++
Intravascular catheters are used
for a wide range of adjunctive therapies in pediatric patients,
such as administering total parenteral nutrition and chemotherapy
and facilitating blood drawing. For the purposes of discussing complication risks
and preventive strategies, these catheters can be subdivided into
short-term, intermediate-term, and long-term devices. Approaches
to catheter device placement and care that prevent infection are
discussed in Chapters 34 and 107.
++
The pathogenesis of bloodstream infections for both short-, intermediate-,
and long-term devices includes migration of potential pathogens from
the skin at the exit site along the external surface of the catheter
to the catheter tip, intraluminal migration of organisms from the
catheter hub, contaminated infusates, and rarely, seeding of the
catheter hematogenously from a distant focus.1 The
true incidence of bowel translocation of microorganisms with subsequent
seeding of the catheter is unknown, but this is proposed as a potential
mechanism of catheter-related bloodstream infections (CR-BSI) in
patients with dysfunctional bowel. Short-term and intermediate-term,
noncuffed, nontunneled catheters are more prone to migration of
organisms along the external surface of the catheter, and as a result they
are infected by a greater proportion of skin flora including coagulase
negative staphylococci and Staphylococcus aureus.2-5 Long-term, cuffed,
tunneled catheters are less likely to be infected by organisms along
the external catheter surface because the cuff acts as a fibrotic
dam to migration. The definitions of different types of catheter-related
infections are presented in Table 239-1.
++
Risk factors for catheter-associated bloodstream infections include
prolonged use of systemic antimicrobials and the selective pressure for
resistant organisms that results, catheter location (short- and
intermediate-term lines placed in the femoral vein were more prone
to infection in some pediatric studies),6-9 infusion
of hyperalimentation with lipids, prolonged duration of catheterization,
age < 2 years (especially premature infants with immature skin
integrity), burn patients, immunocompromised patients, and those
with intestinal integrity issues.1,10
+++
Prevention of Catheter-Related Infections
++
Maximal barrier precautions during insertion, including the use
of sterile masks, gowns, gloves, and large sterile fields, have
been shown to significantly reduce the incidence of catheter infections.11-14 Antisepsis
with chlorhexadine should be used before insertion of a catheter.15,16 Prior to
manipulation of the catheter, hand hygiene and hub antisepsis are
mandatory to reduce the incidence of catheter contamination. Sterile gauze,
a transparent dressing,17 or a chlorhexadine patch18can
be used to cover the site. Preference is usually given to the transparent
dressing either coated with or without chlorhexadine because it
is easier to evaluate the exit site. Dressings should only be changed
when they are loose, soiled, or damp.1 The exit
site should be cleansed with chlorhexadine in 70% alcohol,
although povidone iodine is still used.19-22 Detailed
sample management programs for the care of central venous catheters
and peripherally inserted central catheters (PICC lines) are provided
on the DVD (Appendix 239-2). To reduce infections,
these catheters have been coated with either external chlorhexadine-silver sulfadiazine23 or
internal and external minocycline/rifampin.24,25 Both
types of catheters have proven effective in reducing the incidence
of catheter-related infections in prospective randomized trials
in adults when compared to catheters without antibiotic coating,
but a recent comparison has found the minocycline/rifampin catheter
superior in this respect.24 However, second-generation
antiseptic-coated catheters may be more effective in this regard.26-28 A
recent meta-analysis found little data indicating that these catheters
can be used for durations greater than 2 weeks.29 They
have not been extensively studied in children.
++
The care of long-term central venous catheters is detailed in
Chapter 39 and on the DVD (Appendix 239-1).
Recent work has shown that vancomycin instilled into the catheter
lumen and allowed to dwell for a period of time (antibiotic lock)
is effective in reducing the number of Gram-positive infections
in these long-term lines.30 The largest prospective study
performed in a pediatric population evaluated the use of ciprofloxacin/vancomycin
in a similar manner to reduce Gram-negative infections but failed
to show any further reduction in infections when compared to vancomycin
alone.31
+++
Diagnosis of Catheter-Related Infection
++
Our ability to accurately diagnose these infections has evolved
over time but remains imperfect. Typically, blood cultures are drawn
from patients with intravascular catheters if they develop a fever.
In the past nontunneled lines were removed immediately
if a blood culture became positive. The line tip was then cultured, and
if there were ⩾ 15 colonies of the same organism on the catheter
tip as in the blood, a line-associated infection was diagnosed.2 This method
is accurate, but it requires the removal of the catheter, and with
the advent of tunneled and totally implanted catheters, a diagnostic strategy
that did not require removal of the catheter was sought. Initially,
blood cultures were drawn from the catheter and a peripheral vein.
