoNS
contamination episodes (1 positive culture out of ⱖ2 obtained) between the interven-
tion (rapid PCR, with and without active stewardship) and control (standard care) arms.
When combined with real-time audit and feedback from their antibiotic stewardship
team, the rapid testing group had a significant reduction in treatment of presumed
contaminants (from 25% in the control group to 8% [P ⫽ 0.015]). Rapid test results
alone (with templated comments in the report but no active stewardship) also dem-
onstrated a significant, though less pronounced, reduction in treatment of contami-
nants (11%). Nagel and colleagues likewise examined the impact of MALDI-TOF MS
identification and an antibiotic stewardship intervention on the duration of antibiotic
therapy for CoNS contamination episodes (63). Using a quasiexperimental, before-after
study design, they demonstrated a reduction in days of unnecessary antibiotic therapy
(from 3.89 to 1.31 days [P ⬍ 0.001]) in the postintervention cohort, as well as a 50%
reduction in the number of vancomycin trough levels performed.
Therefore, the implementation of rapid diagnostic testing for positive blood cultures
is likely to reduce, but not eliminate, the adverse clinical consequences of blood culture
contamination events. These approaches are most effective when combined with an
active antibiotic stewardship team. Such interventions are also justified based on many
published studies that report an overall mortality reduction, summarized in a recent
meta-analysis (64). However, these interventions (advanced rapid molecular diagnos-
tics, stewardship-trained pharmacists, and infectious diseases specialists) may not be
accessible to many hospitals. For example, the 2017 National Healthcare Safety Net-
work (NHSN) patient safety annual survey revealed that only 30% of 4,792 acute-care
hospitals used MALDI-TOF MS technology for bacterial organism identification, with

Downloaded from https://journals.asm.org/journal/cmr on 24 July 2024 by 2409:40f4:1:3c3a:456a:7f5a:e81e:35f9.

those facilities using it more likely to be larger and affiliated with a medical school (65).
Such hospitals may also be more likely to have dedicated phlebotomy services, also
known to reduce blood culture contamination rates.

THE ECONOMIC CONSEQUENCES OF BLOOD CULTURE CONTAMINATION

Numerous investigations, using various study designs, have examined the economic
impact of contaminated blood cultures. Studies dating back several decades have
consistently identified an economic cost associated with blood culture contamination.
Two case series have described the added costs associated with blood culture
contamination in pediatric patient populations (66, 67). In a retrospective, hospital-wide
study conducted between 1993 and 1996 in Boston, a total of 9,465 blood culture
specimens were obtained from febrile children; 87 blood cultures (0.9%) from 87
patients were considered to be contaminated (66). Among these 87 patients, 15 were
treated in the hospital and 72 were managed on an outpatient basis. Hospital costs
were not reported in this study, but additional charges attributable to the contami-
nated blood cultures were estimated to be $10,821 for patients treated in the outpa-
tient setting and $16,200 for patients admitted to the hospital. A second retrospective
case series study evaluated 8,306 children with blood cultures drawn in an ED in
Washington, DC, between 1994 and 1996 (67). Of these cultures, 85 (1%) yielded
contaminants. The average added cost for the care of patients with contaminated
blood cultures in this study was $928 per patient. This cost was incurred as a result of
hospital admissions (61% of total costs), reevaluation in the ED (31% of total costs), and
antimicrobial susceptibility testing (8% of total costs).
Similar observations have been made for adult patients. A retrospective case series
conducted in 1997 assessed the cost implications of contaminated blood cultures in
adults (22). In this study, 3,276 blood cultures obtained from 1,433 adult patients
hospitalized in Spokane, WA, were examined. The primary intent of this investigation
was to assess the attributable hospital costs of blood culture contamination on
antimicrobial therapy. Fifty-nine patients in this study were noted to have contami-

January 2020 Volume 33 Issue 1 e00009-19 cmr.asm.org 7

Doern et al. Clinical Microbiology Reviews

nated blood cultures; 52 (88.1%) of these cultures yielded CoNS. No species identifi-
cation was performed. Among these 59 patients, 24 (41%) were treated with a systemic
antimicrobial, most commonly vancomycin (n ⫽ 20 [34%]). No description of empirical
versus directed therapy was reported, but the average length of vancomycin therapy
was 6.5 days, suggesting that antimicrobial therapy was continued after identification
of the organism. The per-patient average cost to perform pharmacokinetic monitoring
and purchase of vancomycin in this study was $645.
Although retrospective case series such as those discussed above do not prove
causality, they do provide insight into the potential for added costs associated with
contaminated blood cultures. There have, however, been six published studies that
have compared the economic differences between patients with contaminated blood
cultures directly to those with negative blood cultures (6–8, 19, 68, 69). These com-
parative investigations, because of their design, provide more direct evidence of the
cost implications of blood culture contamination.
A retrospective case-control study conducted in Ireland (2007 to 2008) compared
142 hospitalized patients with contaminated blood cultures matched by age, comor-
bidities, and length of time in the hospital prior to collection of blood cultures to an
equal number of hospitalized patients with negative blood cultures (19). The total
length of hospital stay was 5.4 days longer for patients with contaminated blood
cultures (95% confidence interval [CI], 2.8 to 8.1) and was associated with increased
hospital costs of $7,502 (U.S. dollars) (95% CI, $4,925 to $10,078) per patient. Per-patient
pharmacy costs increased by $95 and laboratory costs by $61. Another comparative
study conducted in St. Louis, MO (1995 to 1996), examined cost differences between
patients with contaminated versus negative blood cultures (68). This prospective cohort
study of adult, hospitalized patients was conducted as part of a randomized controlled
trial evaluating different disinfectant methods. All patients with contaminated blood
cultures (n ⫽ 120) were compared to all other patients that participated in the study
(n ⫽ 3,731). The average length of stay was 2 days longer for patients with contami-

Downloaded from https://journals.asm.org/journal/cmr on 24 July 2024 by 2409:40f4:1:3c3a:456a:7f5a:e81e:35f9.

nated blood cultures (11 ⫾ 12 days) than for patients with negative blood cultures
(9 ⫾ 10 days), but these results were not statistically significant (P ⫽ 0.11). However,
statistically significant increases in cost were noted in the total cost of hospitalization
($4,100 increase [95% CI, $740 to $7,400]), the total cost of antimicrobials ($700 increase
[95% CI, $20 to $1,400]), and the total laboratory costs ($330 increase [95% CI, $140 to
$300]).
Three other cohort studies assessed hospital charges as opposed to hospital costs,
two involving hospitalized patients (6, 8) and the other involving patients in the ED (7).
The first of these was a prospective study in Boston, MA, in 1989 in which 94 of 1,191
hospitalized patients (7.9%) who had blood cultured were found to have contaminated
blood cultures (6). The length of stay was significantly longer (P ⫽ 0.003) for patients
with contaminated blood cultures (median, 12.5; interquartile range [IQR], 7 to 19 days)
compared to patients with negative blood cultures (median, 8; IQR, 5 to 17 days).
Statistically significantly increased hospital charges were also noted, including total
hospital charges ($4,385 increase [P ⫽ 0.004]), antimicrobial charges ($382 increase
[P ⫽ 0.004]), and laboratory charges ($631 increase [P ⫽ 0.0003]). In multivariate anal-
ysis, total charges increased by 12% for patients with contaminated blood cultures
compared to patients with negative blood cultures. In a retrospective cohort study of
hospitalized patients conducted in 2002 in Denver, CO, Zwang and Albert compared 56
patients with contaminated blood cultures to 816 patients with negative blood cultures
(8). An increase in total hospital stay of 3 additional days was noted for patients with
contaminated blood cultures. This was associated with increased hospital charges of
$8,756. Based on their overall 6% rate of blood culture contamination, they estimated
that contamination of blood cultures would result in $1.8 million in excess annualized
costs, most of that accounted for by an estimated 1,455 to 2,200 extra days of
hospitalization per year. Gander et al. also examined the impact of blood culture
contamination on total hospital charges in Dallas, TX, in 2006 to 2007 (7). This
investigation was part of a randomized controlled study aimed at evaluating phlebot-

January 2020 Volume 33 Issue 1 e00009-19 cmr.asm.org 8

Blood Culture Contamination Clinical Microbiology Reviews

omist versus nonphlebotomist blood culture teams in the ED. All patients with con-
taminated blood cultures during the study period (n ⫽ 120) were compared to 960
randomly selected patients with negative blood cultures. Hospital charges were sig-
nificantly higher for patients with contaminated blood cultures (median, $27,472; IQR,
$21,063 to $37,841) than for patients with negative blood cultures (median: $18,752;
IQR: $17,046 to $20,315).
And finally, in a recent retrospective comparative observational study predicated on
matched survival data derived from two large teaching hospitals in Boston, Geisler et
al. estimated a cost per blood culture contamination event of $6,463; 79% of the
increased cost of care was the result of prolonged hospital stays and increased duration
of antimicrobial therapy (69).
All six of the cohort studies presented above demonstrate consistently increased
hospital costs or charges associated with blood culture contamination. Parenthetically,
it is also interesting that over time the increased length of stay associated with
contaminated blood cultures seems to have decreased. Possible explanations for this
observation include the positive impact of rapid diagnostic procedures in the clinical
microbiology laboratory, the increasing reliance on antimicrobial stewardship programs
as a means for optimizing antimicrobial therapy, and/or the heightened general
emphasis on shortening hospital stays.
Lastly, two investigations have assessed the economic benefits of routine use of
interventions to decrease costs associated with blood culture contamination (70, 71).
Self et al. compared routine use of sterile collection kits as well as phlebotomy teams
for drawing blood cultures as means for reducing blood culture contamination rates
(70). Using a decision analysis model, routine use of a sterile kit resulted in net
annualized savings of $483,219 in comparison to the routine blood culture collection
procedure that had been in place. An annual savings of $288,980, was observed when
trained phlebotomists collected blood cultures. Skoglund et al. performed a similar

Downloaded from https://journals.asm.org/journal/cmr on 24 July 2024 by 2409:40f4:1:3c3a:456a:7f5a:e81e:35f9.

decision analysis investigation in assessing the routine use of an initial specimen
diversion device to reduce blood culture contamination. The Steripath diversion device
(Magnolia Medical Technologies, Seattle, WA; discussed below) was compared to
standard of care in patients with blood cultures drawn in the ED (71). In this study, it
was estimated that the routine use of the diversion device to prevent contamination
would result in overall hospital cost savings of $272 per blood culture obtained in the
ED. Both of these cost-benefit analyses provide convincing evidence as to the cost
effectiveness of interventions that reduce blood culture contamination.

