• Summary
  Click here to print a Quick Reference Guide
• Background
• Clinical Diagnosis
• Investigation
• Treatment
• Non-Antimicrobial Treatment
• Empirical Antimicrobial Treatment
• Directed Antimicrobial Treatment (when microbiology results are known)
• Duration of Treatment
• Switch to oral agent(s)
• Treatment Failure

Key Links

• Table 1 . Empirical antimicrobial treatment of Catheter-related infection in adults – empirical and most common Gram positive pathogens.
• Table 2. Antimicrobial treatment of Catheter-related infection in adults – most common Gram positive pathogens.
• Table 3. Antimicrobial treatment of uncomplicated CRI in adults according to the most common Gram-negative pathogens.
• Table 4. Antimicrobial treatment of uncomplicated CRI caused by fungi.

Diagnosis

• This guideline applies to patients with temporary central venous catheters (CVCs) and/or arterial catheters (AC).
• Monitor for exit site infection using VCVC score
• Diagnose exit site infection when pain + erythema or pus at exit site.
• Consider intravascular catheter-related bloodstream infection (CRBSI) in any patient with a CVC/ AC and systemic symptoms or signs of infection and/or unexplained bacteraemia/fungaemia.
• CRBSI can occur with exit site infections.

Investigation

• Swab inflamed or purulent exit sites for culture.
• Send Blood Cultures from all potentially infected lumens and a peripheral vein for “DTP”– according to LTHT guidelines.

Non-antimicrobial management

• Remove infected CVCs or arterial lines whenever possible. (This should be discussed with a registrar or consultant prior to removal – see full guideline.)
• Attempting salvage of infected CVC or Acs is not generally recommended.
• Where multiple vascular access devices are in situ and CRBSI is diagnosed/suspected a full line change/removal should ideally be undertaken.
• “re-wiring” of new catheters should be avoided.

Empirical antimicrobial management

1. Exit site infection with no signs of bloodstream infection:
   Wait for swab results if possible, otherwise Flucloxacillin 500mg 6-hourly orally. If MRSA colonised/infected discuss with microbiology.
2. Suspected CRBSI but not severe sepsis/septic shock:
   Teicoplanin IV (see dosing guidelines. Off-label use and the patient should be informed of this)
3. Suspected CRBSI severe sepsis/septic shock:
   Teicoplanin IV (see dosing guidelines. Off-label use and the patient should be informed of this)
   PLUS either:
   Ceftazidime  1g 8-hourly iv (patient <65 years and not recently treated with Ceftazidime ) OR,
   Piperacillin/tazobactam  4.5g 8-hourly iv (patient >65 years and not recently treated with Piperacillin/tazobactam )
   If in doubt contact microbiology

Directed (organism-specific) antimicrobial management
See table

Back to top

Different types of vascular access device have been developed with different purposes in mind. The term “temporary CVC” is used in this guideline to refer to any CVC which does not have a portion of its length “tunneled” subcutaneously before it exits the skin and does not have a subcutaneous “cuff”.

Temporary CVCs may have multiple lumens and are intended for short term use (usually <14 days). This guideline concerns infection affecting both temporary CVCs and arterial catheters because the principles are the same.

Infection is the major complication associated with the insertion and maintenance of both temporary CVCs and arterial lines. CVCs are most commonly affected than arterial lines (Fletcher, 1999). Infections associated with CVCs can be divided clinically into local (exit site infections) and bloodstream infections, but these entities may coexist. The incidence of intravascular catheter related bloodstream infections (CRBSI) ranges widely between institutions and patient groups.

Intravascular catheter-related infections continue to be associated with mortality despite the availability of effective broad-spectrum antimicrobial agents (Fletcher, 1999). In addition, secondary infections such as endocarditis, vertebral osteomyelitis and septic thrombophlebitis are serious complications.

Micro-organisms can gain access to the lumen of a catheter via contamination of the hub or via a contaminated infusate (the latter being very rare). Hub contamination most likely occurs during handling to connect and disconnect infusions. Organisms may also be direct transferred from the lumen of one catheter to the lumen of another during a rewiring procedure (Olson et al., 1992).

