Central Line–associated Bloodstream Infection in Children

Long-term central venous catheters (CVCs) are essential for modern pediatric practice. These devices terminate in a large central vein, usually the superior vena cava, and are used for administration of drugs, fluids and blood products; for blood collection and for hemodialysis. The most common, serious complication of CVC use is central line–associated bloodstream infection (CLABSI). Rates of CLABSI vary widely according to device type and patient population and can range from approximately 0.2 episodes per 1000 catheter-days in children with sickle cell disease to 11 per 1000 days in infants with intestinal insufficiency.1,2 The most common infecting organisms are coagulase-negative staphylococci (especially Staphylococcus epidermidis), Staphylococcus aureus, Enterococcus spp., Escherichia coli, Klebsiella spp., other enteric Gram-negative bacteria and Candida spp. Microorganisms are introduced predominantly through the hub during routine use or at the time of catheter insertion. After the first 14 days, intraluminal colonization is the most important source of infection.3 Important consequences of CLABSI include extended hospital stay (median of 12 days in 1 study), interruption of chemotherapy or other treatment, catheter removal (up to 50% of episodes), intravascular thrombosis, endocarditis, sepsis (up to 10%c) and rarely death.4,5 Techniques to prevent CVC colonization during insertion, such as maximum sterile barrier precautions, reduce the risk of CLABSI. After insertion, appropriate CVC dressings, careful catheter access technique, alcohol catheter caps and the use of prophylactic lock therapy with taurolidine, antibiotic or preservative-containing heparin solution can reduce the risk of infection in some groups.6–9 This review focuses on the management of CLABSI in children, outside the neonatal period, who have long-term tunneled CVCs (eg, Broviac or Hickman catheters, Bard Access Systems, Salt Lake City, Utah) or implantable ports (eg, Port-A-Cath, Smiths Medical, Dublin, Ohio). Major controversies include duration and choice of systemic antibiotic therapy, indications for catheter removal, screening for complications and the roles of adjunctive treatment and secondary prevention techniques. This review assesses the available literature and identifies pragmatic treatment strategies based on the best available evidence basis. The Infectious Diseases Society of America guidelines, which focus predominantly on adult patients, are referenced where a consensus recommendation is required but evidence is limited or conflicting. DIAGNOSIS AND DEFINITIONS The classic presentation of CLABSI is the development of fever and chills immediately after accessing a catheter that has been locked for some time. However, the range of clinical presentations is broad, and the catheter may not always be immediately considered as the source of fever. Most episodes are not associated with any visible abnormality at the site of the catheter.3 Unless an alternative source is identified, all bloodstream infections in patients with a CVC are classified as CLABSI. This surveillance definition may overestimate the true number of infections that are attributable to the CVC. Therefore, when there is evidence to confirm that the colonized device is the true source of infection, the more specific diagnosis of catheter-related bloodstream infection (CRBSI) is used (Fig. 1).10 Confirmatory tests include culture of the catheter tip, quantitative blood cultures or differential time to positivity (DTP) of blood cultures drawn from different sites. Data from pediatric oncology patients suggest that up to one third of CLABSI episodes are related to infections at other sites and are therefore not CRBSI.10 Definitive diagnosis of CRBSI can be important to identify those patients who might benefit from catheter removal or adjunctive therapy. CVC tip culture can identify CRBSI, but precludes salvage of the catheter.FIGURE 1: Categories of bloodstream infections.10 BSI indicates bloodstream infection.The most readily available technique to confirm CRBSI without catheter removal is the calculation of DTP between blood cultures drawn from the catheter and from a peripheral vein or separate lumen. In cases of true CRBSI, blood obtained through a colonized lumen will usually indicate growth before uncolonized lumens or peripheral blood because of high intraluminal microorganism burden. To apply, identical volumes of blood must be collected for culture simultaneously from each site, and a continuously monitored blood culture system is needed. In pediatric oncology patients who had simultaneous peripheral and central cultures, a DTP of ≥150 minutes had a specificity of 100% and sensitivity of 89% for CRBSI. For cultures from both lumens of a double-lumen catheter, a DTP of ≥180 minutes had a specificity of 94% and sensitivity of 61%.11,12 These data suggest that DTP can be confidently used to diagnose CRBSI, but that the negative predictive value is poor, especially when comparing 2 lumens of a double-lumen catheter. To maximize the sensitivity and specificity of diagnosis in cases of suspected CRBSI, blood cultures should be collected from the catheter and from the peripheral blood before initiation of antibiotics.13 Blood should generally be collected from all lumens of a multilumen catheter (although there is an absence of data to support this consensus recommendation), as some can be