PREVENTION OF BLOOD CULTURE CONTAMINATION

The clinical and cost implications of blood culture contamination together with the
impact of contamination on clinical microbiology laboratory function are summarized
in Fig. 1. The question then arises, what can be done to reduce the scope and
magnitude of this problem?

Patient Selection
Prevention of blood culture contamination starts with selecting appropriate patients
for blood cultures. The pretest probability for bacteremia has a major impact on the
positive predictive value of the blood culture result. Performing blood cultures on
patients with a very low likelihood of bacteremia results in positive cultures more
frequently representing false positives. Therefore, clinicians should be educated regard-
ing clinical conditions more frequently associated with bacteremia and appropriate
laboratory stewardship. For example, there is growing recognition of the waste asso-
ciated with unrestrained blood culture use and the lack of utility of routinely repeating
blood cultures for patients with Gram-negative bacteremia (72, 73). Bacteremia predic-
tion models may eventually prove useful in guiding clinical decision making (74, 75).
There is potential for decision support tools, associated with the electronic medical
record, to positively affect the appropriateness of provider orders for blood cultures.

January 2020 Volume 33 Issue 1 e00009-19 cmr.asm.org 9

Doern et al. Clinical Microbiology Reviews

FIG 1 Consequences of blood culture contamination.

Skin Antisepsis
Extensive investigation has been conducted to discern the preferable disinfectant to
be used for skin disinfection in blood culture procurement. Immediate antimicrobial
activity at the venipuncture site is required, but prolonged residual effect is not needed.
The results of a College of American Pathologists Q-Probe survey published in 1998
noted that the median blood culture contamination rate among adult inpatients was
significantly lower in hospitals using tincture of iodine (2.1%) than in those institutions
using an povidone iodine (2.9%) for blood culture skin disinfection (P ⫽ 0.036) (14). A
year later, similar observations regarding the superiority of tincture of iodine versus
povidone iodine as a skin disinfectant were made by Little et al. (68). Further reduction
in reported contamination rates was observed when 0.5% chlorhexidine gluconate,

Downloaded from https://journals.asm.org/journal/cmr on 24 July 2024 by 2409:40f4:1:3c3a:456a:7f5a:e81e:35f9.

used with alcohol, became available (76).
Subsequent reports, however, did not demonstrate a statistically significant differ-
ence with alcoholic chlorhexidine in comparison to aqueous povidone-iodine or tinc-
ture of iodine (4, 77–79). Maiwald and Chan performed a systematic review and
meta-analysis on the efficacy of chlorhexidine in skin antisepsis in which 12 articles,
including 2 previous systematic reviews, were included (80). Alcoholic chlorhexidine
was noted to be superior to aqueous povidone iodine. However, there was no evidence
to support alcoholic chlorhexidine over alcoholic povidone iodine or other comparative
agents that contained alcohol. In fact, alcohol alone may be an adequate skin disin-
fectant (79, 81). Investigators at M. D. Anderson Cancer Center noted that the addition
of 5% dimethyl sulfoxide (DMSO) to 70% isopropyl alcohol and 1% iodine resulted in
a significant decrease in blood culture contamination (82). Guidelines generally recom-
mend disinfection of the phlebotomy site with 2% alcoholic chlorhexidine or 70%
isopropyl alcohol followed by 2% chlorhexidine, with at least 30 s allowed for drying
(83, 84). We recommend careful disinfection of the skin prior to phlebotomy for blood
cultures with an alcohol-containing disinfectant and allowance for drying. Data do not
support preference for alcoholic chlorhexidine over an alcoholic iodine-containing
preparation, alcohol followed by another disinfectant, or even alcohol alone.

Blood Culture Bottle Disinfection

Although they are covered with a lid, the rubber septa of blood culture vials are not
sterile, and it is standard practice to disinfect the tops of culture bottles prior to
inoculation (14, 15, 44, 83–85). Seventy percent isopropyl alcohol is most often used for
this purpose. Iodine products should not be used alone to disinfect blood culture bottle
tops, as the iodine may result in erosion of the rubber and introduction of contaminants
(44).

Blood Culture Collection Site

It is preferable to obtain blood for culture via venipuncture rather than from

January 2020 Volume 33 Issue 1 e00009-19 cmr.asm.org 10

Blood Culture Contamination Clinical Microbiology Reviews

intravascular catheters (21, 86). A meta-analysis of nine studies that met carefully
chosen quality metrics (24) demonstrated that blood collected through an intravascular
catheter had, on average, a 2.69-fold greater likelihood of being contaminated than
blood collected by venipuncture (95% CI, 2.03 to 3.57) (21). When blood cultures are
obtained through lines, organisms colonizing the catheter hubs may result in false-
positive blood cultures with not only skin commensals but also recognized pathogens
such as enterococci, S. aureus, and even Gram-negative bacilli (29, 30). This may falsely
elevate an institution’s rate of central line-associated bloodstream infections (CLABSIs)
as established by the NHSN (87). False-positive CLABSI rates may have negative patient
and financial consequences if thresholds are exceeded and could affect an institution’s
reputation as a quality organization (29, 30).
While blood cultures obtained via intravascular catheters are associated with de-
creased specificity and lower positive predictive value than for peripheral venipuncture
blood cultures, they do have greater sensitivity and negative predictive value (86).
Deriving data from 6 studies on this topic, Falagas et al. concluded that for every 1,000
patients with blood cultures obtained via intravascular catheters, an additional 8
patients with true bacteremia would be identified, but 59 false-positives would also
result (86). Blood culture contamination rates may be decreased with newly inserted
central venous catheters if the sample for culture is obtained via a lumen not exposed
to the guidewire (88).
Furthermore, if a central venous catheter is thought to be a possible source of a
patient’s bacteremia, paired blood samples may be obtained for culture from both the
catheter and a peripheral site, and the length of time to positivity compared. In order
for the time-to-positivity comparison to be valid, it is important that equal volumes of
blood be drawn from both sites. Alternatively, both blood samples can be cultured
quantitatively and the quantities of organisms recovered compared (89). Blood cultures
obtained from a central venous catheter should be drawn via a freshly placed needle-

Downloaded from https://journals.asm.org/journal/cmr on 24 July 2024 by 2409:40f4:1:3c3a:456a:7f5a:e81e:35f9.

less connector, as connectors in clinical use are frequently colonized and result in blood
culture contamination (90). In addition, if a catheter-related bloodstream infection is
suspected, some data support obtaining blood for culture from each catheter lumen
(91). However, in patients with multiple catheters and multiple lumens, this practice can
result in a significant amount of blood sampled at a substantial cost as well as
additional opportunities for blood culture contamination with each lumen sampled.
Although the routine practice of drawing blood cultures via vascular catheters is
discouraged, it remains a common practice in EDs when peripheral intravenous cath-
eters are established and experience, particularly in pediatric patients, remains some-
what conflicting (37, 92–95). Interestingly, the use of antiseptic barrier caps or passive
port protectors has been noted to decrease contamination of blood cultures obtained
via central venous catheters (96).

Single Needle versus Double Needle

As discussed above, because the needle used for venipuncture can become con-
taminated, it used to be common practice to apply a new sterile needle to transfer
blood from a syringe to the blood culture vial (double needle transfer technique), and
indeed, as noted above, a meta-analysis of 8 studies demonstrated a decrease in
contamination rate from 3.7% to 2.0% associated with the double-needle technique
(40). However, due to concern regarding needlestick injury, the double-needle transfer
technique has fallen into disfavor, and instead, it is recommended that equipment be
utilized that allows for the direct collection of blood into culture vials (15, 83).

Sterile Gloves and Hand Hygiene

Gloves should be worn as personal protective equipment when obtaining blood
cultures as part of standard infection prevention/blood-borne pathogen precautions.
However, it should be noted that nonsterile gloves can become contaminated and have
resulted in blood culture contamination (97). The use of sterile gloves was associated
with a significant decrease in blood culture contamination in a single-center crossover

January 2020 Volume 33 Issue 1 e00009-19 cmr.asm.org 11

Doern et al. Clinical Microbiology Reviews

trial (98). However, the baseline rate of contamination at the center was low (1.1%) and
the overall decrease (0.5%) was small.Sterile glove use for blood culture procurement
has been an element of successful multicomponent culture contamination reduction
programs (99, 100). Clearly, sterile gloves should be employed if it is necessary to
repalpate a venipuncture site after it has been disinfected (101, 102). Although the
performance of hand hygiene is universally recommended prior to patient contact as
part of standard infection prevention, and intuitively one would reason that hand
hygiene could decrease blood culture contamination, there are few data to suggest
that hand hygiene plays a role in blood culture contamination rates.

Blood Culture Kits and Standard Procedures

In some studies, the use of blood culture collection kits (with or without sterile
gloves) and standardized operating procedures has been associated with a significant
decrease in blood culture contamination (100, 103–106). However, in a meta-analysis of
7 studies, Snyder and colleagues noted that the utilization of prepackaged blood
culture preparation kits did not result in a significant reduction in the rate of blood
culture contamination (odds ratio [OR], 1.121; 95% CI, 0.94 to 1.35) (21). Also, it should
be noted that in many studies the kits were one of several interventions, such as
education, training, and data feedback, and thus, the relative contribution of the kit to
the blood culture contamination rate is difficult to discern.

Sterile Drapes
The effect of sterile drapes on blood culture contamination has not been studied in
isolation, but sterile drapes have been included in some blood culture kits as part of
multicomponent improvement projects (100).

Blood Sampling and Volume

Sampling an appropriate amount of blood is essential in optimizing the perfor-

Downloaded from https://journals.asm.org/journal/cmr on 24 July 2024 by 2409:40f4:1:3c3a:456a:7f5a:e81e:35f9.

mance characteristics of blood cultures. The detrimental effect on pathogen detection
by culturing an inadequate amount of blood is well known (106). In this regard,
whenever possible, two or three 20-ml volumes of blood should be obtained in the
initial evaluation of adult patients suspected of having bacteremia. In pediatric patients
in whom collection of high volumes of blood is not possible, two blood specimens
should be obtained each consisting of as much blood as can safely be collected.
Importantly, separate sticks should be employed in collecting each blood sample. The
practice of obtaining multiple blood samples during a single phlebotomy as can
happen, for example, at the time an intravenous (i.v.) catheter is placed, is discouraged.
It is also essential that care be taken to inoculate appropriate volumes of blood into
blood culture vials. Both underfilling and overfilling blood culture vials have been
associated with contamination and/or false-positive results (107–109). As noted above,
the clinical microbiology laboratory should be proactive in establishing policies and
providing guidance to care providers with respect to optimizing the collection of blood
for culture.

