Bloodstream infection in children

Lucy Lum Chai See, MBBS, MD, MRCP

Objective: To establish the definitions of bloodstream infection
(BSI) in children for the purposes of identifying BSI for early
therapy, enrollment in sepsis trials, and epidemiology and sur-
veillance studies.
Methods: Generalized medical literature search using various com-
binations of the terms “bloodstream infection,” “children,” and “sepsis.”
Results: The medical literature is sparse on these topics;
therefore, these recommendations are adapted from guidelines
designed for adults. BSI overlaps with other areas of sepsis, such
as catheter-related BSI, which will be covered separately. This
discussion focuses on BSI of unknown origin, also known as
primary BSI.
Conclusion: A BSI is the presence of a pathogen in the blood.
Its clinical significance should be determined by the presence of
the host response as defined by the modified criteria for systemic
inflammatory response syndrome SIRS in children or a clinically
recognizable syndrome. Definitions of BSI for the purposes of
sepsis trials may differ from those for epidemiologic or surveil-
lance studies. (Pediatr Crit Care Med 2005; 6[Suppl.]:S42–S44)
KEY WORDS: bloodstream infection; sepsis

T here is a large amount of lit-
erature on epidemiology and
burden of disease that ad-
dresses bloodstream infection
(BSI) in adults. The same, however, can
not be said of children. The few published
articles that are available address almost
exclusively nosocomial BSI (1). Hence,
the following recommendations are
adapted from guidelines that are designed
for adults.
Gray et al. (2) published a 3-yr survey
of bacteremia and fungemia in a pediatric
intensive care unit. They observed an inci-
dence of 39.0 per 1,000 admissions, or 10.6
per 1000 bed days. Of these, 64.1% were
intensive care unit–acquired and 20.6%
were community acquired. The rest of the
infections (15.3%) were acquired in other
areas of the hospital. Crude mortality in
children with BSI was 26.5%, compared
with 8.1% in those without BSI. In adults,
BSI accounts for 30%– 40% of severe sepsis
and septic shock (3). However, the true
prevalence of BSI as a cause of severe sepsis
is probably underestimated in the hospital
setting. This is because blood cultures are
From the Department of Pediatrics, Faculty of
Medicine, University of Malaya, Kuala Lumpur, Malay-
sia.
This work was supported by the Mannion Family
Fund—Center for the Critically Ill Child, Division of
Critical Care Medicine at Children’s Hospital Boston,
the PALISI Network, and the ISF.
Copyright © 2005 by the Society of Critical Care
Medicine and the World Federation of Pediatric Inten-
sive and Critical Care Societies
DOI: 10.1097/01.PCC.0000161945.98871.52
frequently drawn from patients pretreated
with broad-spectrum antibiotics or are not
taken at all. In addition, smaller volumes
of blood are routinely taken in smaller
children, decreasing the sensitivity of the
cultures.
Definitions of BSI that have been used
by numerous investigators have been ob-
tained from studies done in the 1980s and
1990s (4 – 6). In recent years, the defini-
tions of nosocomial infections of the Cen-
ters for Disease Control and Prevention
(7) have been used in most epidemiologic
studies (2). Although widely used, they
are designed for nosocomial infections
and may need to be modified for commu-
nity-acquired infections. In the 1999 doc-
ument, BSI is divided into laboratory-
confirmed BSI and clinical sepsis for
which blood cultures are not performed,
not detected, or for which the physician
institutes treatment for sepsis. Although
it may be criticized that BSI should be
defined independently of a physician’s de-
cision whether to initiate antimicrobial
therapy, the real situation is that labora-
tory-based surveillance alone will under-
estimate the prevalence and the burden
of BSI. In addition, the Centers for Dis-
ease Control and Prevention guidelines
do not provide a definition for secondary
BSI, and they reference obsolete micro-
biological diagnostic tools such as anti-
gen testing in the blood.
Bacteremia is the presence of a recog-
nized pathogen in the blood. This pres-
ence may be indicated by a positive blood
culture, positive blood film (as in ma-
laria), or a positive serological response
in the presence of a compatible clinical
syndrome associated with a high proba-
bility of infection (as in enteric fevers,
leptospirosis, meningococcemia). This
definition, however, is not perfect. It ex-
cludes clinically well-recognized syn-
dromes that are associated with sepsis,
such as toxic shock syndrome, which is
caused by the exotoxin secreted by Staph-
ylococcus aureus. Frequently, BSI is
strongly suspected without microbiolog-
ical confirmation. Classification, how-
ever, usually depends on identification of
bacteremia.