uncolonized. Similarly, in patients with a single-lumen catheter, at least 2 sets of blood cultures should be collected through the CVC if a peripheral culture is not taken. Normal skin flora are the most common causes of CLABSI and are also frequent blood culture contaminants. Therefore, definitive diagnosis of bloodstream infection with a common skin contaminant requires multiple positive cultures to exclude the possibility of contamination during collection or inoculation. If multiple blood cultures have not been collected before antibiotics, a presumptive diagnosis of CLABSI has to be made, leading to unnecessary treatment in cases where the positive culture has resulted from contamination. Peripheral blood should be collected in all cases of suspected CLABSI before antibiotics are started, but pediatricians often avoid this. There are a number of important situations in which peripheral blood culture aids management. One example is a patient with severe sepsis in whom catheter removal might limit venous access and compromise treatment. Another is attempted catheter salvage complicated by persistent bacteremia during appropriate antibiotic therapy in which evidence that the catheter is not the primary focus could lead to earlier identification of an intravascular focus such as endocarditis. Lastly, definitive diagnosis of CRBSI can support a decision to remove and replace a catheter in cases of S. aureus or candida infection. All these situations can be difficult to predict at initial presentation. After diagnosis of CLABSI, blood cultures should be collected daily until cultures from all lumens are negative. Further peripheral cultures may not be necessary if blood can be collected from all lumens of the catheter and the patient is clinically stable. Diagnosis of Complications Resulting From CLABSI Complicated catheter-associated infections include tunnel or port pocket infection, endovascular infections such as endocarditis or suppurative thrombophlebitis and metastatic infections such as osteomyelitis and liver, spleen or brain abscess. Catheter-related skin and soft tissue infection can be identified by localized erythema, tenderness, fluctuance or swelling.13 An intravascular focus may be identified when there is persistent bacteremia or fungemia after CVC removal, despite appropriate antimicrobial therapy. Clinical signs, such as new murmur, peripheral stigmata and cardiac failure or localized tenderness and swelling of a limb, are often absent in infective endocarditis and suppurative thrombophlebitis, respectively.14,15 CATHETER SALVAGE Catheter salvage is attractive in many cases because of the cost and inconvenience of replacement. However, catheter removal substantially reduces the risk of relapse, persistent bacteremia or fungemia, and metastatic infection.16 The risk of relapse differs by device and organism. For ports, the risk may be >50%, probably related to the complex structure and high internal surface area of the port housing.10 Salvage should not be attempted if the catheter is no longer essential for clinical care. Timely removal of the CVC is also indicated for infections complicated by sepsis, tunnel or port pocket infection, endocarditis or suppurative thrombophlebitis, relapse of CLABSI with an identical organism, and infection with mycobacteria, fungi such as Candida spp., S. aureus, Bacillus cereus and some multiresistant bacteria. With attempted catheter salvage, candida and mycobacteria CRBSIs have treatment failure rates over 70%, and S. aureus and B. cereus infections have relapse rates around 50% with potentially catastrophic complications.13,17 The catheter should therefore be removed in all the above cases, except where replacement will be so difficult that the high risk of treatment failure is considered justifiable. It is appropriate to consider catheter salvage in other patients in view of the relatively high success rates of salvage therapy and potential hazards of CVC replacement. Catheter replacement is costly, carries risks associated with anesthesia and local trauma, does not reduce the risk of new CLABSI (the risk may actually be increased immediately after replacement) and venous occlusion can preclude future use of that insertion site.18 If salvage is attempted, the patient should be closely observed for the development of acute complications, persistent infection or sepsis. The catheter should be removed if complications develop or if bacteremia or fungemia persists after 72 hours of appropriate antibiotic therapy.13 After catheter removal, reinsertion of a new long-term CVC should ideally be delayed until blood cultures collected after removal are negative to prevent immediate contamination of the new device.13 This delay might be unnecessary for patients in whom clearance of bacteremia or fungemia is documented before catheter removal. The absolute risk of contamination of the new device is unknown in either case, so the decision of whether to delay reinsertion should take into account the potential clinical impact of either delayed reinsertion or colonization of the new catheter. The use of guidewire catheter exchange in the management of CLABSI is controversial. The technique has the benefit of reducing venous occlusion and mechanical complications of new device insertion, but it may be associated with contamination of the new catheter and can be technically difficult in young children.13 Guidewire replacement with an antimicrobial impregnated catheter could