Phlebotomy Teams/Education
Aseptic collection of blood for culture is a complex skill that requires special
knowledge and training. Unfortunately, gaps in knowledge and performance flaws are
common (92, 110, 111). Therefore, it is not surprising that the use of trained phlebot-
omists to obtain blood cultures has been associated with decreased contamination of
blood cultures (21, 112–114). In addition, a meta-analysis of five large studies con-
ducted in several U.S. hospitals, four of which were designated good quality, showed
excellent strength of evidence supporting reduced contamination rates by trained
phlebotomists compared to nonphlebotomists (25). The mean odds ratio of all five
studies of 2.58 (95% CI, 2.07 to 3.20) favors phlebotomy teams for decreasing blood
culture contamination (21). Several authors have detailed educational programs asso-
ciated with successful reduction in blood culture contamination (21, 112, 115, 116).

January 2020 Volume 33 Issue 1 e00009-19 cmr.asm.org 12

Blood Culture Contamination Clinical Microbiology Reviews

Multidisciplinary/Multimodal Performance Improvement

In many instances, interventions to minimize blood culture contamination have not
been studied individually, and instead, multiple measures are introduced in multimodal
performance improvement projects that often include education and training, kits,
sterile gloves, phlebotomy teams, etc. (30, 102, 112, 117, 118). Translating evidence into
sustained clinical practice via multidisciplinary improvement projects requires attention
to both technical and adaptive work (119). Adaptive challenges to sustained perfor-
mance improvement may include people’s beliefs, habits, and values.

Surveillance and Feedback

Surveillance for blood culture contamination, most commonly using laboratory
parameters for the definition of contamination, is crucial in establishing baseline
institutional performance metrics and measuring the effect of improvement efforts.
Surveillance and feedback systems have been shown in multiple studies to result in
improved blood culture contamination rates, particularly when contamination rates are
reported in a timely manner and directed individually to those who perform phlebot-
omy (117, 120–123). In one study, education combined with feedback to individual
phlebotomists proved more effective than education alone (124).

Initial Specimen Diversion

Hair follicles, sebaceous glands, and deeper layers of the epidermidis serve as
sanctuary sites for skin flora. It is very difficult to eradicate microbes from these
protected sites with topical antiseptics. Skin fragments that may be colonized with
microbes can be dislodged as a result of venipuncture, potentially resulting in blood
culture contamination (125, 126). The previously discussed methods to prevent blood
culture contamination do not fully address this mode of blood culture contamination.
As an alternative or at least as an adjunct to these methods, diversion of the first
portion of blood, which presumably contains the contaminating bacteria, with culture

Downloaded from https://journals.asm.org/journal/cmr on 24 July 2024 by 2409:40f4:1:3c3a:456a:7f5a:e81e:35f9.

of the remaining secondary aliquot of blood, has been proposed as a means to
decrease contamination. Studies assessing the utility of this approach have reported
overall blood culture contamination rates of 1.6 to 2.4% (127–129).
Recently, two commercially available, self-contained devices that do not require
extra collection tubes but which achieve initial specimen diversion have been intro-
duced to the market, the Steripath Gen2 system (Magnolia Medical Technologies,
Seattle, WA) and the Kurin Lock device (Kurin, Inc., San Diego, CA). The question arises,
can even lower blood culture contamination rates be achieved through use of these
diversion devices?
The performance characteristics of the Steripath Gen2 device have been elucidated
in several published reports, three of which are outlined here (9, 130, 131). In a
prospective, controlled, matched-pair trial in an ED setting, an environment in which
blood culture contamination is notoriously a major problem, Rupp and colleagues
observed a nearly 90% decrease in blood culture contamination rates based on use of
the Steripath device (9). Overall blood culture contamination rates decreased from a
level of 1.78% to a level of 0.22% (P ⫽ 0.001) through use of this device during the
12-month intervention period involving 904 subjects and 1,808 blood cultures. Limi-
tations of the study include a potential selection bias, lack of pediatric aged patients,
and the fact that the study was conducted at a single center. In a multicenter controlled
clinical study in four EDs, the historical blood culture contamination rate over a
year-long period (35,392 blood cultures) was compared to contamination rates during
a 7-month intervention period (6,293 blood cultures) (130). The cumulative average
blood culture contamination rate decreased from a level of 3.52% to a level of 0.6%
(P ⫽ 0.0001) when the Steripath device was used to process blood specimens obtained
by both venipuncture and at the time that peripheral i.v. lines were placed, an 83%
reduction in contamination rate. And finally, in a prospective study conducted on a
single medical ward over a 6-month period involving 671 blood cultures obtained via
peripheral phlebotomy or at the time of initial peripheral intravenous catheter insertion, use

January 2020 Volume 33 Issue 1 e00009-19 cmr.asm.org 13

Doern et al. Clinical Microbiology Reviews

TABLE 2 Interventions to prevent blood culture contaminationa

Intervention Comment
Patient selection Blood cultures should be performed for patients with a reasonable likelihood of bacteremia.
Skin disinfection Use of an alcohol containing disinfectant is recommended.
Blood culture bottle cap disinfection Tops of blood culture vials should be disinfected prior to inoculation of blood.
Phlebotomy site (intravascular catheter Blood cultures should not be obtained via intravascular catheters unless the catheter is thought to
vs peripheral vein) be the source of bacteremia.
Single-needle vs double-needle transfer Although double-needle technique may be helpful, it is not recommended due to risk of
needlestick injury. Use direct transfer technique.
Sterile gloves/hand hygiene Limited data to support sterile glove use
Standardized kits Standardized kits and procedures helpful in preventing blood culture contamination
Sterile drapes Not studied as an isolated intervention; sometimes included in blood culture kits
Appropriate blood volume Contamination and false-positive results associated with under- and overfilling blood culture
bottles
Phlebotomy team/education Proven useful in decreasing blood culture contamination in numerous studies
Multidisciplinary quality improvement Requires both technical and adaptive work
Surveillance and feedback A key part of any comprehensive program to decrease blood culture contamination
Initial specimen diversion Commercially available device shows promise as a cost-effective means to decrease contamination
aDerived from references 4, 14, 21, 24, 30, 31, 37, 68, 71, 76–128, 130, 131, 133.

of the Steripath device (207 cultures) was associated with a 1.0% contamination rate,
compared to a contamination rate of 5.2% in 464 cultures obtained via standard
practice (P ⬍ 0.008) (131). There is one published study pertaining to the Kurin Lock
device (132). In a small, single-center study, blood culture contamination rates obtained
during a 3-month period in which the Kurin device was used to process blood cultures
were reported as 0.44%. Contamination rates of 1.71% were noted during a 3-month
comparison period. No information was provided regarding the numbers of blood
cultures performed, the characteristics of the patient populations assessed, the criteria
used to judge the clinical significance of blood culture isolates, or the identification of

Downloaded from https://journals.asm.org/journal/cmr on 24 July 2024 by 2409:40f4:1:3c3a:456a:7f5a:e81e:35f9.

the microorganisms recovered from blood cultures. There are no head-to-head com-
parative trials with these two commercially available specimen diversion devices.
Table 2 summarizes those factors that are important in avoiding contamination of
blood cultures.

Unanswered Questions
A number of unanswered questions remain related to the practice of obtaining
blood cultures and their interpretation.
The first of these is, should there be a new universal standard which defines
“acceptable” overall institutional blood culture contamination rates, and, if so, what
should that threshold value be? Currently, rates of ⱕ3.0% are considered acceptable. It
is our opinion, however, that overall institutional contamination rates of ⱕ1.0% are now
achievable, and therefore, consideration should be given to the establishment of a new
universal threshold value of ⱕ1.0%.
Second, in settings in which overall contamination rates rise above 1%, should
objective, stepwise quality improvement programs designed to improve patient care
and reduce unnecessary costs be implemented? The answer is clearly yes. In this regard,
organizations concerned with patient safety and health care quality control such as The
Joint Commission, the Centers for Disease Control and Prevention, and the Agency for
Healthcare Research and Quality should assume a leadership role.
A third question is whether it is advisable for institutions to have a single, uniform
standard operational policy for collecting blood cultures, one that reduces to the extent
possible the problem of contamination. Again, the answer to this question is clearly yes,
at least as it pertains to the routine evaluation of patients suspected of having
bacteremia. However, in selected patients, e.g., patients with endocarditis or infected
endovascular devices such as pacemakers or vascular grafts, patients with joint proth-
eses, and profoundly immunocompromised patients, a routine policy might be insuf-
ficient. This is because positive blood cultures yielding organisms usually considered to
be contaminants can be and often are of clinical significance for such patients. Further,

January 2020 Volume 33 Issue 1 e00009-19 cmr.asm.org 14

Blood Culture Contamination Clinical Microbiology Reviews

the clinical consequences of a contaminated blood culture vary depending on the

location from which a specimen is drawn (e.g., in an ED or on a ward caring for
immunosuppressed patients). In addition, contaminated blood cultures are more likely
to occur in patients with limited or difficult venous access such as neonates, patients
with concurrent vascular lines, obese patients, or in patients in which blood cultures are
drawn emergently. Thus, in these settings, even greater effort is warranted in attempt-
ing to avoid contamination of blood cultures. This may include collection of blood for
culture by use of initial specimen diversion devices and/or use of specialized, highly
trained personnel to draw blood cultures using sterile gloves or a double-needle
method. A related question is whether special laboratory techniques to rapidly identify
blood culture isolates and antimicrobial resistance determinants should be utilized in
these situations. The answer to this secondary question is yes.
A fourth question is whether better criteria are needed to distinguish between
contaminants and true pathogens when unusual organisms recovered from blood
cultures are identified using non-culture-based molecular methods, e.g., MALDI-TOF
MS, nucleic acid amplification, PNA-FISH, and/or nucleic acid sequencing. Notwith-
standing situations where contamination is likely to have occurred, for example,
multiple blood culture sets obtained with only a single bottle positive, blood culture
isolates identified by these methods are often, a priori, considered to be of clinical
significance. This can result in two unintended consequences. First, patients receive
unnecessary antimicrobial therapy and additional diagnostic studies are performed.
Second, the spurious diagnosis of certain conditions may result in unfair financial
penalties being imposed on institutions based on the dictates of the Centers for
Medicare Services. Expanding the list of known pathogens or otherwise changing the
definition of what constitutes blood culture contamination may serve as solutions to
this problem.
A fifth issue pertains to the increasingly important problem of central line-associated

