NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Holzheimer RG, Mannick JA, editors. Surgical Treatment: Evidence-Based and Problem-Oriented. Munich: Zuckschwerdt; 2001.

Cover of Surgical Treatment

Surgical Treatment: Evidence-Based and Problem-Oriented.

Show details

Bloodstream and intravascular catheter infections

, M.D. and , M.D.

Author Information

Bacteremia occurs in ~ 250,000 hospitalized patients annually in the U.S., causing considerable morbidity and increasing the risk of death nearly twenty fold (13). The magnitude of this problem only now is becoming recognized. Most patients undergo intravascular catheter placement at some time during their hospitalization, including ~ 145 million peripheral venous catheters and ~ 3–5 million central venous catheters per annum in the U.S., and ~ 18–25% of these catheters become colonized at some time during the course of their use. Ensuing infection is the most common complication associated with intravascular devices and subsequent bacteremia (occurring in ~ 3–5% of central venous catheters and ~ 0.5% of peripheral catheters) can be a highly morbid and oftentimes lethal consequence (46).

Sepsis syndrome

Sepsis syndrome represents a subset of the systemic inflammatory response syndrome (SIRS) in which patients harbor an infectious process and in which two or more of the following parameters are present: 1) temperature > 38 °C or < 36 °C, 2) heart rate > 90 beats/minute, 3) respiratory rate > 20 breaths/minute or PaCO2 < 32 torr, 4) white blood cell count > 12,000 cells/mm3, < 4000 cells/mm3, or > 10% immature (band) forms of neutrophils present on the peripheral blood smear (7, 8). Mortality associated with sepsis syndrome is ~ 35–40%, although fewer than half of patients have a microbial pathogen cultured from their bloodstream (7, 911). It is highly likely, however, that most episodes of bacteremia and fungemia are either intermittent, transient, or both precluding isolation of the offending pathogens during every clinical event.

Gram-negative bacteremia

Gram-negative bacteria account for ~ 30% of all episodes of bacteremia at most institutions. The mortality associated with gram-negative bacteremia in normal individuals is ~ 10% and may exceed 50% in immunocompromised patients. The most common causative microbes include Escherichia coli, Klebsiella pneumoniae, Pseudomonas aeruginosa, Enterobacter aerogenes and cloacae, although a vast array of organisms can be responsible. Infections caused by Pseudomonas and Klebsiella species occur more commonly in patients with serious underlying diseases (e.g., extensive trauma, burns, or malignancy) (7, 12).

Because of the high mortality associated with gram-negative bacteremia, empiric antibiotic therapy should be initiated as soon as the diagnosis is suspected. The initial choice of antimicrobial agent should be based upon the antibiotic resistance pattern of pathogens commonly encountered within the institution. In general, many β-lactam drugs (e.g. third-generation cephalosporins, acylampicillins, monobactams, carbapenems), or quinolones exhibit satisfactory activity against gram-negative bacteria. Trimethoprim-sulfamethoxazole may be useful for specific organisms as well. Aminoglycosides remain potent agents, but have fallen into disfavor because of their associated nephro- and ototoxicity and the perceived need for drug level monitoring. The diagnosis of gram-negative bacteremia is confirmed by isolation of a gram-negative bacterial pathogen from a blood culture in a febrile patient. Once an organism has been identified and antibiotic sensitivity testing performed, refinements in antimicrobial agent therapy can be made.

Clinical trials in which febrile neutropenic patients were treated with either single agents or two agents in combination - usually a β-lactam drug plus an aminoglycoside - have provided evidence that the dual agent therapy is more effective than single agent therapy in this particular immunosuppressed patient population (13, 14). Although use of this type of dual agent therapy is frequently extended to other patient groups, no clinical trials have been performed to provide evidence of similar efficacy in the general patient population. Thus, the added benefit of two agents has not been demonstrated in patients who are not neutropenic, and single agent therapy targeting the infecting organism should be utilized in most circumstances. However, when a microbe is identified that is known to be highly resistant to many agents (e.g., Pseudomonas, Aeromonas, Xanthomonas species) dual agent therapy should be considered.

Gram-positive bacteremia

Gram-positive organisms currently account for ~ 50–60% of nosocomial bacteremic events. Staphylococcus epidermidis is the most common gram-positive organism isolated from blood (~ 30% of isolates) and accounts for the majority of infections that are associated with an intravascular catheter. Staphylococcus aureus also causes a significant number of bloodstream and intravascular catheter infections and the incidence of clostridial and streptococcal infections has increased (2, 4). Enterococci have become a significant cause of bacteremia in surgical patients and have been isolated increasingly from patients with burns or multiple injuries (15).