potentially be beneficial to prevent colonization of the new device. One retrospective study of nontunneled impregnated devices reported excellent outcomes, but there are no prospective studies supporting the method, and appropriate devices for tunneled use are not yet available.19 SYSTEMIC ANTIBIOTIC THERAPY Empiric antibiotic therapy for suspected CLABSI should take into account the clinical condition of the patient, documented past colonization or infection with resistant organisms, known allergies and local resistance patterns. Initial therapy should usually be with vancomycin combined with a broad-spectrum agent active against Gram-negative bacteria including Pseudomonas aeruginosa, such as an aminoglycoside (eg, gentamicin), and antipseudomonal penicillin (eg, piperacillin-tazobactam), or a cephalosporin (eg, ceftazidime or cefepime). If multiresistant Gram-negative infection is likely, the use of a carbapenem (eg, meropenem) should be considered. In patients with clinical sepsis, an additional third agent of a different antibiotic class with activity against Gram-negative bacteria is recommended, until susceptibility results are available because of the risk of resistance. Although evidence is limited, it seems logical to administer treatment through all lumens of the CVC by splitting or rotating doses. Empiric therapy for fungal infection is not routinely recommended except in high-risk patients; considerations include severity of illness and presence of multiple risk factors for candidemia, such as prolonged intensive care unit stay, administration of total parenteral nutrition, immunosuppression, prolonged antibiotic exposure, previous candidemia without catheter removal and known candida colonization.20 Once the causative organism is identified, targeted therapy should be selected based on susceptibility testing. Ideally, a single narrow spectrum agent should be used. If the catheter is retained, patients should receive 10–14 days of pathogen-specific systemic antibiotic therapy from the date of the first negative blood culture through each lumen of the catheter.13 A longer duration of therapy is indicated for cases complicated by endocarditis, suppurative thrombophlebitis or metastatic infection. Use of antibiotics, that have superior in vitro activity against biofilm bacteria, such as ciprofloxacin and rifamp(ic)in, has been proposed to potentially offer a greater chance of eradication of infection, but there is currently no evidence that these drugs improve clinical outcomes. Rifampin should not be used as a single agent for the treatment of infection due to rapid development of antibiotic resistance. If the catheter is removed, patients should receive 10–14 days of appropriate systemic therapy. Oral therapy may be considered if an appropriate drug is available and intravenous access is difficult to maintain.21 A shorter 5- to 7-day treatment course is reasonable for CLABSI due to coagulase-negative staphylococci if the catheter is removed and blood cultures clear promptly.13 There is no evidence to support the use of synergistic aminoglycoside in the absence of endocarditis. TREATMENT FAILURE Despite appropriate antibiotic therapy, treatment failure (persistent bacteremia, premature device removal due to infection or relapse of infection) occurs in at least 25% of CLABSI episodes in which salvage is attempted. Relapse can occur many months after infection.10 Failure is highest in infections caused by certain organisms (such as S. aureus, Candida spp., Bacillus spp. and mycobacteria), tunnel and pocket infection, and in ports.13 This is attributable to biofilm, organized communities of microorganisms in a sessile state, predominantly on the luminal surface of the catheter.3 Mechanisms of treatment failure are described in Figure 2.9 In addition to the risk of treatment failure, prior CLABSI increases the risk of future episodes.4 This might be associated with line-care technique, biofilm or catheter-associated thrombus acting as a nidus for reinfection, or the presence of multispecies biofilm.FIGURE 2: Mechanisms of treatment failure in CLABSI.CATHETER LOCK THERAPY Catheter lock therapy (CLT) may be used in addition to systemic therapy with the aim of reducing treatment failure by targeting biofilm bacteria. A small dose of antimicrobial agent is instilled to fill the catheter lumen (usually ≤1.5 mL) and is allowed to dwell for an extended period of time (usually hours to days). This permits high concentrations of antimicrobial agent at the site of infection and can be repeated in each lumen throughout the course of antibiotic treatment. For example, local vancomycin concentrations of up to 5 mg/mL, 1000 times higher than the usual minimum inhibitory concentration, are achievable. This targets intraluminal colonization and would not be expected to improve outcomes for extraluminal colonization or infected catheter-associated thrombus. Specific agents used for CLT are outlined below. Antibiotic lock therapy (ALT) with vancomycin, daptomycin, ceftriaxone, aminoglycosides, ciprofloxacin and other agents, individually or in combination, have been reported. Published studies on ALT are heterogeneous and mostly nonrandomized.8 One randomized study of vancomycin or ceftazidime lock therapy in addition to systemic antimicrobial therapy showed a nonsignificant trend in the reduction of treatment failure (33% versus 57%; P = 0.1) and relapse of infection (14% vs 39%; P = 0.06) in the treatment group.22 In adults, ALT