Not all positive blood cultures, how-
ever, are true BSIs. In the study by Wein-
stein et al. (8), 41.5% of inpatient positive
blood culture episodes were judged to
represent contamination, and another
5.3% were of indeterminate clinical sig-
nificance. Thus, only half of all positive
cultures represented true BSI. These
rates might be explained by the use of
intravascular devices such as arterial or
central venous catheters for the purpose
of obtaining blood cultures. The rate of
contamination might be higher in chil-
dren, in whom obtaining blood cultures
from an indwelling device may be
thought to prevent the trauma of a veni-
puncture. Distinguishing between true
contamination and local infection of the
device (catheter colonization) is also dif-
ficult in children if peripheral blood cul-
tures are not obtained.
In critically ill adult patients in whom
paired blood culture specimens were ob-

S42 Pediatr Crit Care Med 2005 Vol. 6, No. 3 (Suppl.)

tained through a central venous catheter
and a peripheral venipuncture, the sensi-
tivity was 82.4% and 64.7%, respectively,
and specificity was 92.5% and 95.9% (9).
The positive predictive value was 58.3%
for catheter-obtained samples and 66.7%
for peripheral venipuncture samples, and
the respective negative predictive values
were 97.6% and 95.5%. Although the
negative predictive value of blood sam-
ples obtained by both techniques is good,
the sensitivity of blood samples obtained
by either catheter draw or peripheral ve-
nipuncture alone is not adequate to rec-
ommend the elimination of blood sam-
ples obtained from the other site. In
children, smaller volumes of blood are
usually drawn (ⱕ3 mL in children vs. 10
mL in adults), leading to a markedly
lower sensitivity. Peripheral blood sam-
ples may be valuable when interpreting
positive blood culture results for com-
mon skin or central venous catheter con-
taminants.
Given the caveats described above, the
following consensus definitions are pro-
posed for BSI in children after the new-
born period (separate definitions are be-
ing proposed for newborns).
Definite BSI
1. Microbiological confirmation of the
presence of recognized pathogens
that are not common to skin flora,
or
2. Isolation of organisms that are com-
mon skin flora together with clinical
signs and symptoms of infection,
from
a. Two or more separate blood cul-
tures,
b. One blood culture and another site,
or
c. One blood culture in a patient with
an intravascular device in whom
there is resolution of clinical signs
and symptoms after removal of the
device or after appropriate therapy.
Catheter-related BSI will be discussed
in more detail elsewhere. A definite BSI
can be further divided into:
1. Primary BSI: a BSI not related to an
identifiable focus of infection or an
intravascular catheter-related BSI.
2. Secondary BSI: A BSI caused by mi-
croorganisms related to an infection
at another site (e.g., pneumonia, in-
traabdominal abscess).
Pediatr Crit Care Med 2005 Vol. 6, No. 3 (Suppl.)
Probable BSI
Presence of systemic inflammatory re-
sponse syndrome (SIRS) or a clinically
compatible syndrome with a negative
blood culture plus a non– culture-positive
marker of inflammation, such as in-
creased C-reactive protein level or in-
creased procalcitonin level, plus a sero-
logical response to immunoglobulin M in
the acute phase of infection.
Possible BSI
Presence of SIRS or a clinically com-
patible syndrome, plus laboratory mark-
ers of inflammation, such as an increased
C-reactive protein level or an increased
procalcitonin level, but a negative blood
culture or a negative serological response
of a particular pathogen. The difference
between probable and possible BSI is that
indirect evidence of a pathogen (positive
serology) is present, which is absent in
possible BSI.
For the purpose of enrolling children
in sepsis trials, the diagnosis of BSI is
often a tentative diagnosis until definitive
culture results are available together with
a confirmed absence of infection from
other sources. To decrease the probability
of patients without BSI being enrolled
into clinical sepsis trials, those patients at
highest risk of BSI should be considered
eligible. A clinically significant BSI
should have a laboratory-confirmed pres-
ence of a pathogen and be accompanied
by a host response, which manifests itself
in physiologic disturbances, including a
nonspecific inflammatory process as de-
fined in SIRS.