The appearance of antibiotic resistance among gram-positive pathogens is particularly concerning. Methicillin resistant strains of Staphylococcus aureus (MRSA) have become increasingly prevalent and are an endemic and occasionally an epidemic problem in many hospitals. Staphylococcus epidermidis now is clearly recognized as an important nosocomial pathogen and ~ 80% of strains are resistant to methicillin. Vancomycin is one of the few antimicrobial agents that consistently is active against these pathogens, although isolation of vancomycin resistant Staphylococcus aureus (VRSA) recently has been reported. Many enterococci have become resistant to aminoglycosides as well as β-lactam antibiotics and the increasingly prevalence of vancomycin resistant enterococci (VRE) is of considerable concern. VRE isolates that cause infection require use of agents such tetracycline, rifampin, chloramphenicol, quinupristine-dalfopristine or other drugs alone or in combination according to the sensitivity pattern of the particular organism (16).

Because staphylococci are responsible for the majority of gram-positive bacteremic events, initial antibiotic therapy should target these organisms. Due to the appearance and rapid spread of VRE that has been associated with widespread vancomycin use, consideration should be given to the use of semisynthetic penicillins such as nafcillin or methicillin or a first generation cephalosporin as initial empiric therapy. In the situation of the β-lactam allergic patient or a patient who has developed a life-threatening infection, a defined, short course (at most 3 days) of vancomycin should be administered. In critically ill patients who have received a prior course of antimicrobial agents, the addition of antimicrobial coverage directed against Enterococcus fecalis and fecium (ampicillin or acylampicillin plus an aminoglycoside, or vancomycin alone for the β-lactam drug allergic patient or those who develop serious infections) should be considered. Empiric therapy subsequently can be tailored based on culture and sensitivity data.


Candida species are the most common isolates in fungemic patients, and C. albicans accounts for over half of the fungi cultured. In non-neutropenic patients, either intravenous fluconazole or amphotericin B can be used initially. Fluconazole is effective against many strains of candida, although most isolates of C. kruzeii, tropicalis, glabrata, and guilliermondii are resistant (17). If the infection is unresponsive based upon clinical parameters or if an organism resistant to fluconazole is identified, amphotericin B therapy should be instituted. Fungemic patients who exhibit hemodynamic instability or neutropenia should receive amphotericin B as the initial therapeutic agent. Patients who require amphotericin B and who exhibit renal dysfunction (serum creatitine > 2.5 g/dL) should receive liposomal amphotericin B.

Intravascular catheter infection

Superficial infections limited to the peripheral or central venous catheter exit or subcutaneous port site of an intravascular catheter are diagnosed readily by the presence of redness, swelling, pain, and occasionally purulent exudate. Such infections are treated effectively in the vast majority of cases by catheter (and port, if present) removal. Culture of the catheter tip or insertion site generally is unnecessary, unless purulence is present and the patient exhibits systemic manifestations of infection (e.g. fever, sepsis syndrome), or the patient is significantly immunosuppressed. Empiric antibiotics directed at the most common pathogens (i.e. gram-positive cocci) should be used to treat any accompanying cellulitis or ascending lymphangitis, and only occasionally - upon diagnosis of suppurative thrombophlebitis - is surgical intervention consisting of removal of the involved segment of extremity vein or venotomy and thrombectomy (central veins) plus intravenous antibiotics required for treatment. Note that suppurative thrombophlebitis of the great veins associated with central intravascular catheters is rare.

Invasive intravascular catheter infections are more difficult to diagnose, and it is often extremely hard to determine whether an infection at the catheter exit site solely is superficial and limited to the immediate region or invasive, extending to the intravascular portion of the catheter (tunnel or subcutaneous port site infection), let alone whether bacteremia is present. Several diagnostic principals should be considered when attempting to diagnose these insidious infections:


infections extending along the track of a tunneled catheter or subcutaneous port pocket are potentially serious and do not invariably manifest either local and/or systemic symptoms and signs, and


catheter infection should be suspected in any patient with an intravascular catheter who exhibits bacteremia and no other obvious source of infection.

A variety of methods of have been proposed to establish the diagnosis which entail catheter removal. In the most commonly used semiquantitative method of diagnosis, the distal 2 cm of the catheter is rolled across an agar plate and growth of > 15 colonies per plate is considered diagnostic of catheter infection (18). Because the most serious sequela of catheter infection is bacteremia, the presence of a positive blood culture is probably of considerable importance in diagnosing a catheter-related infection. In general, patients with signs of infection and a positive blood culture drawn peripherally or through the catheter or a positive catheter tip culture (> 15 colonies per plate) should be treated for catheter infection.