If both cultures grew the same organism, a catheter-associated bloodstream
infection was diagnosed, and an attempt to clear the line with appropriate
antimicrobials was made as long as the patient remained clinically
stable.Recently, experimental evidence has revealed that it is possible
to differentiate between a line-associated infection and bacteremia
from another source by utilizing equal-volume quantitative blood
culture methods.32-37 If the infection is truly
line related, then the largest burden of organisms would be at the catheter
tip. As organisms are shed into the bloodstream, they are filtered
through the lungs, and as a result, they have a lower concentration
in the peripheral blood. Studies have shown that there should be
5 to 10 times the number of colonies from a blood culture drawn through
the catheter than that drawn from a peripheral vein in order to
diagnose a line-related infection.38 This method
is sensitive and specific, but it is not widely used because of
its labor intensiveness and expense. It may have its greatest utility
in specific patient populations, such as those with bowel dysfunction,
where one might expect a greater incidence of bacterial translocation
to take place.A new method has been proposed. Newer blood culture
analyzers are capable of storing time of entry and time of positivity
in the computer; therefore, it is possible to generate standard
curves with known quantities of specified organisms. In this way
an approximate organism burden can be generated once the machine
has categorized a bottle as positive.39-44 This
method could become the gold standard after further validation because
it is less labor intensive and cheaper than the current quantitative
methods.
++
Controversy remains as to the optimal combination of central
and of peripheral blood cultures to obtain in pediatric patients.
Studies looking at the sensitivity, specificity, and positive (PPV) and
negative (NPV) predictive values of central versus peripheral blood
cultures have been predominantly performed in adults.45-47 The
simultaneous acquisition of equal volumes of blood from a peripheral
vein and from a potentially infected catheter provides the best
combination of sensitivity, specificity, PPV, and NPV; however,
central cultures have greater sensitivity while peripheral cultures
have greater specificity and PPV. In an adult patient, where adequate
blood volume is generally assured, the use of multiple peripheral
blood cultures to diagnose a catheter-related bloodstream infection
might be sufficient. The greater sensitivity of the central culture
would be expected to occur as a result of colonization or contamination,
while the greater specificity and PPV as well as the equivalent
NPV of the peripheral cultures would lead to a more accurate diagnosis
of a catheter-related bacteremia.In children, especially infants
and toddlers, equal blood volume is not always guaranteed. As a
result, because volume of blood has been shown to predict culture
positivity,48 inadequate peripheral cultures could
lead to a high false-negative rate compared to a central culture,
where blood volume is not as important an issue. Consequently, it
has become more common in children to draw blood cultures only from
the catheter. The optimal combination of central and peripheral
blood cultures must take into consideration the age of the child, the
type of catheter, the number of catheter lumens and ability to draw
from each lumen, the clinical status of the patient, and the type
of blood culture vial (pediatric blood culture vials have an optimal
liquid medium–to–blood ratio for smaller volumes
of blood). Every effort should be made to obtain an adequate peripheral
culture before antimicrobials are initiated as per recommendations. As
always, clinical judgment is paramount in interpreting results and
in deciding on the appropriate management strategy.
++
Short- or immediate-term catheters should ideally be removed
when a catheter-associated infection is documented.49 A
great deal of attention has been paid to the duration of catheterization and
its role in the pathogenesis of catheter-related infections. It
has been shown that the colonization rate of these catheters increases
with the duration of catheterization, but it has also been shown
that the rate of infection does not change over time while the absolute
number of infections increases.50 In clinical studies
as well, there has been no clear reduction in infectious complications
if one compares scheduled periodic changing of short-term catheters via
guidewire exchange versus removal and replacement of these devices
as needed.51 Removal of a long-term catheter may
not be practical in some clinical situations since this risks the loss
of life-saving venous access. It is now routine practice to attempt
to treat an infection in a long-term catheter while the catheter
is in place. Guidance on decisions for immediate versus delayed removal
of a catheter are provided in Table 239-2.
A proposed algorithm49 for the management of short-
and intermediate-term catheters suggests: (1) removal of catheters
when they are no longer needed for patient care and (2) replacement
over a guidewire if clinically indicated, ie, mechanical difficulty
or suspected but as yet undocumented infection. Replacement by guidewire
exchange has been shown to reduce insertional complications. If
a line is exchanged because of presumed infection, the old line
should be cultured, and if positive, the new line should be removed
and another new catheter placed in a different location. If a catheter
is removed because of a documented infection, then a new line should
be placed in a different location. An alternative approach would be
removal of all malfunctioning lines with reinsertion at a different
location.
++
++
Decisions regarding whether to remove a long-term line secondary
to recrudescence of infection is more problematic and should be decided
by the pathogenicity of the offending organism, the status of the
patient, and the need to maintain appropriate access; however, most
experts would recommend line removal in such situations.
++
The choice of empiric antibiotics depends upon the host, previous
history of bloodstream infection, and nosocomial versus community acquired
pathogens. There are a large variety of acceptable regimens chosen
by the likely organisms and patient status, as shown in eTable 239.1. In a neutropenic or unstable
patient, combination Gram-negative antibiotic coverage should be
initiated with agents that work by different mechanisms (ie, a pencillin derivative
and aminoglycoside or a quinolone and aminoglycoside) until the
patient has stabilized and an organism is identified.
