A Clinical guide to pediatric infectious disease

15

Catheter Infection

Epidemiology

The most common mechanism of catheter infection is the colonization by bacteria at the insertion site with subsequent migration to the catheter tip. A less frequent cause of catheter infection is the hematogenous seeding of the catheter from a bacteremia originating at a separate site.

Presentation

Consideration of catheter infection should be made in all patients with indwelling catheters presenting with fever. This is particularly true in patients who have signs of inflammation around the catheter itself. The evaluation of a febrile patient who has an indwelling catheter should always include the following:

1. Vital signs.Decreased peripheral perfusion and hypotension may be signs of progressive sepsis and should always be part of the evaluation.
2. Timing of catheter placement.Patients who have had catheter placement within the past 2 to 3 weeks are at a higher risk for infection with Staphylococcus aureus.
3. History of previous catheter infection.A patient recently completing antibiotic therapy, especially if the treatment did not include catheter removal, may be presenting with a relapsed infection with the sameorganism.
4. Any other explanations for fever, including signs of upper or lower respiratory infection.

P.151

Diagnosis of Catheter Infections

The hallmark of the diagnosis of catheter infection is a positive blood culture from the catheter. One of the challenges is that the blood culture may be positive in a patient with a catheter who has a catheter infection or has sepsis originating at a separate site. Furthermore, a bacteremia originating at another site may ultimately seed and infect that catheter.

Several methods have been proposed for the diagnosis of catheter infection:

1. Culture of the catheter tipis a common method used for the diagnosis of catheter infection. In the semiquantitative method, the catheter segment is rolled across the agar plate, and colony-forming units are counted after 24 hours of incubation. The quantitative method of catheter culture is done by flushing or vortexing the catheter segment with broth, followed by serial dilution or plating on an agar plate. A colony count of at least 15 colony-forming units by semiquantitative method or at least 100 colony-forming units from the quantitative method is indicative of catheter infection. An obvious disadvantage of this method is that the catheter has to be removed before the diagnosis of infection can be confirmed.
2. Paired quantitative blood cultures.Patients in whom catheter infection is considered should always have two sets of blood cultures obtained: a blood culture from the catheter in question and a peripheral blood culture. A blood culture from an infected catheter should have 5 to 10 times more colony-forming units per cc than the peripheral blood culture. The use of paired quantitative blood cultures offers the advantage of evaluating for catheter infection without removal of the catheter.
3. Time to positivity. New technology in the processing of blood cultures allows for continuous monitoring, allowing for a time to positivity to be documented. It has been reported that there is a difference between the time to positivity between blood cultures from peripheral veins and blood cultures obtained from infected catheters. This time difference is related to the size in the inoculum of bacteria between a blood culture obtained through an infected catheter and that from a peripheral site. A time difference of greater than 2 hours is highly sensitive for the diagnosis of catheter-related infection. A limitation to this methodology is that an ability to monitor the automated blood culture system must be available.
4. Persistent positive cultures. This is often one of the most practical ways of proving a catheter infection. Repeated positive cultures through the catheter despite the administration of appropriate antibiotics strongly suggests a source of infection, the source usually being the catheter itself. After 48 to 72 hours of appropriate antibiotic treatment, continued positive blood cultures through the catheter suggest the diagnosis of catheter infection.

A summary of the diagnostic methods for catheter infection is provided in Table 15.1.

P.152

Management

Empiric Antibiotic Therapy

Depending on the clinical status of the patient, empiric therapy should be broad. Many clinicians, when faced with a potential catheter infection, include vancomycin (as coverage for coagulase-negative staphylococcus, S. aureus, methicillin-resistant S. aureus, and enterococcus). Some physicians who wish to limit the empiric use of vancomycin begin with semisynthetic penicillin such as nafcillin and resort to vancomycin only when a pathogen requiring vancomycin is documented. Because of the rapid mortality caused by gram-negative organisms, aggressive gram-negative coverage is usually begun. This often consists of a third-generation cephalosporin combined with an aminoglycoside. In patients who are severely ill, with decreased peripheral profusion and hypotension, empiric antifungal treatment may be started. The initial empiric therapy can always be reduced after culture results are available.

Treatment of Catheter Infections

There are two options for the treatment of catheter infections:

1. Antibiotic therapy alone
2. Catheter removal with antibiotic therapy

Organisms that infect catheters often produce a biofilm that causes the organism to adhere to the catheter. It has been shown that antibiotic concentration must be 100 to 1,000 times greater to kill bacteria residing within a biofilm. This explains the difficulty in treating infected catheters with antibiotic therapy alone.

Guidelines have recently been published by the Infectious Disease Society of America. This important document states that randomized trials regarding treatment of infected catheters are lacking; definitive recommendations therefore are often not available. Recommendations are based on a consensus from a panel of experts, taking into account a large amount of historical experience with catheter infections.

P.153

These guidelines are grouped according to the pathogens causing the catheter infection.

It is important to remember that a clinical or microbiological relapse following completion of antibiotics with the same pathogen in a patient in whom the catheter was retained, warrants immediate consideration for catheter removal, regardless of the pathogen.

