CONTENTS

• Rapid Reference 🚀
• Risk factors & prevention
• Clinical features
• Diagnosis
• Diagnostic criteria
• Management
  • Overall strategy
  • Empiric antibiotic selection
  • Treatment duration
  • Persistent bacteremia
• Specific situations
  • Septic thrombophlebitis
  • Arterial line infection
  • Chest port infection
• Other topics
  • Physiology
  • Prevention
• Podcast
• Questions & discussion
• Pitfalls

---

rapid reference

(back to contents)

---

---

risk factors & prevention

(back to contents)

---

---

risk factors for central line infection include:

• Immunocompromise (e.g., neutropenia).
• Body mass index >40. (37733310)
• Non-subclavian insertion site.
• Multiple manipulations of the catheter (e.g., repositioning).
• Hemodialysis catheter.
• Duration of catheter use.

---

prevention of central line infection

[1] avoidance of a central line

• Peripheral pressors: Epinephrine and phenylephrine are safe when given peripherally. 🌊 Norepinephrine can cause extravasation, but it may nonetheless be safe with adequate monitoring. 🌊 Patients who require low dose pressors for a limited period of time don't necessarily need a central line.
• Midline catheters: may be utilized for vasopressor infusion, without any risk of central line infection.

[2] dirty/crash central lines

• It's fine to insert a dirty/crash central line without full sterility, but it must be removed within <48 hours.
• If a line is correctly designated as dirty and removed within <48 hours this will not cause CLABSI.

[3] meticulous insertion of clean central lines

• Consider delaying central line placement. Immediate stabilization can often be achieved using peripheral vasopressors and/or intraosseous lines. Subsequently, a fresh central line may be placed in a controlled and meticulously clean manner.
• Place central lines with full sterility: Use a hat, mask, gown, full body drape, etc.
• Use chlorhexidine: This is shown to be more effective than povidone iodine across a variety of contexts.
• When inserting a central line, don't make a large nick with a scalpel along the guidewire. It is better to make a small nick, so that the skin incision is nice and tight (this will require using a bit more force to advance the dilator). Using a small nick will produce a central line site which remains clean and dry.
• Use the subclavian site: Compared to the internal jugular or femoral sites, the subclavian site has a lower risk of thrombosis or line infection. If possible, this site is recommended by United States guidelines. (21460264)
• Avoid catheter manipulation or repositioning: If the catheter site isn't perfect but is likely in a safe position, consider just using it in situ without attempting to reposition it. 🌊

[4] prompt discontinuation of central lines 

• Discontinue central lines ASAP. The necessity of the line should be reviewed daily.
• Transition to a PICC line for long-term central access. PICC lines have a lower per-day infection rate than traditional central lines, so consider transitioning from a central line to a PICC line earlier rather than later.
• Consider early placement of tunneled hemodialysis catheter: For patients who will require chronic dialysis, placement of a tunneled dialysis catheter should be performed sooner rather than later (to allow for removal of the temporary hemodialysis catheter).

---

clinical features

(back to contents)

---

timing of line infection

• Line infection is extremely rare within <48 hours of line placement. Infection before 48 hours doesn't meet the definition of CLABSI (central line associated bloodstream infection). (37733310)
• Infection is uncommon within the first week of line placement.

[#1/3] non-focal manifestations

• Fever.
• Septic shock.

[#2/3] local signs of infection

• These may include:
  • Pain.
  • Erythema.
  • Purulent exudate (this should be sent for gram stain & culture). (37014953)
• These signs are usually absent. Obvious evidence of infection at the line site is specific, but insensitive.(30241718)

[#3/3] late-onset line dysfunction

• Dysfunction of a central line shortly after placement usually reflects a mechanical complication (e.g., the line is kinked or some ports are pressed up against a vessel wall).
• If a central line functions fine initially, but later stops working well this may support the diagnosis of line infection. Late-onset line dysfunction often reflects thrombosis, which correlates with infection.

---

diagnosis

(back to contents)