Prevention of catheter infections requires adherence to strict sterile technique and subsequent meticulous insertion site care. The risk of catheter infection is increased if inserted through infected or contaminated skin (e.g., burns) and these areas should be avoided if at all possible. Femoral vein catheters are more likely to become infected than subclavian or internal jugular vein catheters, and multilumen catheters may also have a higher rate of infection than single lumen catheters, but this may not be clinically significant since multilumen catheters are typically used for short periods of time (19, 20). Some authors have suggested that hyperalimentation solutions be administered only through single lumen catheters, because its use predisposes patients to catheter infection, particularly if the catheter is used for administration of medication as well (21). This approach invariably is impractical in critically ill patients requiring venous access for administration of numerous medications or hemodynamic monitoring and in these cases use of a dedicated hyperalimentation lumen probably is sufficient. Neither use of tunneled or cuffed (with or without antibiotic cuff impregnation) catheters, nor the addition of prophylactic antibiotics while catheters are in place have been demonstrated to reduce infection rates (2226). In a controlled trial of scheduled replacement of central venous and pulmonary artery catheters, routine replacement of catheters (either at a new site or by guidewire exchange) did not decrease infection. Routine guidewire exchange increased infection while placement of lines at new sites increased the rate of technical complications (27, 28).

More recently, several groups of investigators have examined the ability of antibiotic bonding to polyurethane catheters to reduce the incidence of catheter colonization and associated bacteremic episodes. Two studies bear mention, the first of which examined 281 patients in whom 298 triple lumen catheters were placed, 151 of which were routine catheters, 147 of which were minocycline-rifampin bonded on both the intra- and extralumenal surfaces. In the control group, 36 catheters became colonized (26%) and 7 bloodstream infections were identified (5%), while these same rates for patients in whom antibiotic bonded catheters were inserted were 8% (p < .001) and 0% (p < .01), respectively (29). A second, larger trial involved 817 patients who underwent placement of 865 triple lumen polyurethane catheters, either bonded with chlorhexidine/silver sulfadiazine on the external surface only (n = 451) or minocycline-rifampin bonded on both surfaces (n = 414). The former group evidence rates of 23% colonization (n = 87) and 3% bloodstream infections (n = 13), while the rates for the latter group were 8% (n = 28, p < .001)) and 0.3% (n = 1, p < .002), respectively (30). Of note, however, is the observation that a number of these catheters were left in place for long periods of time (up to 28 days) and the added costs of these catheters have yet to be determined. Thus, the indications for the use of these devices remain to be established.

Patients generally can be categorized into several groups in order to guide therapy:


patients with evidence of localized, nonagressive exit site or superficial skin and subcutaneous infection without either positive blood cultures or systemic manifestations,


patients demonstrating presence of positive blood cultures with no external site manifestations of infection and only minimal signs or symptoms that could be related to incipient sepsis syndrome,


patients with an intravascular catheter (with or without evidence of exit site or subcutaneous infection) and evidence of sepsis syndrome, with or without positive blood cultures.

Increasingly, patients in the first two categories are undergoing antibiotic therapy without catheter removal with considerable success. However, if significant purulence is present at the catheter exit site, catheter removal coupled with exploration and drainage of the subcutaneous tunnel or port site using local anesthesia is mandatory to exclude the presence of more aggressive infection. Infected wounds should be left open and packed with gauze. In general, patients in the third category should undergo prompt catheter removal unless no alternative vascular access sites are available and catheter removal itself would be life threatening.

The traditional approach for treating an intravascular catheter infection has been catheter removal, administration of parenteral antibiotic(s) (7–14 day course), and replacement of the catheter at a separate site after a 24–72 hour period, to avoid catheter reinfection curing concurrent bacteremia. However, this approach may be impractical in the critically ill patient with long-term need for central venous access or limited venous access sites. Consequently, a number of approaches have been proposed that do not require immediate catheter removal and replacement at a new site. One approach is to change the potentially infected line over a guidewire and culture the catheter tip. If the catheter tip culture shows that the original catheter was infected, a new catheter can be placed at a fresh site (31, 32). In addition, ~ 80% of catheter infections are due to gram-positive microbes and can be treated with a 10–21 day course of an intravenous antibiotic via the catheter while leaving it in situ (33). Patients who are not candidates for this type of treatment include individuals who:


exhibit evidence of sepsis syndrome from the outlet (category 3, above),


exhibit ongoing evidence of potentially serious infection (persistent bacteremia with or without systemic manifestations or persistent systemic manifestations of infection without a specific site of infection) or


recrudescence soon after completion of an initial treatment.