++
++
Duration of antibiotic therapy depends on the clinical status
of the patient and the laboratory findings. If the catheter culture
is positive and the peripheral culture is negative, then a short
course (3–5 days for coagulase negative staphylococci and
7 days for most other bacterial pathogens) is acceptable as long
as the host does not have an underlying immunodeficiency. If the
patient is also bacteremic, then the duration of antimicrobials
depends on the ability to clear the organism and may prolong therapy
by at least a week. Repeat cultures should be drawn before starting therapy
if a coagulase negative staphylococcus is isolated because of the
possibility of contamination from a skin site; however, this might
not always be practical in patients with an underlying immunodeficiency
when empiric therapy is initiated quickly. A workup for bacterial
dissemination to other sites (spontaneous bacterial endocarditis,
bone) be initiated for all patients who fail to clear their peripheral
blood culture despite line removal or who remain persistently symptomatic
despite negative blood cultures. If this workup is negative at the
end of a standard 2-week course of antimicrobial therapy, then it
is reasonable to discontinue treatment. In addition, bloodstream
infection with Candida (spp) should prompt a search
for additional foci of infection, especially in neutropenic patients. It
should include an ophthalmology exam, echocardiogram, and CT scan
of chest and abdomen (head if appropriate shunting might occur across
the heart).
++
Granulocytopenic oncology patients without positive cultures
can be started on dual therapy with pipericillin/tazobactam
or cefipime and an aminoglycoside; however, if local epidemiology
is consistent with increased isolation of methicillin-resistant S
aureus, then vanocmycin should be added to empiric therapy
until culture results are known. Possible community-acquired infections
may be covered with a combination of oxacillin and an aminoglycoside
or third-generation cephalosporin. Nosocomial infections are usually covered
with vancomycin and pipericillin/tazobactam or cefipime.
Once a pathogen is identified, antibiotics can be narrowed, and special
attention should be given to removing vancomycin as soon as possible.
A workup for bacterial dissemination to other sites (spontaneous
bacterial endocarditis, bone) should be initiated for all patients
who fail to clear their peripheral blood culture despite line removal or
who remain persistently symptomatic despite negative blood cultures.
If this workup is negative at the end of a standard 2-week course
of antimicrobial therapy, then it is reasonable to discontinue treatment.
In addition, bloodstream infection with Candida spp. should
prompt a search for additional foci of infection, especially in
neutropenic patients. It should include an ophthalmology exam, echocardiogram,
and CT scan of chest and abdomen (head if appropriate shunting might
occur across the heart).
+++
Adjunctive Treatment
Strategies
++
In addition to the use of antibiotic locks for the prevention
of catheter-related bloodstream infections, antibiotic and ethanol
locks have been used as an adjunct in the treatment of catheter-related
infections. Antibiotic lock therapy has been recommended in the
Infectious Diseases Society of America guidelines for the treatment of
catheter infections,49 and ethanol lock therapy is
a promising new technique for the treatment of such infections.53,54 The
major complication of ethanol lock therapy is catheter occlusion,
presumably from biofilm destruction. This results in the use of
catheter declotting procedures but rarely requires catheter removal
(see DVD: Appendix 239-3).
+++
Treatment of
Exit/Tunnel Infections
++
The treatment of more localized exit/tunnel infections
depends on the catheter in use.49 For short-term
catheters, an exit site infection associated with bacteremia should
prompt catheter removal and the introduction of systemic as well
as topical antimicrobials. If the exit site infection is not associated
with bacteremia, then an attempt can be made to treat the infection
with topical agents alone. The treatment of long-term catheter exit
site infections is essentially the same, except that unlike the
short-term catheter, long-term catheters can be salvaged, even in
the face of bacteremia, if the infection responds to appropriate
antimicrobial therapy with sterilization of the blood culture and
resolution of the local infection. Tunnel tract infections are generally
seen with long-term catheters only. They require systemic antimicrobial
therapy and usually necessitate catheter removal. Unless bacteremia
is present, a trial of empiric therapy for S aureus and Pseudomonas
aeruginosa is warranted if the patient is stable. If no
improvement is noted, then the catheter should be removed and the
tract cultured. Antibiotics can then be tailored to the identified
pathogen. Pocket infections require catheter removal for eradication
of the offending organism.
++
Major complications of central line infections include sepsis,
septic emboli, endocarditis, endovascular infection, abscess formation, and
death. The true incidence of these complications is unknown in the
pediatric population, but prompt removal of the catheter and initiation
of long-term antibiotic therapy are required to ensure a good outcome.
Certain organisms, including Candida (spp) Pseudomonas (spp)
and S aureus, have greater metastatic potential;
thus, early consideration should be given to removing central lines
when these pathogens are present. Studies analyzing the outcomes
of central line–related infections have shown increased
length of stay with associated increases in medical costs, and morbidity
and mortality rates as high as 10–20%.55-57 Complications
in immunocompromised patients would be expected to be greater and
to carry a larger morbidity and mortality burden. However, there
is a need for studies of this kind in pediatric patients, with special
attention given not only to the differences in organisms, but to
the outcomes of the different catheter types.