The external surface of a catheter may become colonised as a result of contamination during insertion, via exit site colonisation/infection or secondary to a bacteraemia from a distant source (Elliott et al., 1997; Maki et al., 1997). It is axiomatic that a catheter must be colonised before invasion of tissues or bloodstream can occur. The longer a catheter remains in place, the more likely it is to become colonized (Sandoe et al., 2003).

Coagulase-negative staphylococci, particularly Staphylococcus epidermidis, are the most common cause of CRI, but Staphylococcus aureus, Candida species and enterococci are also commonly implicated (Pearson, 1996). Other biofilm forming Gram-negative organisms such as Pseudomonas aeruginosa and Stenotrophomonas maltophilia can also colonize catheters and subsequently cause infection. Enterobacteriaceae (such as Escherichia coli, Citrobacter, Enterobacter and Serratia species) are not infrequently encountered and may be more common in associated with femoral CVCs.

This guideline should be used in conjunction with the LTHT Infection Control Policy; “Guideline for the Prevention of Infection Associated with Central Venous Catheters (CVCs)” . The use of catheters designed and marketed to carry a lower risk of CRI is an issue that needs to be resolved after an evaluation of the evidence and of the particular risks in various clinical areas. This guideline does not assess such prophylactic measures, just the management of suspected infection.

All intravascular catheters should be suitable for the intended purpose and the use limited to that intended purpose in the interests of minimizing CRI.

Back to top

CRI affecting CVCs and arterial lines manifests as an “exit site” infection or bloodstream infection. These clinical entities may coexist and any of them may be accompanied by an isolated fever or signs of systemic inflammatory response syndrome (SIRS) or septic shock or severe sepsis (Bone et al., 1992) often without other localizing symptoms or signs.

Exit site infections are not difficult to diagnose clinically and manifest with erythema and pain at the exit site (>1cm) and/or a purulent exudate.

The exit sites of all CVCs and arterial lines should be observed at least once daily, the appearance documented according to the scoring tool (see appendix) and actions carried out according to the tool.
[Evidence level D]

Clinical diagnosis of CRBSI is notoriously unreliable (Pearson, 1996). Early work in patients receiving total parenteral nutrition determined that only 15% of intravascular catheters, which were removed because of suspected infection, and the tip cultured, were actually shown to be the cause of CRBSI (Ryan et al., 1974).

Intravascular catheter-related bloodstream infection (CRBSI) should be considered in any patient with a CVC and/or arterial line and systemic symptoms or signs of infection or bacteraemia/fungaemia without an alternative explanation. [Evidence level C]

Patients with a documented fungaemia and CVCs and/or arterial lines in situ should have all devices removed or changed as part of their treatment.
[Evidence level B]

Simple investigations can now establish a diagnosis of CRBSI with a high degree of confidence, without removing the catheter – see below.

Back to top

Exit site infection affecting any type of vascular access device is a clinical diagnosis. The pathogen responsible can be determined by sending a pus sample or exit site swab to the microbiology laboratory for culture.
[Evidence level B].

Differential time to positivity (DTP) is a blood culture-based method for diagnosing luminal colonisation or CRBSI (Blot et al., 1999). If a known volume of blood is collected via the lumen of an intravascular catheter and the same volume of blood is then drawn from a peripheral vein, the time taken for each blood culture set to become positive can be used to determine whether the sampled catheter is the likely source of infection. A DTP >2 hours in favour of “through-line” Blood Cultures has a sensitivity of 72-94%, and specificity of 91-95% for CRBSI (Blot et al., 1999; Catton et al., 2005). In other words, using DTP >2 hours will fail to detect some CRBSI, but a positive result has a high probability of being CRBSI. A negative “through-line” blood culture at >100h incubation had a negative predictive value of 100% (Blot et al., 1999; Catton et al., 2005).
[Evidence level B]

Paired Blood Cultures should be sent from all potentially infected lumens and a peripheral vein (Catton et al., 2005).
[Evidence level B]

Acridine orange leukocyte cytospin, quantitative Blood Cultures and endoluminal brush techniques are sensitive and specific methods of in situ diagnosis of intravascular catheter colonisation but are labour intensive, expensive and not routinely available in Leeds.
[Evidence level B]

Semi-quantitative catheter tip culture (often called the “Maki roll”) is a simple and inexpensive method for retrospectively diagnosing colonisation of the external surface of an intravascular catheter. It lacks sensitivity because only the external surface is cultured (missing luminal colonisation) and lacks specificity because the surface of the catheter may become contaminated with organisms colonizing the exit site. Intravascular catheter tips should only be sent for culture to confirm a clinical suspicion of CRI, not as a matter of routine.
[Evidence level B].