Downloaded from https://journals.asm.org/journal/cmr on 24 July 2024 by 2409:40f4:1:3c3a:456a:7f5a:e81e:35f9.

bloodstream infections (CLABSIs). CLABSIs are not only of clinical significance: CLABSI
rates are publicly reported by the Center for Medicare and Medicaid Services (CMS) and
can result in penalties via CMS pay-for-performance programs (value-based purchasing
and hospital-acquired conditions programs) (133). However, as noted above, blood for
culture when obtained through an indwelling vascular catheter is more likely to be
contaminated than blood obtained by peripheral venipuncture. This underscores the
importance of accurately assessing the significance of blood culture isolates in patients
with central lines. The NHSN has provided criteria for defining the presence of a true
CLABSI (134). Among these criteria, two or more blood cultures which yield organisms
defined by the CDC as “common commensals” are to be considered significant. NHSN
developed and maintains a list of organisms to be considered common commensals
(135). Currently, there are over 700 hundred organisms on this list, the vast majority of
which are unknown to all but a few clinical microbiology laboratories in the United
States. Because of the nature of this list, it requires frequent updates, the most
recent of which was in January of 2019 (136). Further, identification of most of the
organisms on the list necessitates use of expensive molecular methods, e.g.,
MALDI-TOF and/or nucleic acid sequencing techniques, that may not be available in
many clinical microbiology laboratories. As a consequence, there exists a tendency
to overestimate the clinical significance of blood culture isolates from central lines,
in turn leading to the potential for all of the negative downstream clinical conse-
quences of false-positive blood cultures discussed above. Blood culture contami-
nation events can also have a negative impact on hospital reimbursement if the
positive blood culture result contributes to meeting the NHSN definition of central
line-associated bloodstream infection (133). Clearly, better criteria are needed to
more accurately define the presence of a true CLASBI, preferably, criteria that are
predicated on technologies and approaches that are actually in place in clinical
microbiology laboratories today.
Additional questions pertain to use of initial specimen diversion devices as a means
for reducing blood culture contamination rates. To wit, should such devices be used

January 2020 Volume 33 Issue 1 e00009-19 cmr.asm.org 15

Doern et al. Clinical Microbiology Reviews

routinely, or should they be applied selectively in circumstances where there exists a

high likelihood of blood culture contamination? Also, can such devices be used to
reduce contamination associated with blood drawn directly through indwelling venous
catheters? The availability of more extensive data in peer-reviewed publications
which assess the utility of the Kurin Lock device would be most instructive, as would
published data describing the results of a head-to-head comparison of the Steripath
and Kurin Lock devices. Additional data comparing contamination rates obtained
with diversion devices versus simple initial specimen discard would be of value, as
would data which assess the utility of diversion device over prolonged study
periods, preferably with application to all patients rather than selected patient
populations.

CONCLUSIONS
Blood cultures have been a critically important and potentially life-saving diagnostic
test for over a century, yet the problem of blood culture contamination persists. There
exists a universal consensus that contaminated blood cultures are costly and lead to
important unintended consequences for patients. These unintended consequences are
diverse and complex, but most are directly or indirectly related to unnecessary and
prolonged antibiotic exposure, increased diagnostic testing, and prolonged periods of
hospitalization. In general, in the past, overall institutional blood culture contamination
rates of 3% have been considered acceptable.
However, we now have a comprehensive understanding of how blood culture
contamination occurs and equally importantly, a clear understanding of how to dimin-
ish the scope and magnitude of the problem. The frequency of contamination and
subsequent secondary harm and excess cost related to contamination can be reduced
by careful selection of patients who really need blood cultures, meticulous and proper
skin antisepsis, the use of dedicated phlebotomy teams to obtain blood samples,

Downloaded from https://journals.asm.org/journal/cmr on 24 July 2024 by 2409:40f4:1:3c3a:456a:7f5a:e81e:35f9.

and/or adequate training of staff on how to properly obtain blood samples. Whenever
possible, blood for culture should not be obtained from vascular catheters. Further, the
use of initial specimen diversion devices as an alternative to standard blood specimen
collection methods represents an exciting new advance that has the potential for
reducing overall contamination rates to levels not previously considered to be attain-
able. In view of the foregoing, we believe that a new universal standard of ⱕ1% should
be considered in defining allowable overall institutional blood culture contamination
rates.

ACKNOWLEDGMENTS
We appreciate the critical evaluation of the manuscript and helpful suggestions for
revision by Christopher D. Doern, Director, Clinical Microbiology Laboratories, Medical
College of Virginia, Richmond, VA.
We disclose the following associations: G.V.D., consultant (Magnolia Medical Tech-
nologies) and honoraria (UpToDate); K.C.C., research support (GenMark, Singulex and
Accelerate); D.J.D., research support (bioMérieux); K.W.G., research support (Magnolia
Medical Technologies, Merck & Co, Summit Therapeutics, and Tetraphase); M.E.R.,
research support (Magnolia Medical Technologies, ContraFect, and XBioTech), con-
sultant (3M, Citius, Telefex/Arrow, and Ariste), advisory panel (3M), and webinar
participant (Magnolia Medical Technologies); M.P.W., research support (Beckman-
Coulter, JMI Laboratories, and the NJ Department of Health), speaker’s honoraria
(Beckman-Coulter), and consultant (SeLux Diagnostics and T2 Biosystems); D.J.S.,
medical advisory board (Magnolia Medical Technologies), stock ownership (Mag-
nolia Medical Technologies), honoraria (UpToDate), research support (Centers for
Disease Control and Prevention, the Duke-UNC Prevention Epicenter, and the
National Football League), and consultant (Johnson & Johnson). Each of the authors
was the recipient of an unrestricted educational grant from Magnolia Medical
Technologies in support of this review.

January 2020 Volume 33 Issue 1 e00009-19 cmr.asm.org 16

Blood Culture Contamination Clinical Microbiology Reviews

REFERENCES
1. Richter SS, Beekman SE, Croco JL, Diekema DJ, Koontz FP, Pfaller MA, 20. Dargère S, Cormier H, Verdon R. 2018. Contaminants in blood cultures:
Doern GV. 2002. Minimizing the workup of blood culture contaminants: importance, implications, interpretation and prevention. Clin Microbiol
implementation and evaluation of a laboratory-based algorithm. J Clin Infect 24:964 –969. https://doi.org/10.1016/j.cmi.2018.03.030.
Microbiol 40:2437–2444. https://doi.org/10.1128/jcm.40.7.2437-2444.2002. 21. Snyder SR, Favoretto AM, Baetz RA, Derzon JH, Madison BM, Mass D,
2. Weinstein MP, Towns ML, Quartey SM, Mirrett S, Reimer LG, Parmigiani Shaw CS, Layfield CD, Christenson RH, Liebow EB. 2012. Effectiveness of
G, Reller LB. 1997. The clinical significance of positive blood cultures in practices to reduce blood culture contamination: a laboratory medicine
the 1990s: a prospective comprehensive evaluation of the microbiol- best practices systematic review and meta-analysis. Clin Biochem 45:
ogy, epidemiology, and outcome of bacteremia and fungemia in 999 –1011. https://doi.org/10.1016/j.clinbiochem.2012.06.007.
adults. Clin Infect Dis 24:584 – 602. https://doi.org/10.1093/clind/24.4 22. Souvenir D, Anderson DE, Palpant S, Mroch H, Askin S, Anderson J,
.584. Claridge J, Eiland J, Malone C, Garrison MW, Watson P, Campbell DM.
3. Pien BC, Sundaram P, Raoof N, Costa SF, Mirrett S, Woods CW, Reller LB, 1998. Blood cultures positive for coagulase-negative staphylococci:
Weinstein MP. 2010. Evaluation of the microbiology, epidemiology, and antisepsis, pseudobacteremia, and therapy of patients. J Clin Microbiol
outcome of bloodstream infections in hospitalized adults. Am J Med 36:1923–1926.
123:819 – 828. https://doi.org/10.1016/j.amjmed.2010.03.021. 23. Weinstein MP. 2003. Blood culture contamination: persisting problems
4. Story-Roller E, Weinstein MP. 2016. Chlorhexidine versus tincture of and partial progress. J Clin Microbiol 41:2275–2278. https://doi.org/10
iodine for reduction of blood culture contamination rates: a prospec- .1128/jcm.41.6.2275-2278.2003.
tive randomized crossover study. J Clin Microbiol 54:3007–3009. 24. Christenson RH, Snyder SR, Shaw CS, Derzon JH, Black RS, Mass D, Epner
https://doi.org/10.1128/JCM.01457-16. P, Favoretto AM, Liebow EB. 2011. Laboratory medicine best practices:
5. Gilligan P. 2013. Blood culture contamination: a clinical and financial systematic evidence review and evaluation methods for quality im-
burden. Infect Control Hosp Epidemiol 34:1–2. provement. Clin Chem 57:816 – 825. https://doi.org/10.1373/clinchem
6. Bates DW, Goldman L, Lee TH. 1991. Contaminant blood cultures and .2010.157131.
resource utilization: the true consequences of false-positive results. JAMA 25. Surdulescu S, Utamsingh D, Shekar R. 1998. Phlebotomy teams reduce
265:365–369. https://doi.org/10.1001/jama.1991.03460030071031. blood-culture contamination rate and save money. Clin Perf Qual
7. Gander RM, Byrd L, DeCrescenzo M, Hirany S, Bowen M, Baughman J. Health Care 6:60 – 62.
2009. Impact of blood cultures drawn by phlebotomy on contamina- 26. Weinbaum FI, Lavie S, Danek M, Sixsmith D, Heinrich GF, Mills SS. 1997.
tion rates and health care costs in a hospital emergency department. J Doing it right the first time: quality improvement and the contaminant
Clin Microbiol 47:1021–1024. https://doi.org/10.1128/JCM.02162-08. blood culture. J Clin Microbiol 35:563–565.
8. Zwang O, Albert RK. 2006. Analysis of strategies to improve cost 27. Norberg A, Christopher NC, Ramundo ML, Bower JR, Berman SA. 2003.
effectiveness of blood cultures. J Hosp Med 1:272–276. https://doi.org/
Contamination rates of blood cultures obtained by dedicated phlebot-
10.1002/jhm.115.
omy vs intravenous catheter. JAMA 289:726 –729. https://doi.org/10
9. Rupp M, Cavalieri S, Marolf C, Lyden E. 2017. Reduction in blood culture
.1001/jama.289.6.726.
contamination through use of initial specimen diversion device. Clin
28. King TC, Price PB. 1963. An evaluation of iodophors as skin antiseptics.
Infect Dis 65:201–205. https://doi.org/10.1093/cid/cix304.
Surg Gynecol Obstet 116:361–365.