Specific Pathogens of Catheter Infection

Coagulase-negative Staphylococcus

Coagulase-negative staphylococci are the most common cause of catheter infection. These bacteria may have lower virulence than other pathogens, and one can consider treatment with antibiotic therapy alone. The treatment is with oxacillin or nafcillin for methicillin-sensitive isolates and with vancomycin for organisms that are methicillin resistant. Some investigators believe that infection with coagulase-negative staphylococcus represents an infection with a heterogenous population of organisms; that is, the organisms involved in the infection may be both oxacillin sensitive and resistant. Thus, some clinicians use vancomycin for any coagulase-negative staphylococcus infection thought to require treatment. Treatment is for 10 to 14 days in patients in whom the catheter is retained.

Staphylococcus aureus

1. aureusis a pathogen of higher virulence, often associated with secondary seeding (i.e., endocarditis and osteomyelitis). Many clinicians believe the treatment of an S. aureuscatheter infection with antibiotics alone is difficult and often results in continued bacteremia and increased risk for infection of secondary sites. The most current recommendation includes removal of the catheter and antibiotic therapy for 14 days.

The diagnosis of endocarditis in patients with S. aureus bacteremia remains controversial. In adults, transesophageal echocardiography (TEE) is routinely used to assess for vegetations. In patients who have had a prolonged course of bacteremia or have abnormalities on TEE, treatment with antibiotics is often given for 4 to 6 weeks.

In pediatrics, the use of TEE is limited, making the diagnosis of S. aureus endocarditis more difficult. Some clinicians, when faced with persistent bacteremia preceding or following catheter removal, make a presumptive diagnosis of endocarditis and consider a prolonged course of antibiotic therapy.

Gram-negative Bacilli

The gram-negative bacilli are increasingly causing catheter-related infections. No controlled studies have addressed definitively whether catheters infected by gram-negative bacilli must be removed. In some settings in which the patient is hemodynamically

P.154

stable, attempts with medical therapy alone can be tried. In patients who present with a septic picture, hypotension, or system failure, immediate catheter removal should be considered. A recent study examining gram-negative catheter infection in neonates found that the bacteremia was successfully treated with medical management in only 45% of cases; successful therapy was most likely to occur when there was only a single day's duration of bacteremia. Infection was rarely resolved in infants who had more prolonged bacteremia and bacteremia-associated thrombocytopenia, unless the catheter was removed.

Candida Species

Candida species are another major cause of morbidity and mortality in patients with catheter infections. Once a fungal catheter infection is documented, removal of the catheter is necessary. Retention of the catheter will only result in persistent fungemia, which can cause significant morbidity and even mortality.

Although minimal inhibitory concentration (MIC) breakpoints are not available for amphotericin B, breakpoints for candidal species to the azole class (e.g., fluconazole, itraconazole) are available. The Infectious Disease Society of America has recently published recommendations regarding the treatment of a variety of Candida species with antifungal agents. Infections with Candida albicans, Candida tropicalis, and Candida parapsilosis can usually be treated with fluconazole. Candida krusei and Candida glabrata are generally considered resistant to fluconazole and may have reduced sensitivity to amphotericin B, requiring either increased dosing of amphotericin B or the use of caspofungin. Candida lusitaniae is considered to be potentially resistant to amphotericin B. Current guidelines recognize that resistance patterns may change over time; testing for azole resistance may need to be increasingly used, particularly if faced with clinical or microbiological failure.

Duration of therapy for fungemia is a common question. Adequate therapy is desired because of the high risk for secondary spread to bone, kidneys, or cerebrospinal fluid. In the past, all fungemia was treated with a 6-week course of antifungal agents. Recently, efforts to determine which patients may be candidates for a shorter course of antifungal therapy have been made. Early catheter removal is essential. A vigorous search for secondary sites of infection by urine culture, urinalysis, funduscopic examination, and renal ultrasound are often recommended. If repeat blood cultures following catheter removal are negative and there is no evidence of a secondary infected site, many clinicians will administer 2 weeks of additional therapy following first negative culture. Secondary sites of infection needprolonged therapy (Table 15.2).

Routine Changes of Central Venous Catheters

Recently, several studies have examined the prevention of catheter-related bloodstream infections in the intensive care setting with scheduled replacement of these

P.155

catheters over guide wires. These studies have not shown a decrease in infection; there was actually an increased risk for mechanical complication when the catheters were replaced. The most current recommendation is that routine scheduling and changing of catheters does not prevent catheter-related bloodstream infections.

Selected Readings

Benjamin, DK, Miller W, Garges H, et al. Bacteremia, central catheters, and neonates: when to pull the line. Pediatrics2001;107(6):1272–1276.

Donowitz LG, Hendley JO. Short course amphotericin B therapy for candidemia in pediatric patients. Pediatrics 1995;95(6):888–891.

Gaur AH, Flynn PM, Giannini MA, et al. Difference in time to detection: a simple method to differentiate catheter-related from non-catheter related blood steam infection in immunocompromised pediatric patients. Clin Infect Dis 2003;37(4):469–475.

Haimi-Cohen Y, Shafinoori S, Tucci V, et al. Use of incubation time to detect in BACTEC 9240 to distinguish coagulase-negative staphylococcal contamination from infection in pediatric blood cultures. Pediatr Infect Dis J 2003;22:968–973.

Mermel LA, Farr BM, Sheretz RJ, et al. Guidelines for the management of intravascular catheter-related infections. Clin Infect Dis2001;32(9):1249–1272.

Nazemi KJ, Buescher ES, Kelly RE, et al. Central venous catheter removal versus in situ treatment in neonates with Enterobacteriaceae bacteremia. Pediatrics 2003;111(3):E269–274.