There is no evidence that surveillance Blood Cultures of intravascular catheters are of benefit and this service is not therefore recommended by the microbiology department.
[Evidence level D]

Back to top

Antimicrobial therapy
Antibiotic therapy for catheter-related infections may be initiated empirically, when the patient is unwell and it is considered unsafe to wait for the results of microbiological tests. For example, if a patient has severe sepsis or septic shock.
[Evidence level D]

The initial choice of antibiotics will depend on the likely pathogen(s) involved.
[Evidence level C]

Empirical therapy should be modified when the causative organism is known i.e to directed or targeted therapy.
[Evidence level C]

Antimicrobial allergy
contact microbiology for advice.

Back to top

Management of CRI is influenced by the type of catheter, severity of symptoms, underlying disease(s), availability of alternative vascular access sites, the pathogen(s) involved and the location of the infection (Bouza et al., 2002; Mermel et al., 2001).

The options for treatment for CRI are:

• removal of the catheter together with antimicrobial therapy.
• removal of the catheter alone.
• antimicrobial therapy without removing the catheter ( “catheter salvage”).

Removal of the intravascular catheter is the most reliable means of eliminating infection usually in combination with antimicrobial therapy. Attempting salvage of an infected temporary CVC or arterial line is not generally recommended. [Evidence level B]

Management decisions are made either empirically on the basis of clinical suspicion, without any supporting microbiological evidence, or when microbiological information is available.
[Evidence level D]

The decision to remove an intravascular catheter as a matter of urgency is a clinical one, dependent upon the severity of sepsis and the likelihood of CRI. Removal of a CVC should be discussed with the registrar or consultant prior to removal. Indications for urgent catheter removal are summarised below: (Krishnasami et al., 2002; Mermel et al., 2001)

• Bacteraemia or sepsis persisting >48-72h after commencing therapy.
• Presence of local complications (e.g. exit site infections).
• Micro-organisms known to be difficult to eradicate (e.g. S. aureus, Bacillus spp., Corynebacterium spp., mycobacteria, Stenotrophomonas maltophilia and fungi).
• Presence of metastatic complications (e.g. infective endocarditis, infected pulmonary embolism).

Catheter removal alone may be sufficient for intravascular catheters colonised by low-grade pathogens, such as coagulase-negative staphylococci, in low-risk patients.
[Evidence level C] NB. “Low risk” in this setting means the patient does not have any intravascular prosthetic material (pacing system, prosthetic heart valves, previous endocarditis, structural cardiac defect known to predispose to endocarditis, Dacron/PTFE vascular grafts) and is not immunosuppressed (neutropaenic). (Bouza et al., 2002; Mermel et al., 2001).

Exit site infection, once established, is difficult to eradicate with antimicrobial therapy. Temporary CVCs and peripheral arterial catheters should be removed as soon as possible after a diagnosis of exit site infection has been made and antimicrobial therapy started (see below).
[Evidence level D]

Where intravascular catheters are still required they should ideally be re-sited away from the area of infection.
[Evidence level D]

CRBSI is a microbiological diagnosis. Given the poor predictive value of clinical diagnosis of CRBSI and the risks of changing intravascular catheters, individual management of suspected bloodstream infection should be based on a risk assessment. This decision needs to take into consideration that tests to assist in the in situ diagnosis of CRBSI (i.e without removing the catheter) are currently culture-based and take at least 24 hours to give a result (see above). The necessary action therefore depends on the severity of the patient’s condition and the likelihood of the problem being CRI.
[Evidence level D]

Initially the need for vascular access should be reviewed and catheters that are no longer needed should be removed. This is based on the fact that removing the portal of entry removes the risk of infection and the fact that catheter removal is the most reliable way of managing infection associated with intravascular catheters (Mermel et al., 2001; Sandoe et al., 2002).
[Evidence level C]