Downloaded from https://journals.asm.org/journal/cmr on 24 July 2024 by 2409:40f4:1:3c3a:456a:7f5a:e81e:35f9.

10. Archer G. 1985. Coagulase-negative staphylococci in blood cultures:
29. Boyce JM, Nadeau J, Dumigan D, Miller D, Dubowsky C, Reilly L, Hannon
the clinician’s dilemma. Infect Control 6:477– 478. https://doi.org/10
CV. 2013. Obtaining blood cultures by venipuncture versus from central
.1017/S019594170006358X.
lines: impact on blood culture contamination rates and potential effect
11. Kim S-D, McDonald LC, Jarvis WR, McAllister SK, Jerris R, Carson LA,
on central line-associated bloodstream infection reporting. Infect Con-
Miller JM. 2000. Determining the significance of coagulase-negative
trol Hosp Epidemiol 34:1042–1047. https://doi.org/10.1086/673142.
staphylococci isolated from blood cultures at a community hospital: a
30. Garcia RA, Spitzer ED, Beaudry J, Beck C, Diblasi R, Gilleeny-Blabac M,
role for species and strain identification. Infect Control Hosp Epidemiol
Haugaard C, Heuschneider S, Kranz BP, McLean K, Morales KL, Owens S,
21:213–217. https://doi.org/10.1086/501747.
12. Kirchhoff LV, Sheagren JN. 1985. Epidemiology and clinical significance Paciella ME, Torregrosa E. 2015. Multidisciplinary team review of best
of blood cultures positive for coagulase-negative staphylococci. Infect practices for collection and handling of blood cultures to determine
Control 6:479 – 486. https://doi.org/10.1017/S0195941700063591. effective interventions for increasing the yield of true-positive bacte-
13. Rupp ME, Archer GL. 1994. Coagulase-negative staphylococci: patho- remias, reducing contamination, and eliminating false-positive central
gens associated with medical progress. Clin Infect Dis 19:231–243. line-associated bloodstream infections. Am J Infect Control 43:
https://doi.org/10.1093/clinids/19.2.231. 1222–1237. https://doi.org/10.1016/j.ajic.2015.06.030.
14. Schifman RB, Strand CL, Meier FA, Howanitz PJ. 1998. Blood culture 31. Jorgensen JH, Mirrett S, McDonald LC, Murray PR, Weinstein MP, Fune
contamination. A College of American Pathologists Q-Probes study J, Trippy CW, Masterson M, Reller LB. 1997. Controlled clinical labora-
involving 640 institutions and 497134 specimens from adult patients. tory comparison of BACTEC Plus Aerobic/F resin medium with BacT/
Arch Pathol Lab Med 122:216 –221. Alert Aerobic FAN medium for detection of bacteremia and fungemia.
15. Wilson ML, Mitchell M, Morris A. 2007. Principles and procedures for J Clin Microbiol 35:53–58.
blood cultures; approved guideline. CLSI document M47-A. Clinical and 32. McDonald LC, Fune J, Gaido LB, Weinstein MP, Reimer LG, Flynn TM,
Laboratory Standards Institute, Wayne, PA. Wilson ML, Mirrett S, Reller LB. 1996. Clinical importance of the in-
16. Baron EJ, Weinstein MP, Dunne WM, Jr, Yagupsky P, Welch DF, Wilson creased sensitivity of BacT/Alert FAN aerobic and anaerobic blood
DM. 2005. Cumitech 1C, blood cultures IV. ASM Press, Washington, DC. culture bottles. J Clin Microbiol 34:2180 –2184.
17. Weinstein MP, Reller LB, Murphy JR, Lichtenstein KA. 1983. The clinical 33. Smith JA, Bryce EA, Ngui-Yen JH, Roberts FJ. 1995. Comparison of
significance of positive blood cultures: a comprehensive analysis of 500 BACTEC 9240 and BacT/Alert blood culture systems in an adult hospital.
episodes of bacteremia and fungemia in adults. I. Laboratory and J Clin Microbiol 33:1905–1908.
epidemiologic observations. Rev Infect Dis 5:35–53. https://doi.org/10 34. Weinstein MP, Mirrett S, Reimer LG, Wilson ML, Smith-Elekes S, Chuard
.1093/clinids/5.1.35. CR, Joho KL, Reller LB. 1995. Controlled evaluation of BacT/Alert stan-
18. Washer LL, Chenoweth C, Kim H-W, Rogers MAM, Malani AN, Riddell T, dard aerobic and FAN aerobic blood culture bottles for detection of
IV, Kuhn L, Noeyack B, Jr, Neusius H, Newton DW, Saint S, Flanders SA. bacteremia and fungemia. J Clin Microbiol 33:978 –981.
2013. Blood culture contamination: a randomized trial evaluating the 35. Weinstein MP, Mirrett S, Wilson ML, Harrell LJ, Stratton CW, Reller LB.
comparative effectiveness of 3 skin antiseptic interventions. Infect 1991. Controlled evaluation of BACTEC Plus 26 and Roche Septi-Chek
Control Hosp Epidemiol 34:15–21. https://doi.org/10.1086/668777. aerobic blood culture bottles. J Clin Microbiol 29:879 – 882.
19. Alahmadi YM, Aldeyab MA, McElnay JC, Scott MG, Darwish Elhajji FW, 36. Wilson ML, Weinstein MP, Mirrett S, Reimer LG, Feldman RJ, Chuard CR,
Magee FA, Dowds M, Edwards C, Fullerton L, Tate A, Kearney MP. 2011. Reller LB. 1995. Controlled evaluation of BacT/Alert standard anaerobic
Clinical and economic impact of contaminated blood cultures within and FAN anaerobic blood culture bottles for the detection of bactere-
the hospital setting. J Hosp Infect 77:233–236. https://doi.org/10.1016/ mia and fungemia. J Clin Microbiol 33:2265–2270.
j.jhin.2010.09.033. 37. Isaacman D, Karasic R. 1990. Lack of effect of changing needles on

January 2020 Volume 33 Issue 1 e00009-19 cmr.asm.org 17

Doern et al. Clinical Microbiology Reviews

contamination of blood cultures. Pediatr Infect Dis J 9:274 –278. https:// identification of microorganisms in positive blood cultures. Biomed Res
doi.org/10.1097/00006454-199004000-00010. Int 21:7013470. https://doi.org/10.1155/2018/7013470.
38. Krumholz HM, Cummings S, York M. 1990. Blood culture phlebotomy: 58. Lee CC, Lin WJ, Shih HI, Wu CH, Chen PL, Lee HC, Lee NY, Chang CM,
switching needles does not prevent contamination. Ann Intern Med Wang LR, Ko WC. 2007. Clinical significance of potential contaminants
113:290 –292. https://doi.org/10.7326/0003-4819-113-4-290. in blood cultures among patients in a medical center. J Microbiol
39. Leisure MK, Moore DM, Schwartzman JD, Hayden GF, Donowitz LG. Immunol Infect 40:438 – 444.
1990. Changing the needle when inoculating blood cultures: a no- 59. van der Heijden YF, Miller G, Wright PW, Shepherd BE, Daniels TL,
benefit and high-risk procedure. JAMA 264:2111–2112. https://doi.org/ Talbot TR. 2011. Clinical impact of blood cultures contaminated with
10.1001/jama.1990.03450160081034. coagulase-negative staphylococci at an academic medical center. In-
40. Spitalnic SJ, Wollard RH, Mermel LA. 1995. The significance of changing fect Control Hosp Epidemiol 32:623– 625. https://doi.org/10.1086/
needles when inoculating blood cultures: a meta-analysis. Clin Infect 660096.
Dis 21:1003–1006. 60. Bell J, Goyal M, Long S, Kumar A, Friedrich J, Garfinkel J, Chung S,
41. MacGregor RR, Beaty HN. 1972. Evaluation of positive blood cultures. Fitzgibbons S. 2018. Anatomic site-specific complication rates for central
Guidelines for early differentiation of contaminated from valid positive venous catheter insertions. J Intensive Care Med 19:885066618795126.
cultures. Arch Intern Med 130:84 – 88. https://doi.org/10.1001/archinte https://doi.org/10.1177/0885066618795126.
.1972.03650010072013. 61. Hauck K, Zhao X. 2011. How dangerous is a day in the hospital? A
42. Anonymous. 1992. Clinical Laboratory Improvement Amendments of model of adverse events and length of stay for medical inpatients. Med
1988. Register F, http://www.gpo.gov/fdsys/pkg/STATUTE-102/pdf/ Care 49:1068 –1075. https://doi.org/10.1097/MLR.0b013e31822efb09.
STATUTE-102-Pg2903.pdf. 62. Banerjee R, Teng CB, Cunningham SA, Ihde SM, Steckelberg JM, Mori-
43. Beekmann SR, Diekema DJ, Doern GV. 2005. Determining the clinical arty JP, Shah ND, Mandrekar JN, Patel R. 2015. Randomized trial of rapid
significance of coagulase-negative staphylococci isolated from blood multiplex polymerase chain reaction-based blood culture identification
cultures. Infect Control Hosp Epidemiol 26:559. https://doi.org/10.1086/ and susceptibility testing. Clin Infect Dis 61:1071–1080. https://doi.org/
502584. 10.1093/cid/civ447.
44. Hall KK, Lyman JA. 2006. Updated review of blood culture contamina- 63. Nagel JL, Huang AM, Kunapuli A, Gandhi TN, Washer LL, Lassiter J, Patel
tion. Clin Microbiol Rev 19:788 – 802. https://doi.org/10.1128/CMR T, Newton DW. 2014. Impact of antimicrobial stewardship intervention
.00062-05. on coagulase-negative Staphylococcus blood 368 cultures in conjunc-
45. Riedel S, Carroll KC. 2010. Blood cultures: key elements for best prac- tion with rapid diagnostic testing. J Clin Microbiol 52:2849 –2854.
tices and future directions. J Infect Chemother 16:301–316. https://doi https://doi.org/10.1128/JCM.00682-14.
.org/10.1007/s10156-010-0069-1. 64. Timbrook TT, Morton JB, McConeghy WK, Caffrey AR, Mylonakis E,
46. Mirrett S, Weinstein MP, Reimer LG, Wilson ML, Reller LB. 2001. Relevance LaPlante KL. 2017. The effect of molecular rapid diagnostic testing on
of the number of positive bottles in determining clinical significance of clinical outcomes in bloodstream infections: a systematic review and
coagulase-negative staphylococci in blood cultures. J Clin Microbiol 39:
meta-analysis. Clin Infect Dis 64:15–23. https://doi.org/10.1093/cid/
3279–3281. https://doi.org/10.1128/jcm.39.9.3279-3281.2001.
ciw649.
47. Baron EJ, Scott JD, Tompkins LS. 2005. Prolonged incubation and
65. Centers for Disease Control and Prevention. December 2018. National
extensive subculturing do not increase recovery of clinically significant
Healthcare Safety Network newsletter. https://www.cdc.gov/nhsn/
microorganisms from standard automated blood cultures. Clin Infect
newsletters/index.html.

