Effect of a Second-Generation Venous Catheter Impregnated with Chlorhexidine and Silver Sulfadiazine on Central CatheterRelated Infections

Context Bacterial colonization of central venous catheters is relatively common, and subsequent bacteremia is a serious iatrogenic complication of critical illness. Initial studies of antimicrobial-coated catheters have suggested that this approach might decrease catheter-associated infection. Contribution This randomized, double-blind, controlled study of a new antiseptic-coated catheter versus an uncoated catheter shows a substantial decrease in bacterial colonization in patients receiving the coated device. Caution The study was unable to show a substantial decrease in bloodstream infections, possibly because of the low infection rate as a result of meticulous aseptic techniques used during catheter insertion. The Editors Infections associated with central venous catheters are a substantial problem. Each year in the United States, at least 80 000 patients in intensive care units experience central venous catheterassociated bacteremia (1, 2). These infections are associated with an overall attributable mortality of approximately 3% (3), but estimates vary from 0% to greater than 30% depending on patient population, definitions, and pathogens (4). The attributable cost per infection ranges from $3240 to more than $50 000 (5-8). Many strategies have been used to prevent catheter-associated infection. These measures can be divided into 2 groups: those that prevent microbes from gaining access to the catheters and those that discourage microbes from adhering and proliferating on the catheter, such as coating the catheters with various antimicrobial agents. The latter approach has shown promise and has included the use of chlorhexidine and silver sulfadiazine. In a randomized clinical trial, Maki and colleagues (9) observed a statistically significant decrease in colonization and bacteremia in patients who received a catheter coated with chlorhexidine and silver sulfadiazine compared with controls who received an uncoated catheter. In a randomized, comparative trial, Darouiche and colleagues (10) found that catheters impregnated with minocycline and rifampin were associated with fewer infectious complications than catheters coated with chlorhexidine and silver sulfadiazine. However, one of the main differences between the catheters was that the chlorhexidinesilver sulfadiazine coating involved only the external surface of the catheter, whereas the minocycline and rifampin catheter was coated on the internal and external surfaces. More recently, a second-generation antiseptic catheter was formulated that increased the chlorhexidine concentration on the external surface of the catheter 3-fold and incorporated chlorhexidine on the luminal surface of the catheter, extension lines, and hubs. This trial was conducted to assess the efficacy and safety of the second-generation antiseptic catheter compared with an uncoated control catheter. Methods Patients and Study Design This study was a randomized, double-blind, controlled trial conducted between July 1998 and June 2001 at 9 university-affiliated hospitals. The objective was to determine whether the second-generation antiseptic central venous catheter was effective in preventing microbial colonization and bloodstream infection in comparison with an uncoated control catheter. The null hypothesis was that the incidence of bloodstream infection would be the same or worse for the patients who received the antiseptic catheter compared with the patients who received the control catheter. Secondary goals consisted of product safety evaluation, assessment of the microbiology of catheter-associated infection, and microbial susceptibility to chlorhexidine and silver sulfadiazine. The institutional review boards at each hospital approved the protocol. Adult patients who were cared for in critical care units and who required a triple-lumen central venous catheter were eligible for participation. Patients who were pregnant, were allergic to chlorhexidine or sulfa drugs, were hospitalized for burn injuries, had a chronic inflammatory skin disorder at the catheter insertion site, were suspected of having a catheter-associated infection, or were enrolled in another investigational trial were not eligible for participation. All patients or their authorized surrogates granted informed consent. The study sample size was calculated on the basis of an expected catheter-related bloodstream infection rate of approximately 4.5% in the control group and 1.5% in the antiseptic catheter group. Allowing for a 12% dropout rate, 793 patients were required to yield a study with an 80% power at the 0.05 level of statistical significance. Catheters All catheters were 7-French, 20-cm long polyurethane triple-lumen central venous catheters manufactured by Arrow International, Inc. (Reading, Pennsylvania). Control catheters were standard, uncoated triple-lumen catheters. Antiseptic catheters (ARROWgard II Blue Plus, Arrow