---

blood cultures

• Draw two sets of peripheral blood cultures (from different sites) and one set of cultures from the distal port of the line.
  • Some evidence suggests that obtaining two sets of blood cultures from two lumens of the central line could increase the sensitivity for central line infection (20455693). However, obtaining more cultures will also serve to increase the likelihood of false-positive cultures due to contamination.
• Four general patterns may be discerned:
  • (1) Peripheral and central line cultures all turn positive with incubation times <2 hours of one another: this suggests bacteremia due to another cause (e.g., endocarditis). However, this can definitely occur with catheter infection, so clinical judgement is required. If no obvious source of bacteremia is present, then catheter infection may become increasingly likely. (29432818)
  • (2) Central line cultures turn positive first, then subsequently peripheral cultures turn positive with incubation times >2 hours longer. This is fairly diagnostic of central line infection, (~85%) but other possible sources still warrant consideration. (15767623)
  • (3) One of the peripheral cultures turns positive, but none of the other cultures do. This suggests a contaminated culture. (⚠️ However gram-negative organisms or yeast should generally be assumed to be real, even if they only turn positive in a single culture.)
  • (4) The central line culture turns positive, but the peripheral blood cultures do not. This presents a bit of a quandary. It might represent contamination, but it could represent early colonization or infection of the line. Line removal is indicated. Systemic antibiotics aren't always needed, with judgement required based on clinical context (e.g., presence of intravascular hardware). If the organism involved is Staph. aureus, Candida spp., Pseudomonas spp., or Acinetobacter spp. then antibiotics are advisable – more on this below. 📖 (32894389)
• Low-virulence skin organisms may indicate a line infection if they are grown from two different sites (indicating true bloodstream infection rather than contamination). Such organisms include:
  • Coagulase negative staphylococci.
  • Corynebacterium spp.
  • Bacillus spp.
  • Malassezia furfur
  • Micrococcus spp.
  • Cutibacterium spp. (previously: Propionibacterium spp.).
• (Quantitative culture techniques can also be used to compare bacterial burden in peripheral versus central cultures. However, most hospitals lack these techniques. Additionally, differential time to positivity may be more accurate than quantitative cultures.)(17304454)

line tip culture

• This involves cutting off the line tip and sending it for culture (after removing the line).
• Limitations to this technique:
  • i) It requires removing the line.
  • ii) The tip can get contaminated during line removal, leading to a “false-positive” diagnosis of line infection.
• Reasonable approach?
  • Generally, avoid culturing line tips (the preferred technique for diagnosis of line infection is differential time to blood culture positivity, as discussed above).
  • Line tip culture may be rarely utilized in unusual situations (e.g., inability to obtain peripheral blood cultures, inability to draw blood through the catheter due to thrombosis).

procalcitonin

• Available data is summarized above.
  • A very low procalcitonin argues against bacteremia (e.g., procalcitonin <0.3 ng/ml).
  • Elevated procalcitonin is nonspecific (e.g., it could represent line infection, infection elsewhere, or renal dysfunction).
• Use of procalcitonin in suspected line infection?
  • Procalcitonin should not be used in the decision of whether to initiate antibiotics.
  • A low procalcitonin may be used to support a decision to discontinue antibiotics (in combination with clinical judgement).
• The evidentiary basis supporting procalcitonin here remains weak, so this should be interpreted judiciously. For example, in a patient with known pneumonia, procalcitonin will likely be positive whether or not there is a line infection.

---

diagnostic criteria

(back to contents)

---

Diagnostic criteria aren't really useful for clinical management, but this will become relevant in terms of reporting infections.  

CLABSI (central line associated bloodstream infection)

• Criteria:
  • [1] Central venous catheter in place for >48 hours prior to positive blood culture.
  • [2] No other source of bacteremia or fungemia identified.
• This is a epidemiological surveillance tool that will overestimate the true incidence of infection.

CRBSI (catheter-related bloodstream infection)

• CRBSI is a subset of CLABSI.
• Criteria:
  • [1] Requires the presence of CLABSI as outlined above.
  • [2] Clinical evidence of infection (e.g., fever, leukocytosis, erythema at the catheter site, differential time to positivity of blood cultures).
• CRBSI rates are usually not reported because this diagnosis is too complex to be easily applied in a broad fashion.

---

overall management strategy

(back to contents)

---

indications for line removal and empiric antibiotics:

• [1] Definite line infection:
  • Unequivocal local signs of infection (e.g., purulent drainage).
  • Positive blood cultures from the line.
• [2] Possible infection –plus– increased risk of harm from leaving the line in, due to:
  • Septic shock.
  • Endovascular hardware (e.g. pacemaker, prosthetic valve), or fresh arterial graft. These increase the risk of an infection spreading to a site where it would be difficult to eradicate.

are there situations where the line can be left in place despite infection?

• Occasionally, for patients with infected ports or tunneled hemodialysis lines, it might be possible to treat with antibiotics alone (without hardware removal). This should be considered only if all of the following conditions are met:
  • (1) Patient isn't in septic shock.
  • (2) The organism isn't Staphylococcus, Pseudomonas, a multi-drug-resistant gram-negative, Candida, Acinetobacter, Micrococcus spp., or Propionibacterium spp.
  • (3) Infectious disease consultant agrees that this is a wise course of action.

can the line be exchanged over a guidewire?