Downloaded from https://journals.asm.org/journal/cmr on 24 July 2024 by 2409:40f4:1:3c3a:456a:7f5a:e81e:35f9.

Dis 41:1677–1680. https://doi.org/10.1086/497595.
66. Waltzman ML, Harper M. 2001. Financial and clinical impact of false-
48. Ilstrup D. 1978. Sepsis: organisms from blood cultures at the Mayo
positive blood culture results. Clin Infect Dis 33:296 –299. https://doi
Clinic: 1968 –1975, p 23–25. In Washington J, II (ed), The detection of
.org/10.1086/321881.
septicemia. CRC Press, West Palm Beach, FL.
67. Segal GS, Chamberlain JM. 2000. Resource utilization and contami-
49. Krisher KK, Whyburn DR, Koepnick FE. 1993. Comparison of the BacT/
nated blood cultures in children at risk for occult bacteremia. Arch
Alert pediatric blood culture system, Pedi-BacT, with conventional
Pediatr Adolesc Med 154:469 – 473. https://doi.org/10.1001/archpedi
culture using the 20-millimeter Becton-Dickinson supplemented pep-
.154.5.469.
tone broth bottle. J Clin Microbiol 31:793–797.
50. Kennedy GT, Barr JG, Goldsmith C. 1995. Detection of bacteraemia by 68. Little JR, Murray PR, Traynor PS, Spitznagel E. 1999. A randomized trial
the continuously monitoring BacT/Alert system. J Clin Pathol 48: of povidone-iodine compared with iodine tincture for venipuncture
912–914. https://doi.org/10.1136/jcp.48.10.912. site disinfection: effects on rates of blood culture contamination. Am J
51. Banerjee R, Ozenci V, Patel R. 2016. Individualized approaches are Med 107:119 –125. https://doi.org/10.1016/s0002-9343(99)00197-7.
needed for optimized blood cultures. Clin Infect Dis 63:1332–1339. 69. Geisler BP, Jilq N, Patton RG, Pietzsch JB. 27 March 2019. A model to
https://doi.org/10.1093/cid/ciw573. evaluate the impact of hospital-based interventions targeting false-
52. Riedel S, Carroll KC. 2016. Early identification and treatment of patho- positive blood cultures on economic and clinical outcomes. J Hosp
gens in sepsis: molecular diagnostics and antibiotic choice. Clin Chest Infect. https://doi.org/10.1016/j.jhin.2019.03.012.
Med 37:191–207. https://doi.org/10.1016/j.ccm.2016.01.018. 70. Self WH, Talbot TR, Paul BR, Collins SP, Ward MJ. 2014. Cost analysis of
53. Messacar K, Hurst AL, Child J, Campbell K, Palmer C, Hamilton S, Dowell strategies to reduce blood culture contamination in the emergency
E, Robinson CC, Parker SK, Dominguez SR. 2017. Clinical impact and department: sterile collection kits and phlebotomy teams. Infect Con-
provider acceptability of real-time antimicrobial stewardship decision trol Hosp Epidemiol 35:1021–1028. https://doi.org/10.1086/677161.
support for rapid diagnostics in children with positive blood culture 71. Skoglund E, Dempsey C, Chen H, Garey KW. 2019. Estimated clinical
results. J Pediatric Infect Dis Soc 6:267–274. and economic impact through use of a novel blood collection device to
54. Box MJ, Sullivan EL, Ortwine KN, Parmenter MA, Quigley MM, Aguilar- reduce blood culture contamination in the emergency department: a
Higgins LM, MacIntosh CL, Goerke KF, Lim RA. 2015. Outcomes of rapid cost-benefit analysis. J Clin Microbiol 57:e01015-18. https://doi.org/10
identification for gram-positive bacteremia in combination with anti- .1128/JCM.01015-18.
biotic stewardship at a community-based hospital system. Pharmaco- 72. Wiggers JB, Xiong W, Daneman N. 2016. Sending repeat cultures: is there
therapy 35:269 –276. https://doi.org/10.1002/phar.1557. a role in the management of bacteremic episodes? (SCRIBE study). BMC
55. Pliakos EE, Andreatos N, Shehadeh F, Ziakas PD, Mylonakis E. 2018. The Infect Dis 16:286. https://doi.org/10.1186/s12879-016-1622-z.
cost effectiveness of rapid diagnostic testing for the diagnosis of 73. Canzoneri CN, Akhavan BJ, Tosur Z, Andrade PEA, Aisenberg GM. 2017.
bloodstream infections with or without antimicrobial stewardship. Clin Follow-up blood cultures in gram-negative bacteremia: are they
Microbiol Rev 31:e00095-17. https://doi.org/10.1128/CMR.00095-17. needed? Clin Infect Dis 65:1776 –1779. https://doi.org/10.1093/cid/
56. Lamy B, Dargère Arendrup MC, Parienti JJ, Tattevin P. 2016. How to cix648.
optimize the use of blood cultures for the diagnosis of bloodstream 74. Eliakim-Raz N, Bates DW, Leibovici L. 2015. Predicting bacteraemia in
infections? A state-of-the art. Front Microbiol 7:697. https://doi.org/10 validated models—a systematic review. Clin Microbiol Infect 21:
.3389/fmicb.2016.00697. 295–301. https://doi.org/10.1016/j.cmi.2015.01.023.
57. Verhoeven PO, Haddar CH, Rigaill J, Fonsale N, Carricajo A, Grattard F, 75. Woods-Hill CZ, Fackler J, McMillan KN, Ascenzi J, Martinez DA, Toerper
Pozzetto B. 2018. Comparison of the fully automated FilmArray BCID MF, Voskertchian A, Colantuoni E, Klaus SA, Levin S, Milstone AM. 2017.
assay to a 4-hour culture test coupled to mass spectrometry for day 0 Association of a clinical practice guideline with blood culture use in

January 2020 Volume 33 Issue 1 e00009-19 cmr.asm.org 18

Blood Culture Contamination Clinical Microbiology Reviews

critically ill children. JAMA Pediatr 17:157–164. https://doi.org/10.1001/ contaminated cultures in neonates. J Paediatr Child Health 49:105–108.
jamapediatrics.2016.3153. https://doi.org/10.1111/jpc.12088.
76. Mimoz O, Karim A, Mercat A, Cosseron M, Falissard B, Parker F, Richard 94. Pourcyrous M, Korones SB, Bada HS, Patterson T, Baselski V. 1988.
C, Samii K, Nordmann P. 1999. Chlorhexidine compared with povidone- Indwelling umbilical arterial catheter: a preferred sampling site for
iodine as skin preparation before blood culture. A randomized, con- blood culture. Pediatrics 81:821– 825.
trolled trial. Ann Intern Med 131:834 – 837. https://doi.org/10.7326/ 95. Hall RT, Domenico HJ, Self WH, Hain PD. 2013. Reducing the blood
0003-4819-131-11-199912070-00006. culture contamination rate in a pediatric emergency department and
77. Barenfanger J, Drake C, Lawhorn J, Verhulst SJ. 2004. Comparison of subsequent cost savings. Pediatrics 31:e292– e297. https://doi.org/10
chlorhexidine and tincture of iodine for skin antisepsis in preparation .1542/peds.2012-1030.
for blood sample collection. J Clin Microbiol 42:2216 –2217. https://doi 96. Mermel LA. 6 March 2019. Drawing blood cultures through intravas-
.org/10.1128/JCM.42.5.2216-2217.2004. cular catheters: controversy and update. Infect Control Hosp Epidemiol.
78. Trautner BW, Clarridge JE, Darouiche RO. 2002. Skin antisepsis kits https://doi.org/10.1017/ice.2019.37.
containing alcohol and chlorhexidine gluconate or tincture of iodine 97. York MK. 1990. Bacillus species pseudobacteremia traced to contami-
are associated with lower rates of blood culture contamination. Infect nated gloves used in collection of blood from patients with acquired
Control Hosp Epidemiol 23:397– 401. https://doi.org/10.1086/502073. immunodeficiency syndrome. J Clin Microbiol 28:2114 –2116.
79. Caldeira D, David C, Sampaio C. 2011. Skin antiseptics in venous 98. Kim N-H, Kim M, Lee S, Yun NR, Kim K-H, Park SW, Kim HB, Kim N-J, Kim E-C,
puncture-site disinfection for prevention of blood culture Park WB, Oh M-D. 2011. Effect of routine sterile gloving on contamination
contamination: systematic review with meta-analysis. J Hosp Infection rates in blood culture: a cluster randomized trial. Ann Intern Med 154:
77:223–232. https://doi.org/10.1016/j.jhin.2010.10.015. 145–151. https://doi.org/10.7326/0003-4819-154-3-201102010-
80. Maiwald M, Chan ES. 2012. The forgotten role of alcohol: a systematic 00003.
review and meta analysis of the clinical efficacy and perceived role of 99. Park WB, Myung SJ, Oh M-D, Lee J, Kim N-J, Kim E-C, Park JS. 2015.
chlorhexidine in skin antisepsis. PLoS One 7:e44277. https://doi.org/10 Educational intervention as an effective step for reducing blood culture
.1371/journal.pone.0044277. contamination: a prospective cohort study. J Hosp Infect 91:111–116.
81. Martinez J, Macias JH, Arreguin V, Alvarez JA, Macias AE, Mosqueda- https://doi.org/10.1016/j.jhin.2015.04.022.
Gomez JL. 2017. Isopropyl alcohol is as efficient as chlorhexidine to 100. Self WH, Mickanin J, Grijalva CG, Grant FH, Henderson MC, Corley G,
prevent contamination of blood cultures. Am J Infect Control 45: Blaschke DG, II, McNaughton CD, Barrett TW, Talbot TR, Paul BR. 2014.
350 –353. https://doi.org/10.1016/j.ajic.2016.11.027. Reducing blood culture contamination in community hospital emer-
82. Tarrand JJ, LaSala PR, Han XY, Rolston KV, Kontoyiannis DP. 2012. gency departments: a multicenter evaluation of a quality improvement
Dimethyl sulfoxide enhances effectiveness of skin antiseptics and re- intervention. Acad Emerg Med 21:274 –282. https://doi.org/10.1111/
duces contamination rates of blood cultures. J Clin Microbiol 50: acem.12337.
1552–1557. https://doi.org/10.1128/JCM.05106-11. 101. Bowen CM, Coleman T, Cunningham D. 2016. Reducing blood culture
contaminations in the emergency department: it takes a team. J Emerg
83. Emergency Nurses Association. 2016. Clinical practice guideline: pre-
Nurs 42:306 –311. https://doi.org/10.1016/j.jen.2015.10.021.
vention of blood culture contamination. https://www.ena.org/docs/
102. Denno J, Gannon M. 2013. Practical steps to lower blood culture
default-source/resource-library/practice-resources/cpg/
contamination rates in the emergency department. J Emerg Nurs
bcccpg2c37f1815b664d2fa8d7e9fd0f475a41.pdf?sfvrsn⫽6d1899fb_12.
39:459 – 464. https://doi.org/10.1016/j.jen.2012.03.006.
84. Department of Health. 2007. Saving lives: taking blood cultures-a sum-