If a patient's clinical course fails to improve after 24 to 48 hours of antimicrobial therapy, the catheter should be removed. Fungemia and gram-negative bacteremia due to an infected catheter rarely are treatable with antimicrobial agents alone and can become life threatening. In most cases, catheters in these patients should be removed without a trial of antimicrobial agent therapy (34).


Although gram-negative organisms continue to account for up to one third of intravascular catheter infections, gram-positive organisms have become increasingly prevalent pathogens Virulent antibiotic resistant bacterial strains have emerged and present a formidable treatment challenge. Simultaneously, management of catheter infection has evolved. Although patients who develop fungemia, gram-negative bacteremia, or sepsis syndrome are best treated by catheter removal in addition to antimicrobial therapy, an increasing body of evidence suggests that many gram-positive bacterial catheter infections can be treated by use of antimicrobial agents without catheter removal. The initial need for catheter placement, adherence to meticulous sterile surgical technique during insertion, and subsequent fastidious catheter maintenance remain the mainstays of preventing these potentially disastrous infections.


Pittet D, Wenzel R P. Nosocomial bloodstream infections. Arch Intern Med. (1995);155:1177–1184. [PubMed: 7763123]
Banerjee S N, Emori T G, Culver T H. et al. Secular trends in nosocomial primary bloodstream infections in the United States, 1980–1989. Am J Med. (1991);91:SB86–89. [PubMed: 1928197]
Centers for Disease Control. Increase in national hospital discharge survey rates for septicemia - United States. JAMA. (1990);263:937–938. [PubMed: 2299753]
Pittet D, Tarara D, Wenzel R P. Nosocomial bloodstream infection in critically ill patients: excess length of stay, extra costs, and attribuatable mortality. JAMA. (1994);271:1598–1601. [PubMed: 8182812]
Elliott T S J, Faroqui M H, Armstrong R F. et al. Guidelines for good practice in central venous catheterization. J Hosp Infect. (1994);28:163–176. [PubMed: 7852731]
Whitman E D. Complications associated with the use of central venous access devices. Curr Prob Surg. (1996);33:311–378. [PubMed: 8601333]
Burd RS, Cody CS, Dunn DL (1992) Immunotherapy of gram-negative bacterial sepsis. Armstrong Printing Co, Austin, TX .
Bone R C, Balk R A, Cerra F B, Dellinger R P, Fein A M, Knaus W A, Schein R M, Sibbald W J. Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. Chest. (1992);101:1644–1655. [PubMed: 1303622]
Ziegler E, Fisher C, Sprung C. et al. Treatment of gram-negative bacteremia and septic shock with HA-1A human monoclonal antibody against endotoxin. N Engl J Med. (1991);324:429–436. [PubMed: 1988827]
Greenman R, Schein R, Martin M. et al. A controlled clinical trial of E5 murine monoclonal IgM antibody to endotoxin in the treatment of gram-negative sepsis. The XOMA Sepsis Study Group. JAMA. (1991);266:1097–1102. [PubMed: 1865542]
Fisher C J, Dhainaut J F, Opal S M. et al. Recombinant human interleukin 1 receptor antagonist in the treatment of patients with sepsis syndrome. JAMA. (1994);271:1836–1843. [PubMed: 8196140]
Dunn D L. Gram-negative bacterial sepsis and sepsis syndrome. Surg Clin N Amer. (1994);74:621–635. [PubMed: 8197534]
The EORTC International Antimicrobial Therapy Cooperative Group. Ceftazidime combined short or long course of amikacin for empiric therapy of gram-negative bacteremia in cancer patients with granulocytopenia. N Engl J Med. (1987);317:1692–1698. [PubMed: 2892130]
Klastersky J, Glauser M P, Schimpff S C. et al. Prospective randomized comparison of three antibiotic regimens for empiric therapy of suspected bacteremic in fection in febrile granulocytopenic patients. Antimicrob Agents Chemother. (1986);29:263–270. [PMC free article: PMC176388] [PubMed: 3521475]
Jones W G, Barie P S. Enteroccocal infections in surgical patients. Infect Surg. (1988);7:733–738.