seems to be well-tolerated with few adverse effects reported. Infectious Diseases Society of America treatment guidelines for CLABSI recommend the routine use of adjunctive ALT for adults and children.13 However, the uptake of ALT use in pediatric practice seems to be inconsistent. This may be related to concerns about the development of antibiotic resistance, a limited range of studied antibiotics, a paucity of pediatric safety data, the potential for heparin overdose in infants, the need for prolonged dwell times and the lack of commercially available lock solutions. Antibiotic solutions for ALT are produced extemporaneously and usually admixed with heparin to prevent catheter occlusion. Recommended heparin concentrations vary from 0 to 5000 units/mL depending on the antibiotic, but high concentrations are contraindicated in small infants due to the risk of systemic anticoagulation. More data are needed before a strong recommendation can be made on the routine use of adjunctive ALT in pediatrics, and ALT should not be considered mandatory in children.13 Ethanol lock therapy (ELT) for the treatment of CLABSI has not been studied in published randomized trials. Ethanol penetrates biofilm and rapidly kills microorganisms, including common CLABSI pathogens such as S. aureus, S. epidermidis, Klebsiella pneumoniae, P. aeruginosa and Candida spp., by protein denaturation. In vitro studies suggest that maximum killing occurs during the first 2 hours of exposure to 70% ethanol.23 One single-center retrospective review of pediatric oncology patients treated with and without ELT has been reported, which demonstrated a reduction in treatment failure, although this was not statistically significant. The small sample size and study design make these results difficult to apply broadly.24 Successful treatment of candida CRBSI using ELT has been reported in an uncontrolled series. This is interesting because of the historically poor outcomes of these infections.25 Other retrospective studies are uncontrolled and do not provide information about efficacy of the treatment.9 Episodes of self-limiting nausea, headache and taste of alcohol have been associated with ELT, but more serious side effects, such as respiratory distress, are rare. In 1 randomized study of ELT prophylaxis in adults, 5% of patients in the ELT arm discontinued the drug due to side effects with in the These effects are by the ELT dose and the lock after the dwell period and to be common in on catheter occlusion or intraluminal during the use of ELT, especially in children, have concerns that ELT may to but the and clinical of this are adjunctive ELT for the treatment of CLABSI is not recommended at this time due to lack of evidence for efficacy and about lock therapy has been as an adjunctive treatment for CLABSI in children with long-term retrospective studies an between lock therapy use and catheter but one an in mechanical complications versus P = The available data do not support the use of lock therapy for the treatment of CLABSI. AND thrombophlebitis, the presence of infected intravascular requires catheter removal and prolonged therapy as as in many However, the routine use of to for catheter-associated in the absence of clinical may be One prospective study intravascular thrombus in of all patients with CLABSI due to S. aureus, many of whom had no local of but the clinical impact of such blood is Another study used to for catheter-associated in all CLABSI episodes in adult patients with acute removed from patients with evidence of catheter-associated The study reported that from sepsis and catheter was common in the with but the study has a number of and no children Therefore, at venous is recommended when there are and of or persistent bacteremia or fungemia after 72 hours of appropriate antibiotic therapy and catheter removal. management of catheter-associated thrombus catheter removal with delayed reinsertion and with a pediatric may be with persistent bacteremia or fungemia hours after appropriate antibiotic treatment and catheter removal and those with cardiac disease are at high risk of If patients have risk factors or have clinical of endocarditis, and of endocarditis therapy are with S. aureus CLABSI have a high risk of endocarditis, and so treatment or presumptive of endocarditis by has been However, the risk of endocarditis associated with S. aureus CLABSI seems to be in children without specific risk factors without cardiac risk factors who develop S. aureus CLABSI therefore do not routine in the absence of clinical of endocarditis or prolonged The long-term risk of CLABSI is increased in children with prior or and careful to catheter care and access may prevent future episodes and should be in all The use of heparin with reduces the risk of CLABSI, but alcohol should be in In specific secondary such as prophylactic antibiotic or lock therapy, may be but there is no evidence for these as secondary future CLABSI include therapy, and other agents, which are against biofilm (eg, and are active against organisms biofilm (eg, taurolidine, and or with biofilm (eg, It is which of these will to trials. Despite the development of to reduce risks of infection related to CVC insertion and CLABSI to be a in pediatric have a high failure predominantly due to relapse and because of the in and intraluminal adjunctive are to be but evidence to support these is More is needed to the management of CLABSI in children and 1: The and of for the of CLABSI in