• Line exchange over a guidewire has been shown to increase the risk of subsequent infection (compared to placement of a fresh line). (9267959)
• Reasons that guidewire exchange may increase infection risk:
  • i) It's nearly impossible to do this with true 100% sterility.
  • ii) The skin tract may be infected, so even if perfect sterility is achieved, the new line may immediately be contaminated by residual bacteria.
• In the modern era of ultrasound-guided line placement, guidewire exchange is a highly dubious practice (placement of a fresh line can generally be accomplished safely).

---

empiric antibiotic selection

(back to contents)

---

microbiology of line infections

• Gram-positives are most common (~75%):
  • Coagulase negative staphylococci (~17-37%).
  • Staphylococcus aureus (13%).
  • Enterococcus spp. (~14%).
• Gram-negative bacilli (~14-20%):
  • Klebsiella spp. (8%).
  • Escherichia coli (5%).
  • Enterobacter spp. (4%).
  • Pseudomonas spp. (4%).
  • Serratia spp.
  • Acinetobacter spp.
• Candida (6-12%). (27573805, 37014953)

Empiric therapy will usually involve two antibiotics as follows, with a goal of covering MRSA and Pseudomonas:

[#1] agent covering drug-resistant gram positives (MRSA)

• Usual choices:
  • Vancomycin 💉
  • Daptomycin 💉
• More on the choice of antibiotic coverage for resistant gram-positives here: 📖

[#2] antipseudomonal beta-lactam

• Options include:
  • Piperacillin-tazobactam. 💉
  • Cefepime. 💉
  • Meropenem. 💉

[#3] rarely may consider empiric coverage for Candida

• Empiric candida therapy is generally not recommended. (32894389)
• The EMPIRICUS trial demonstrated that empirical micafungin didn't improve outcomes among patients with ICU-acquired sepsis, an indwelling vascular catheter, candida colonization, multiple organ failure, and exposure to broad-spectrum antibacterial agents.(27706483) 🌊
• An echinocandin could be reasonable for individual patients with numerous risk factors for candida as well as severe septic shock. Risk factors include:
  • Colonization with candida at multiple sites.
  • Prolonged exposure to broad-spectrum antibiotics.
  • ICU stay >1 week.
  • Total parenteral nutrition.
  • Immunosuppression (chemotherapy, neutropenia, transplantation, hematologic malignancy).
  • Gastrointestinal perforations/anastomotic leaks.
• The initial antibiotic of choice is generally an echinocandin (e.g., caspofungin or micafungin).
• When candidemia is a concern, consider obtaining a Beta-D-Glucan level as well.

---

treatment duration

(back to contents)

---

The following is a rough guide to treatment duration for uncomplicated infections. (32894389)

Staphylococcus aureus, Staphylococcus lugdunensis (SLUG), or Candida spp.

• Line colonization (culture from line positive, but peripheral blood cultures negative): 3-5 days.
• Line infection without remote complications: 14 days.
• Line infection with remote complications: 4-6 weeks.

enterobacteriaceae, Enterococci, CoNS (coagulase-negative Staphylococcus)

• Line colonization (culture from line positive, but peripheral blood cultures negative): There may be no need for antibiotics at all.
• Line infection, without distant complications: 7 days.
• Line infection with remote complications: 4-6 weeks.

Pseudomonas aeruginosa or Acinetobacter baumannii

• Line colonization (culture from line positive, but peripheral blood cultures negative): 3-5 days.
• Line infection, without distant complications: 7 days.
• Line infection with distant complications: 4-6 weeks.

---

persistent bacteremia

(back to contents)

---

test for clearance of bacteremia

• Antibiotics plus catheter removal should sterilize the blood within <72 hours.
• Blood cultures should be repeated after >72 hours. Testing for clearance is especially important in Staph. aureus and enterococcus. Up to half of patients with persistent Staph. bacteremia may lack fever. (16338832)

differential diagnosis of persistent bacteremia includes:

• Incorrect antibiotic, or inadequate dose of antibiotic.
• Endocarditis.
• Infected endovascular hardware (e.g., venous stent).
• Septic thrombophlebitis.
• Septic pulmonary emboli.
• Distant metastatic foci of infection, e.g.:
  • Joint infection.
  • Vertebral osteomyelitis.
  • Splenic abscess.
  • Secondary infection of prosthetic material. (37014953)

investigation of persistent bacteremia

• Vascular ultrasonography to evaluate for deep vein thrombosis (which may imply septic thrombophlebitis).
• Transthoracic echocardiography (TTE).  Transesophageal echocardiography (TEE) should be considered as well, especially if: (32894389)
  • Organism involved tends to cause endocarditis (e.g. Staphylococcus aureus, Enterococcus, Candidiasis).
  • Prosthetic heart valves or pacemaker.
  • Valvular or structural heart disease.
  • History of endocarditis.
• Investigate for any distant metastatic foci of infection.

management

• Endocarditis may require specific management. 📖
• Septic thrombophlebitis may benefit from heparin in addition to antibiotics.(30377625)
• Distant metastatic foci of infection may require drainage or debridement.
• The duration of antibiotic therapy may often need to be extended out to 4-6 weeks.