Wells C L, Juni B A, Cameron S B, Mason K A, Dunn D L, Ferrieri P, Rhame F S. Stool carriage, clinical isolation, and mortality during an outbreak of vancomycin-resistant enterococci in hospitalized medial and/or surgical patients. Clin Infect Dis. (1995);21:45–50. [PubMed: 7578758]
Rex J H, Bennett J E, Sugar A M. et al. A randomized trial comparing fluconazole with amphotericin B for the treatment of candidemia in patients without neutropenia. N Engl J Med. (1994);331:1325–1330. [PubMed: 7935701]
Flynn P M, Shenep J L, Barrett F F. Differential quantitation with a commercial blood culture tube for diagnosis of catheter-related infection. J Clin Microbiol. (1988);26:1045–1046. [PMC free article: PMC266516] [PubMed: 3384897]
Early T F, Gregory R T, Wheeler J R. et al. Increased infection rate in double lumen versus single-lumen Hickman catheters in cancer patients. South Med J. (1990);83:34–36. [PubMed: 2300831]
Mermel L A, McCormick R D, Springman S R. et al. The pathogenesis and epidemiology of catheter-related infection with pulmonary artery Swan-Ganz catheters: a prospective study using molecular subtyping. Am J Med. (1991);91:SB197–205. [PubMed: 1928165]
Christensen M L, Hancock M L, Gattuso J. et al. Parenteral nutrition associated with increased infection rate in children with cancer. Cancer. (1993);72:256–263. [PubMed: 8402497]
Wagman L D, Kirkemo A, Johnston M R. Venous access: a prospective randomized study of the Hickman catheter. Surgery. (1984);95:303–308. [PubMed: 6701786]
Andrivet P, Bacquer A, Ngoc C V. et al. Lack of clinical benefit from subcutaneous tunnel insertion of central venous catheters in immunocompromised patients. Clin Infect Dis. (1994);18:34–36. [PubMed: 8161627]
Lim S H, Smith M P, Machin S J, Goldstone A H. A prospective randomized study of prophylactic teicoplanin to prevent early Hickman-catheter-related sepsis in patients receiving intensive chemotherapy for haematological malignancies. Eur J Haematol. (1993);54:10–13. [PubMed: 8365459]
Maki D G, Cobb L, Garman J K. et al. An attachable silver-impregnated cuff for prevention of infection with central venous catheters: a prospective randomized multicenter trial. Am J Med. (1988);85:307–314. [PubMed: 3046351]
Flowers R H I, Schwenzer K J, Kopel R F. et al. Efficacy of an attachable subcutaneous cuff for the prevention of intravascular catheter-related infection: a randomized, controlled trial. JAMA. (1989);261:878–883. [PubMed: 2492354]
Cobb D K, High K P, Sawyer R G. et al. A controlled trial of scheduled replacement of central venous and pulmonary-artery catheters. New Engl J Med. (1992);327(15):1062–1068. [PubMed: 1522842]
Shaffer D. Catheter-related sepsis complicating long-term tunneled central venous dialysis catheters: Management by guidewire exchange. Am J Kidney Dis. (1995);25(4):593–596. [PubMed: 7702056]
Raad I, Darouiche R, Dupuis J. et al. Central venous catheters coated with minocycline and rifampin for the prevention of catheter-related colonization and bloodstream infections. Ann Intern Med. (1997);127:267–274. [PubMed: 9265425]
Darouiche R O, Raad I I, Heard S O. et al. A comparison of two antimicrobial-impregnated central venous catheters. N Engl J Med. (1999);340:1–8. [PubMed: 9878638]
Prince A, Heller B, Levy J, Heind W C. Management of fever in patients with central venous catheters. Ped Infect Dis J. (1986);5:20–24. [PubMed: 3080737]
Wang E E L, Prober C G, Ford-Jones L. The management of central intravenous catheter infections. Pediatr Infect Dis J. (1984);3:110–113. [PubMed: 6427760]
Johnson D C, Johnson F L, Goldman S. Preliminary results treating persistent venous catheter infections with the antibiotic lock technique in pediatric patients. Ped Infect Dis. (1994);13:930–931. [PubMed: 7854895]
Bullard K M, Dunn D L. Diagnosis and treatment of bacteremia and intravascular catheter infections. Am J Surg. (1996);172:S13–19. [PubMed: 9003685]
Copyright © 2001, W. Zuckschwerdt Verlag GmbH.
Bookshelf ID: NBK7008


  • PubReader
  • Print View
  • Cite this Page

Related information

  • PMC
    PubMed Central citations
  • PubMed
    Links to PubMed

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...