The SIRS criteria were developed for
use in the adult population (10) and
therefore contained a number of clinical
signs and laboratory values not appropri-
ate for children. A number of modifica-
tions of these criteria for the pediatric
population have been proposed. The most
recent one was used in the Recombinant
Human Protein C study in children (11),
which was based on a variation of the
Bone’s Sepsis Syndrome Definitions (12).
These criteria were further refined by the
International Pediatric Sepsis Consensus
Conference (13). These modified criteria
for SIRS in children include the presence
of at least two of the following four cri-
teria, one of which must be abnormal
temperature or leukocyte count:
1. Core temperature of ⬎38.5°C or
⬍36°C
2. Tachycardia, as defined as a mean
heart rate of ⬎2 SD above normal for
age in the absence of external stimu-
lus, chronic drugs, or painful stimuli;
or otherwise unexplained persistent
elevation for a 0.5- to 4-hr time pe-
riod, or for children ⬍1 yr old, brady-
cardia, defined as a mean heart rate
less than the tenth percentile for age
in the absence of external vagal stim-
ulus, beta-blocker drugs, or congeni-
tal heart disease; or otherwise unex-
plained persistent depression for a
0.5-hr time period.
3. Mean respiratory rate ⬎2 SD above
normal for age or mechanical ventila-
tion for an acute process not related to
an underlying neuromuscular disease
or the administration of general anes-
thesia.
4. Leukocyte count elevated or depressed
for age (not secondary to chemotherapy-
induced leucopenia) or ⬎10% imma-
ture neutrophils.
For the evaluation of trials involving
BSI, a clinically significant BSI (i.e., a
laboratory-confirmed presence of a
pathogen plus the presence of the SIRS
criteria or a recognized clinical syndrome
compatible with an infection) should be
considered the gold standard. Although a
bacterial infection in the blood may often
be confirmed by positive blood cultures
or other methods, other pathogens, such
as infections of viral or rickettsial origins,
may not be positively confirmed. It
should be borne in mind that some of
these infections may directly affect the
leukocyte count in a way that the leuko-
cyte count in the SIRS criteria may not
be fulfilled.
For the purpose of identifying BSI for
epidemiologic and surveillance studies,
the Centers for Disease Control and Pre-
vention guidelines that are designed for
nosocomial infections may be applicable
for community-acquired infections with
some modifications.
1. The definition should be modified to
include infections that cannot be di-
agnosed by positive blood cultures but
are accompanied by a clinically recog-
nizable syndrome such as typhoid fe-
ver, leptospirosis, Epstein-Barr virus,
and rickettsial infections.
2. For clinical sepsis, the definition is
modified to include at least two of the
four SIRS criteria, and blood cultures
are negative or are not drawn, there is
no apparent infection at another site,
and physician institutes treatment for
sepsis.
S43
 A BSI is regarded as hospital-acquired
if the initial positive blood culture is ob-
tained ⬎48 hrs after admission to the
hospital. Likewise, BSI is considered to
have been acquired in the intensive care
unit when the initial positive culture is
obtained ⬎48 hrs after admission to the
unit. A community-acquired BSI is a di-
agnosis of exclusion of the above.
Epidemiologic surveillance of BSI
should include both laboratory and clin-
ical surveillance. Laboratory surveillance
will identify laboratory-confirmed BSI,
which by itself does not give any indica-
tion of its clinical significance and as to
whether it is a true or contaminated sam-
ple. We recommend that for epidemio-
logic surveillance, BSIs should be cap-
tured as separate entities of:
1. Laboratory-confirmed BSI with no
clinical significance.
2. Laboratory-confirmed BSI with clini-
cal significance.
3. Clinical sepsis.
Clinical significance is identified by
the presence of two or more of the four
criteria of pediatric SIRS. The epidemiol-
ogy, morbidity, and economic costs of
each of these entities may differ signifi-
cantly (14).
For the purpose of identifying BSI for
diagnosis and optimal therapy, we recom-
mend early recognition for the presence
of BSI and early institution of therapy.