Downloaded from https://journals.asm.org/journal/cmr on 24 July 2024 by 2409:40f4:1:3c3a:456a:7f5a:e81e:35f9.

103. Thomas S, Cheesbrough J, Plumb S, Bolton L, Wilkinson P, Walmsley J,
mary of best practice. https://webarchive.nationalarchives.gov.uk/
Diggle P. 2011. Impact of a blood culture collection kit on the quality
20120118171812/http://hcai.dh.gov.uk/files/2011/03/Document_Blood
of blood culture sampling: fear and the law of unintended conse-
_culture_FINAL_100826.pdf.
quences. J Hosp Infect 78:256 –259. https://doi.org/10.1016/j.jhin.2011
85. Septimus E. 2015. Clinician guide for collecting cultures. https://www
.04.012.
.cdc.gov/antibiotic-use/healthcare/implementation/clinicianguide
104. Madeo M, Jackson T, Williams C. 2005. Simple measures to reduce the
.html.
rate of contamination of blood cultures in accident and emergency.
86. Falagas ME, Kazantzi MS, Bliziotis IA. 2008. Comparison of utility of
Emerg Med J 22:810 – 811. https://doi.org/10.1136/emj.2005.003079.
blood cultures from intravascular catheters and peripheral veins: a 105. Dhillon RH, Clark J, Azadian BS. 2009. Reducing blood culture contam-
systematic review and decision analysis. J Med Microbiol 57:1– 8. ination. J Hosp Infect 73:97–99. https://doi.org/10.1016/j.jhin.2009.06
https://doi.org/10.1099/jmm.0.47432-0. .002.
87. Centers for Disease Control and Prevention. 2013. The National Health- 106. Mermel LA, Maki DG. 1993. Detection of bacteremia in adults: conse-
care Safety Network (NHSN) manual: patient safety component proto- quences of culturing an inadequate volume of blood. Ann Intern Med
col. Division of Healthcare Quality Promotion, National Center for 119:270–272. https://doi.org/10.7326/0003-4819-119-4-199308150-00003.
Infectious Diseases, Atlanta, GA. 107. Gonsalves WI, Cornish N, Moore M, Chen A, Varman M. 2009. Effects of
88. Levin PD, Moss J, Stohl S, Fried E, Cohen MJ, Sprung CL, Benenson S. volume and site of blood draw on blood culture results. J Clin Microbiol
2013. Use of the non-wire central line hub to reduce blood culture 47:3482–3485. https://doi.org/10.1128/JCM.02107-08.
contamination. Chest 143:640 – 645. https://doi.org/10.1378/chest.12 108. Alfa M, Sanche S, Roman S, Fiola Y, Lenton P, Harding G. 1995. Con-
-0863. tinuous quality improvement for introduction of automated blood
89. Mermel LA, Allon M, Bouza E, Craven DE, Flynn P, O’Grady NP, Raad II, culture instrument. J Clin Microbiol 33:1185–1191.
Rijnders BJA, Sherertz RJ, Warren DK. 2009. Clinical practice guidelines 109. Meessen NEL, Jacobs JA. 1998. Blood volume in BACTEC(PLUS)/F cul-
for the diagnosis and management of intravascular catheter-related ture bottles sampled using the direct-draw technique. Clin Microbiol
infection: 2009 update by the Infectious Diseases Society for America. Infect 4:471– 472. https://doi.org/10.1111/j.1469-0691.1998.tb00399.x.
Clin Infect Dis 49:1– 45. https://doi.org/10.1086/599376. 110. Nair A, Elliott SP, Al Mohajer M. 2017. Knowledge, attitude, and practice
90. Sherertz RJ, Karchmer TB, Palavecino E, Bischoff W. 2011. Blood drawn of blood culture contamination: a multicenter study. Am J Infect Con-
through valved catheter hub connectors carries a significant risk of trol 45:547–548. https://doi.org/10.1016/j.ajic.2017.01.008.
contamination. Eur J Clin Microbiol Infect Dis 30:1571–1577. https://doi 111. Rowley S. 2012. Blood culture collection: quality assurance for ANTT. Br
.org/10.1007/s10096-011-1262-6. J Nurs 21:158. https://doi.org/10.12968/bjon.2012.21.3.158.
91. Guembe M, Rodríguez-Créixems M, Sánchez-Carrillo C, Pérez-Parra A, 112. Weightman NC, Kerr KG. 2012. Blood culture contamination: having
Martín-Rabadán P, Bouza E. 2010. How many lumens should be cul- your cake and eating it. J Hosp Infect 80:101–102. https://doi.org/10
tured in the conservative diagnosis of catheter-related bloodstream .1016/j.jhin.2011.08.023.
infections? Clin Infect Dis 50:1575–1579. https://doi.org/10.1086/ 113. Bekeris LG, Tworek JA, Walsh MK, Valenstein PN. 2005. Trends in blood
652766. culture contamination: a College of American Pathologists Q-tracks
92. Moeller D. 2017. Eliminating blood culture false positives: harnessing study of 356 institutions. Arch Pathol Lab Med 129:1222–1225. https://
the power of nursing shared governance. J Emerg Nurs 43:126 –132. doi.org/10.1043/1543-2165(2005)129[1222:TIBCCA]2.0.CO;2.
https://doi.org/10.1016/j.jen.2016.07.001. 114. Mtunthama N, Gordon SB, Kusimbwe T, Zijlstra EE, Molyneux ME,
93. McLaughlin LM, Inglis GD, Hoellering AB, Davies MW. 2013. Relation- French N. 2008. Blood culture collection technique and pneumococcal
ship between blood culture collection method and proportion of surveillance in Malawi during the four-year period 2003–2006: an