International, Inc.) were coated with chlorhexidine acetate and silver sulfadiazine on the external surface and chlorhexidine and chlorhexidine acetate on the luminal surfaces. All catheters were indistinguishable in appearance and packaging. Randomization, Catheter Insertion, and Care Procedures Patients were randomly assigned to receive an individually numbered catheter and had an equal probability of assignment to either group. The randomization code was developed by using a computerized random-number generator to select permuted blocks. The block length was 4. Randomization stratification ensured that antiseptic and control catheters were evenly distributed in the de novo and guidewire exchange groups. Patients were randomly assigned in a 1:1 ratio within each of the study centers. Catheter allocation was concealed, and patients, study personnel, and all health care workers were unaware of whether the catheters were coated or uncoated. A subset of patients at each institution (approximately one third of patients) was allowed to receive an initial study catheter through guidewire exchange. Four institutions were also allotted a small number of exchange insertions in which a study catheter could be exchanged for a matched study catheter (randomization and blinding were protected). Figure 1 shows the distribution of patients. Catheters were inserted by using full sterile barrier precautions, which included the operators wearing a sterile long-sleeve gown, sterile gloves, hat, and mask, and using a large sterile drape. Before insertion, the skin was cleansed with 10% povidone-iodine (chlorhexidine-based antiseptics were not approved by the U.S. Food and Drug Administration for insertion-site preparation). Before a study catheter was inserted over a guidewire into a preexisting site, the hub of the first catheter was cleansed with povidone-iodine. The tip of the preexisting catheter was submitted for microbiological testing. Insertion sites were dressed with a transparent polyurethane dressing (OpSite 3000, Smith & Nephew, Inc., Largo, Florida). No antimicrobial ointment was applied at the insertion site. Depending on institutional routine, dressings were changed every 72 to 96 hours using a standardized kit. At the time of dressing change, the insertion site was cleansed with povidone-iodine. The patient's attending physician made the decision to remove the catheter. Figure 1. Distribution of initial study catheters by type and method of insertion. Measurements and Definitions At the time of catheter insertion, the following data were recorded: patient demographic characteristics, indication for catheter insertion, underlying medical conditions, indication for admission to the intensive care unit, length of hospital stay and length of intensive care unit stay, and severity of illness score (Acute Physiology and Chronic Health Evaluation [APACHE] II score). Study catheters were inspected daily. Local and systemic signs and symptoms of infection were recorded. The presence of other intravascular and indwelling devices was noted, and the antibiotics that were administered were recorded. At the time of catheter removal, a 20-cm2 circular template was placed at the catheter insertion site and a moistened swab (Culturette, Becton Dickinson and Co., Sparks, Massachusetts) was used to sample the pericatheter insertion site. The swab was sent to the institutional microbiology laboratory, where it was used to inoculate a blood agar plate. Catheters were removed by using an aseptic technique. The subcutaneous portion of the catheter was cut from the rest of the catheter, and the 2 portions were placed in separate sterile plastic bags for transport to the laboratory. The subcutaneous portion of the catheter was divided into four 2.5-cm segments. A neutralizing media (D/E Neutralizing Broth or Agar, Remel, Lenexa, Kansas) was used to minimize any potential antimicrobial carryover effect. Proximal and distal segments were cultured by using the roll-plate method (11), and similarly, proximal and distal segments were cultured by using a sonication technique (12). At 2 centers, the catheter hubs were cultured by using moistened swabs. Blood cultures were obtained from the catheter and from a peripheral vein on any patient with suspected catheter-associated infection. Signs and symptoms of a catheter-associated infection included fever (temperature > 38 oC) without another obvious source and local signs of infection, such as erythema, cellulitis, purulent drainage, or excessive tenderness. All microbes recovered from cultures of the patient's blood, catheter, skin, or other sites were shipped to a central laboratory (University of Iowa, Iowa City, Iowa) for confirmatory identification and susceptibility testing. Catheters were defined as colonized if cultures revealed at least 15 colony-forming units per segment by the roll-plate method or at least 100 colony-forming units per segment by the sonication method. Catheter-related bloodstream infection was defined as catheter colonization with positive blo