Before initiation of antibiotic therapy, the
sample of blood drawn for culture should
be as large of a volume as possible, and
the laboratory confirmation of pathogens
plus adequate cultures from other sus-
pected sites of infection, such as central
venous catheters, surgical wounds or
urine, should be done. Apart from the
S44
recognition of the pediatric SIRS criteria,
clinicians must be equipped with suffi-
cient knowledge of clinical syndromes
that have high association with infec-
tions. A detailed history may identify risk
factors—for example, swimming in rivers
(leptospirosis), playing in muddy fields
(Chromobacterium violaceum), visiting
endemic areas of diseases (malaria, ty-
phoid), consuming food from food handlers
(typhoid)—and will guide appropriate anti-
biotic therapy. Clinical examination may
reveal petechial or purpuric rash or pur-
pura fulminans and other organ involve-
ment that are clinically recognizable. A
comprehensive knowledge of the epidemi-
ology of nosocomial infections in the insti-
tution or the ward will be necessary to
recognize and control outbreaks of nosoco-
mial infections. Careful consideration of
the choice and duration of antimicrobial
therapy is necessary to prevent the emer-
gence of multiresistant organisms in the
institution.
REFERENCES
1. Slonim AD, Kurtines HC, Sprague BM, et al:
The cost associated with nosocomial blood
stream infections in the pediatric intensive
care unit. Pediatr Crit Care Med 2001;
2:170 –174
2. Gray J, Gossain S, Morris K: Three-year sur-
vey of bacteremia and fungemia in a pediatric
intensive care unit. Pediatr Infect Dis J 2001;
20:416 – 421
3. Bochud PY, Glauser M, Calandra T: Antibiot-
ics in sepsis. Intensive Care Med 2001; 27:
S33–S48
4. Weinstein MP, Murphy JR, Reller LB, et al.
The clinical significance of positive blood
cultures: A comprehensive analysis of 500
episodes of bacteremia and fungemia in
adults. II: Clinical observations, with special
Pediatr Crit Care Med 2005 Vol. 6, No. 3 (Suppl.)
reference to factors influencing prognosis.
Rev Infect Dis 1983; 5:54 –70
5. Bates DW, Cook EF, Goldman L, et al: Pre-
dicting bacteremia in hospitalized patients: A
prospectively validated model. Ann Intern
Med 1990; 113:495–500
6. Pittet D, Wenzel RP: Nosocomial blood-
stream infections: Secular trends in rates,
mortality, and contribution to total hospital
deaths. Arch Intern Med 1995; 155:
1177–1184
7. National Nosocomial Infection Surveillance.
Atlanta, Center for Disease Control and Pre-
vention, 1999
8. Weinstein MP, Towns ML, Quartey SM, et al:
The clinical significance of positive blood
cultures in the 1990s: A prospective compre-
hensive evaluation of the microbiology, epi-
demiology, and outcome of bacteremia and
fungemia in adults. Clin Infect Dis 1997;
24:584 – 602
9. Beutz M, Sherman G, Mayfield J, et al: Clin-
ical utility of blood cultures drawn from cen-
tral venous catheters and peripheral veni-
puncture in critically ill medical patients.
Chest 2003; 123:854 – 861
10. Levy M, Fink M, Marshall J, et al: 2001
SCCM/ESICM/ACCP/ATS/SIS International
Sepsis Definitions Conference. Crit Care Med
2003; 31:1250 –1256
11. Barton P, Kalil AC, Nadel S, et al: Safety,
pharmacokinetics, and pharmacodynamics
of drotrecogin alfa (activated) in children
with severe sepsis. Pediatrics 2004; 113:7–17
12. Samson LM, Allen UD, Creery WD, et al:
Elevated interleukin-1 receptor antagonist
levels in pediatric sepsis syndrome. J Pediatr
1997; 131:587–591
13. Goldstein B, Giroir B, Randolph A: Interna-
tional Pediatric Sepsis Consensus Confer-
ence: Definitions for sepsis and organ dys-
function in pediatrics. Pediatr Crit Care Med
2005; 6:2– 8
14. Hugonnet S, Sax H, Eggimann P, et al: Nos-
ocomial blood stream infection and clinical
sepsis. Emerg Infect Dis 2004; 10:76 – 81