Where multiple vascular access devices are in situ and CRBSI is diagnosed full line change should ideally be undertaken because of the significant risk that all catheters in situ at the time of infection have been affected.
[Evidence level C]

Re-wiring catheters should be avoided if at all possible because infection is simply transferred from one catheter to the next via the guide wire (Olson et al., 1992).
[Evidence level B]

Guidelines for insertion and care of new catheters should be followed
[Evidence level D]

The possibility of a contemporaneous primary source of infection distant from the intravascular catheter should be considered, particularly in Staphylococcus aureus CRI and when fever or bacteraemia persist >72 hours after catheter removal (Raad & Sabbagh, 1992; Raad et al., 1992).
[Evidence level C]

Back to top

Gram-positive organisms, mainly staphylococci, are the most common cause of CVC/AC infections. Considering the high prevalence of meticillin-resistance in staphylococci, empirical antimicrobial therapy should include the administration of Teicoplanin . The prevalence of Gram-negative bacteria causing CRI is variable but no clinical features can reliably rule-out this possibility so empirical regimens for CRBSI when severe sepsis/septic shock is present should include Gram negative cover. All treatment recommendations are Evidence level D.

The choice of Gram-negative cover will be influenced by the age of the patient, local epidemiology on a given unit, recent antimicrobial therapy and comorbidities. If the options below are considered inappropriate the case should be discussed with the on-call or unit-specific microbiologist, other regimens may be entirely appropriate.

Recommended empirical regimens:

1. Exit site infection with no signs of bloodstream infection:
   Wait for swab results if possible, otherwise Flucloxacillin 500mg 6-hourly orally. If MRSA colonised/infected discuss with microbiology.
2. Suspected CRBSI but not severe sepsis/septic shock:
   Teicoplanin IV (see dosing guidelines. Off-label use and the patient should be informed of this)
3. Suspected CRBSI severe sepsis/septic shock:
   Teicoplanin IV (see dosing guidelines. Off-label use and the patient should be informed of this)
   PLUS either:
   Ceftazidime  1g 8-hourly iv (patient <65 years and not recently treated with Ceftazidime ) OR,
   Piperacillin/tazobactam  4.5g 8-hourly iv (patient >65 years and not recently treated with Piperacillin/tazobactam )

Table 1 . Empirical antimicrobial treatment of Catheter-related infection in adults – empirical and most common Gram positive pathogens.

 ^# Use of Teicoplanin for this use is off-label and the patient should be informed of this

Back to top

Table 2. Antimicrobial treatment of Catheter-related infection in adults – most common Gram positive pathogens.

Pathogen	Treatment with intravascular catheter removal			Attempted intravascular catheter salvage
	Antimicrobial Agent	Dose/route	Duration	Antimicrobial Agent[s]	Dose/route	Duration
Staphylococcus aureus Meticillin-susceptible Meticillin-resistant	Flucloxacillin	1g 6h IV	14 days	Not recommended
	Teicoplanin #	see dosing guidelines	14 days	Not recommended
Coagulase-negative staphylococci Meticillin-susceptible Meticillin-resistant	Line removal sufficient in most situations			Not recommended
	Line removal sufficient in most situations			Not recommended
Coryneforms [diphtheroids]	Line removal sufficient in most situations			Not recommended
Oral streptococci [penicillin-susceptible]	Penicillin	1.2g 6 h/V	2-7days	Not recommended
Enterococcus species Amoxicillin-susceptible Amoxicillin-resistant, Vancomycin susceptible.	Amoxicillin	1g 6h IV	5 days	Not recommended
	Teicoplanin #	see dosing guidelines	5 days	Not recommended
Other Gram positives (e.g. mycobacteria)	Consult microbiology

^# Use of teicoplanin for this use is off-label and the patient should be informed of this

Table 3. Antimicrobial treatment of uncomplicated CRI in adults according to the most common Gram-negative pathogens.

Table 4. Antimicrobial treatment of uncomplicated CRI caused by fungi.

*from removal date

Back to top

See table

Back to top

Once an infected catheter has been removed, if a patient has made a good clinically recovery at 48 hours antimicrobials can be switch to oral alternatives. (exception Staphylococcus aureus)

Back to top

Back to top