---

septic thrombophlebitis

(back to contents)

---

diagnosis

• Septic thrombophlebitis may be suggested by:
  • Signs of deep vein thrombosis (e.g., extremity swelling/erythema, palpable cord).
  • The line stops working (occluded by clot).
  • Persistent sepsis, fever, or positive blood cultures >72 hours after line removal.
  • Septic pulmonary emboli (which typically cause a multifocal nodular-distribution pneumonia with cavitation).
• Diagnosis is based largely on ultrasonography revealing thrombosis, combined with persistently positive blood cultures for >72 hours.

management

• The line must be removed (if it hasn't been already).
  • There may be concern that line removal will dislodge the clot.  This is a theoretical concern which is not evidence-based.  Prompt removal of the line is preferred to facilitate infection control.
• Antibiotic therapy duration may need to be extended (e.g., to 4-6 weeks).
• Anticoagulation may be indicated if the thrombus is within a deep vein (i.e., a deep vein thrombosis).

---

arterial line infection

(back to contents)

---

radial arterial catheter

• Radial arterial catheters seem to have an extremely low rate of infection.
  • One meta-analysis found the infection rate to be 0.3% (1 in 333).(24413576) However, determining the true infection rate is challenging, given that line infection is often a clinical diagnosis (without a single definitive gold-standard diagnostic).
• Radial arterial lines generally stop working after a couple of days, leading to their removal. This naturally limits the ability of radial arterial lines to cause infection.

femoral & axillary arterial catheters

• Femoral arterial lines are associated with a higher rate of infection than radial arterial lines.(24413576) Drivers of this higher rate of infection may be:
  • It is difficult to keep groin lines clean and dry over several days.
  • Femoral lines often stay in longer than radial lines, because femoral arterial lines tend to function well for a longer period of time.
• The infection rate of axillary arterial catheters is unclear. These lines can function well for prolonged periods of time, which could increase the risk of infection.
• Both axillary and femoral arterial catheters should be placed under full sterile conditions, similar to those of a central line.

---

septic shock from an infected chest port

(back to contents)

---

• This is uncommon, but definitely happens. Make sure to always consider any hardware in the patient's body when evaluating for the source of sepsis.
  • This can occur even if the port isn't being actively used (due to seeding of the catheter).
• Diagnosis is similar to that of a central line infection, based on some combination of the following:
  • (1) Local signs of infection (erythema, warmth, pain).
  • (2) Positive blood cultures (especially with a gram-positive organism, but a gram-negative organism may be involved as well).
  • (3) Lack of any other clear focus of infection.
• For patients in shock, immediate removal is necessary to achieve source control:
  • This is a fairly minimal bedside procedure, which can be done under local anesthesia (with or without conscious sedation).
  • Both surgeons and interventional radiologists are entirely capable of performing the procedure. If a patient is actively dying and neither are available, the procedure could probably be done by anyone with basic procedural skills.
• Key point: Saving the patient's life is always more important than saving the port. When facing a septic patient, it's generally safest to err on the side of port removal.

---

physiology

(back to contents)

---

various mechanisms of infection:

• [1] Catheter contamination at insertion site (extraluminal spread of bacteria).
  • Usually occurs at the time of line insertion.
  • Infection most often arises during the first week after catheter placement. (37733310)
• [2] Catheter hub manipulation with intraluminal spread.
  • Usually occurs >1 week after catheter insertion. (37733310)
• [3] Catheter contaminated by secondary infection.
• [4] Contaminated infusate.

---

podcast

(back to contents)

---

Follow us on iTunes

---

questions & discussion

(back to contents)

---

To keep this page small and fast, questions & discussion about this post can be found on another page here.

• Don't ignore blood cultures showing organisms that are typically contaminants in patients with indwelling lines (e.g., coagulase-negative staphylococcus or bacillus spp.). If multiple cultures are positive for one of these organisms, it may reflect a line infection.
• When treating a line infection, repeat cultures after 72 hours to ensure sterilization of the blood. Failure to achieve this may suggest endocarditis or septic thrombophlebitis.

Guide to emoji hyperlinks

• = Link to online calculator.
• = Link to Medscape monograph about a drug.
• = Link to IBCC section about a drug.
• = Link to IBCC section covering that topic.
• = Link to FOAMed site with related information.
• = Link to supplemental media.