January 2020 Volume 33 Issue 1 e00009-19 cmr.asm.org 19

Doern et al. Clinical Microbiology Reviews

observational study. BMC Infect Dis 8:137–138. https://doi.org/10.1186/ 125. Gibson T, Norris W. 1958. Skin fragments removed by injection needles.
1471-2334-8-137. Lancet ii:983–985. https://doi.org/10.1016/s0140-6736(58)90475-6.
115. Larkin S, Baker N, Anderson R, Ward S, Forde S. 2010. An interactive 126. Buchta C, Nedorost N, Regele H, Egerbacher M, Kormoczi G, Hocker P,
approach to reducing blood culture contamination. J Hosp Infect Dettke M. 2005. Skin plugs in phlebotomy puncture for blood dona-
76:273–275. https://doi.org/10.1016/j.jhin.2010.06.001. tion. Wien Klin Wochenschr 117:141–144. https://doi.org/10.1007/
116. Ramirez P, Gordón M, Cortes C, Villarreal E, Perez-Belles C, Robles C, de s00508-005-0310-6.
Hevia L, Marti JV, Botella J, Bonastre J. 2015. Blood culture contamina- 127. Patton RG, Schmitt T. 2010. Innovation for reducing blood culture
tion rate in an intensive care setting: effectiveness of an education- contamination: initial specimen diversion technique. J Clin Microbiol
based intervention. Am J Infect Control 43:844 – 847. https://doi.org/10 48:4501– 4503. https://doi.org/10.1128/JCM.00910-10.
.1016/j.ajic.2015.04.183. 128. Binkhamis K, Forward K. 2014. Effect of the initial specimen diversion
117. Hopkins K, Huynh S, McNary C, Walker A, Nixon R, Craighead JE. 2013. technique on blood culture contamination rates. J Clin Microbiol 52:
Reducing blood culture contamination rates: a systematic approach to 980 –981. https://doi.org/10.1128/JCM.02773-13.
improving quality of care. Am J Infect Control 41:1272–1274. https:// 129. Lalezari A, Cohen NJ, Svinik O, Tel-Zur O, Sinvani S, Block C, Moses AE,
doi.org/10.1016/j.ajic.2013.02.019. Oster Y, Strahilevitz J. A simplified blood culture sampling protocol for
118. Self WH, Speroff T, Grijalva CG, McNaughton CD, Ashburn J, Liu D, reducing contamination and costs: a randomized controlled trial. Clin
Arbogast PG, Russ S, Storrow AB, Talbot TR. 2013. Reducing blood Microbiol Infect, in press.
culture contamination in the emergency department: an interrupted
130. Bell M, Bogar C, Plante J, Rasmussen K, Winters S. 2018. Effectiveness of
time series quality improvement study. Acad Emerg Med 20:89 –97.
a novel specimen collection system in reducing blood culture contam-
https://doi.org/10.1111/acem.12057.
ination rates. J Emerg Nurs 44:570 –575. https://doi.org/10.1016/j.jen
119. Pronovost PJ. 2011. Navigating adaptive challenges in quality improve-
.2018.03.007.
ment. BMJ Qual Saf 20:560 –563. https://doi.org/10.1136/bmjqs-2011
131. Zimmereman FS, Assous MV, Zevin S, Wiener-Well Y. 10 January 2019.
-000026.
Reducing blood culture contamination using an initial specimen diver-
120. Zimmerman FS, Assous MV, Yinnon AM, Wiener-Well Y. 2018. Reducing
blood culture contamination using a departmental report card. J Hosp sion device. Am J Infect Control. https://doi.org/10.1016/j.ajic.2018.12
Infect 99:236 –237. https://doi.org/10.1016/j.jhin.2018.02.023. .004.
121. Youssef D, Shams W, Bailey B, O’Neil TJ, Al-Abbadi MA. 2012. Effective 132. O’Sullivan DM, Steer l. 2019. Reducing false-positive blood cultures:
strategy for decreasing blood culture contamination rates: the experi- using a blood diversion device. Conn Med 83:54 –55.
ence of a Veterans Affairs medical center. J Hosp Infect 81:288 –291. 133. Diekema DJ. 2017. Rising stakes for healthcare associated infection
https://doi.org/10.1016/j.jhin.2012.05.014. prevention: implications for the clinical microbiology laboratory. J Clin
122. Harding AD, Bollinger S. 2013. Reducing blood culture contamination Microbiol 55:996 –1001. https://doi.org/10.1128/JCM.02544-16.
rates in the emergency department. J Emerg Nurs 39:e1– e6. https:// 134. Centers for Disease Control and Prevention. 2019. Surveillance for blood-
doi.org/10.1016/j.jen.2012.10.009. stream infections. https://www.cdc.gov/nhsn/acute-care-hospital/
123. Gibb AP, Hill B, Chorel B, Brant R. 1997. Reduction in blood culture clabsi/index.html.
contamination rate by feedback to phlebotomists. Arch Pathol Lab 135. Centers for Disease Control and Prevention. 2019. National Healthcare
Med 121:503–507. Safety Network Master Organism List. https://www.cdc.gov/nhsn/xls/
124. Lin C-M, Lee W-S, Lin F-Y, Yu F-L, Ou T-Y, Teng S-O. 2012. Reducing master-organism-com-commensals-lists.xlsx.

Downloaded from https://journals.asm.org/journal/cmr on 24 July 2024 by 2409:40f4:1:3c3a:456a:7f5a:e81e:35f9.

blood culture contamination rates by educational intervention and 136. Centers for Disease Control and Prevention. 2019. Bloodstream infec-
one-on-one feedback in the emergency department. J Exp Clin Med tion event definitions. https://www.cdc.gov/nhsn/pdfs/pscmanual/
4:154 –156. https://doi.org/10.1016/j.jecm.2012.04.005. 4psc_clabscurrent.pdf.

Gary V. Doern, Ph.D., is a professor emeritus Karen C. Carroll, M.D., is a professor of pa-
of pathology and medicine at the University thology at The Johns Hopkins University
of Iowa Carver College of Medicine. He re- (JHU) School of Medicine. Dr. Carroll re-
ceived his B.S. degree from Northwestern Uni- ceived her B.A. degree in biology from the
versity and his Ph.D. degree in pathology from College of Notre Dame of Maryland and her
the Medical College of Wisconsin. Following 2 M.D. degree from the University of Maryland
years of postdoctoral training in clinical micro- School of Medicine. She completed a resi-
biology at the University of Oregon Health dency in Internal Medicine in Rochester, NY,
Sciences Center in Portland, he served for 35 in the Associated Hospitals Program, Univer-
years as the Director of the Clinical Microbiol- sity of Rochester School of Medicine. Follow-
ogy Laboratories, first at the University of ing a clinical fellowship in Infectious Dis-
Massachusetts Medical Center in Worcester, MA, and then at the Uni- eases at the University of Massachusetts Medical Center, Dr. Carroll did
versity of Iowa Carver College of Medicine in Iowa City. He retired in a second fellowship in medical microbiology at the University of Utah
2008. He was Chairman of the Clinical Microbiology Division of the School of Medicine. She was granted a faculty appointment in the
American Society for Microbiology, Head of the American Board of Department of Pathology and was the Director of the Infectious Dis-
Medical Microbiology (ABMM), a member of the ICAAC program com- eases Laboratories at ARUP Inc. for 12 years. In 2002, Dr. Carroll joined
mittee, and for 9 years a voting member of the Clinical and Laboratory the faculty at JHU as the Director of the Division of Medical Microbiol-
Standards Institute (formerly the NCCLS). He served for 5 years as the ogy. She is a Fellow of the American Academy of Microbiology, the
editor in chief of the Journal of Clinical Microbiology as well as on the Infectious Diseases Society of America, and the College of American Pa-
editorial boards of seven other journals in the disciplines of infectious thologists. Dr. Carroll’s focused area of clinical research at JHU is in the
diseases and clinical microbiology. He is an elected fellow of the diagnosis of health care-associated infections and bacteremia/sepsis.
American Academy of Microbiology and the Infectious Diseases Society
of America and also a diplomat of the ABMM. He is the recipient of the
ASM’s Becton-Dickinson Award and the ABMLI Award for contributions
to the field of clinical microbiology.

January 2020 Volume 33 Issue 1 e00009-19 cmr.asm.org 20

Blood Culture Contamination Clinical Microbiology Reviews

Daniel J. Diekema, M.D., is a professor of Melvin P. Weinstein, M.D., is board certified

internal medicine and pathology at the Uni- in internal medicine, infectious diseases, and
versity of Iowa Carver College of Medicine. medical microbiology. He is professor of
He is the director of the Division of Infectious medicine and pathology at Rutgers Robert
Diseases, associate director of the Clinical Wood Johnson Medical School, where he
Microbiology Laboratory, and associate hos- was chief of the Division of Infectious Dis-
pital epidemiologist at the University of Iowa eases from 2001 to 2019. A fellow of the
Healthcare. Dr. Diekema’s clinical and re- American Academy of Microbiology and In-
search interests include the role of the diag- fectious Diseases Society of America, Dr.
nostic laboratory in infection prevention and Weinstein was a voting member of the CLSI
antimicrobial stewardship. He has served on Subcommittee on Antimicrobial Susceptibil-
national committees that establish standards in clinical microbiology ity Testing for 18 years and currently serves as its chair. He also serves
and infection prevention, including the Clinical and Laboratory Stan- on the CLSI Committee on Policies and Processes for Blood Cultures. He
dards Institute (CLSI) and the CDC’s Healthcare Infection Control Prac- was previously a member of the FDA Anti-Infective Drug Advisory
tices Advisory Committee (HICPAC), and he is a past president of the Committee and the FDA Microbiology Devices Panel and served as a
Society for Healthcare Epidemiology of America (SHEA). Trustee of the ABMM. He directed the Microbiology Laboratory at RWJ
University Hospital from 1983 to 2015 and currently is codirector. He
has received the BD Award for Research in Clinical Microbiology,
Kevin W. Garey, Pharm.D., M.S., FASHP, is a BioMerieux Sonnenwirth Award for Leadership in Clinical Microbiology,
professor at the University of Houston Col- and ABMM/ABMLI Professional Recognition Award, all from ASM.
lege of Pharmacy and Chair of the Depart-
ment of Pharmacy Practice and Translational
Research. He is an adjunct professor at the Daniel J. Sexton, M.D., is a professor of
University of Texas School of Public Health medicine in the Division of Infectious Dis-
and a clinical specialist and researcher at eases at Duke University Medical Center. He
Baylor St. Luke’s Medical Center, Houston, has been active in all aspects of hospital
TX. He received a bachelor of science in epidemiology since 1977 and was the hos-
pharmacy degree from Dalhousie University pital epidemiologist at Duke for 25 years. His
in Halifax, Nova Scotia, Canada, a doctor of research career has focused on tick-borne
pharmacy degree from SUNY Buffalo in Buffalo, NY, and a master of diseases, infective endocarditis, bloodstream
science degree in biometry from the University of Texas School of infections, and hospital-acquired infections.
Public Health. Postdoctoral training includes a pharmacy practice resi- In 1998 he founded the Duke Infection Con-
dency at Bassett Healthcare, Cooperstown, NY, and infectious disease trol Outreach Network (DICON), which has

Downloaded from https://journals.asm.org/journal/cmr on 24 July 2024 by 2409:40f4:1:3c3a:456a:7f5a:e81e:35f9.

specialty residency and fellowship training at the University of Illinois at since grown to include a total of 68 hospitals in 6 southeastern states.
Chicago, Chicago, IL. He has also held a number of leadership positions in the Duke Depart-
ment of Medicine and the Infectious Diseases Society of America.

Mark E. Rupp, M.D., is a professor and chief

of the Division of Infectious Diseases at the
University of Nebraska Medical Center. He is
the medical director of The Nebraska Medi-
cal Center Department of Infection Control &
Epidemiology. He serves as the chief of staff
for Nebraska Medicine. Dr. Rupp received his
medical degree from Baylor College of Med-
icine and holds a B.S. degree in chemical
engineering from the University of Texas. He
underwent training in internal medicine and
infectious diseases at Virginia Commonwealth University. He is a fellow
of the American College of Physicians, the Infectious Diseases Society of
America, and the Society for Hospital Epidemiology of America (SHEA).
He is a past president of SHEA. Dr. Rupp has published over 450 articles,
chapters, and abstracts. He frequently serves as a guest lecturer and is an
active teacher and researcher. Dr. Rupp’s research interests are in the areas
of health care-associated infections and antimicrobial stewardship.

January 2020 Volume 33 Issue 1 e00009-19 cmr.asm.org 21