| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
*Division of General and Surgical Intensive Care Medicine, Department of Anesthesia and Critical Care Medicine, Institute for Hospital Hygienics and Social Medicine, and Institute for Medical Biostatistics, Innsbruck Medical University; Department of Anesthesia and Critical Care Medicine, Krankenhaus der Barmherzigen Schwestern, Ried I. Innkreis, Austria
Address correspondence and reprint requests to Martin W. Dünser, MD, Division of General and Surgical Intensive Care Medicine, Department of Anesthesia and Critical Care Medicine, Innsbruck Medical University, Anichstrasse 35, 6020 Innsbruck, Austria. Address e-mail to Martin.Duenser{at}uibk.ac.at.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Development of CVC colonization is postulated to be caused by one of five mechanisms. Aside from contamination of the catheter at time of insertion, catheter colonization with skin microorganisms along the external surface seems to be the most important pathogenetic mechanism. After 7 to 10 days, endoluminal colonization caused by contaminated catheter hubs or IV fluids becomes more prominent. Additionally, colonization of the catheter from distant sites can lead to secondary bacteremia and sepsis (3).
Aside from maximal sterile barrier precautions during CVC insertion and maintenance, attempts at reducing CVC colonization have included introduction into clinical practice of catheter hubs containing iodinated alcohol, chlorhexidine-impregnated sponge dressings, and antiseptic and antimicrobial-impregnated CVC (4). However, coating CVC with antimicrobial agents has raised concerns regarding the possible development of antimicrobial resistance (4). These and other considerations have resulted in the development of CVC coated with antiseptics that are not used to treat systemic infections. Aside from CVC coating with iodine complexed with polyvinylpyrrolidone or benzalkonium chloride, impregnation of CVC with silver (S) (5) or chlorhexidine-silversulfadiazine (CH-SS) (6) was shown to reduce catheter colonization in in vitro studies and also in the clinical setting.
In this prospective, randomized, controlled clinical study, the efficacy of antiseptic-coated CVC in prevention of catheter colonization was compared with the efficacy of standard, unimpregnated catheters. Additionally, catheter colonization rates of S-impregnated and CH-SS-impregnated CVC were compared.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
In all study patients, CVC were aimed to be placed via the subclavian route. The internal jugular and femoral sites were considered alternative approaches in patients with bilateral subclavian catheters already in place, permanent transvenous pacemakers, or an anatomical situation making central venous cannulation via the subclavian route impossible. After skin disinfection with aqueous chlorhexidine, only noncuffed, multilumen CVC (3 or 4 lumen) were inserted using the Seldinger technique under maximal sterile hygienic precautions (mask, cap, sterile gloves, sterile gown, and sterile drapes). Catheters were not tunneled, but were fixed to the skin with two subcutaneous sutures. Routine inspection and cleaning of the catheter insertion site and dressing with sterile dry gauze were performed once daily or as required if gauzes became soiled or wet. Before CVC manipulation, strict hand disinfection techniques or clean examination gloves were used. CVC were removed if one or more of the following criteria were fulfilled: (a) the catheter was no longer medically indicated, (b) catheter malfunction occurred, (c) signs of local infection at the insertion site developed, (d) bloodstream infection with methicillin-resistant staphylococci spp. was diagnosed without any other clinically obvious infection source, and (e) catheter-related bloodstream infection was suspected because clinical or laboratory signs of infection could not be attributed to other infection sources (sinusitis, bronchopulmonary infections, urinary tract infections, wound infections, etc.). If CVC had to be changed, de novo insertion through a new venipuncture was performed at another insertion site, if possible. No wire-guided change was performed in any study patient.
Anticoagulation with unfractionated heparin or a low molecular weight heparin (enoxaparin) was used in most patients for either prophylactic or therapeutic anticoagulation. All patients received parenteral nutrition consisting of glucose and amino acids with or without fat-containing solutions via the CVC, simultaneous with a progressive increase in enteral nutrition. None of the patients received immunosuppressive treatment or exhibited clinically relevant neutropenia (granulocytes < 500/µL) during the study period.
After study inclusion, patients were randomly assigned to one of the three study groups. Patients in the standard CVC group received a standard, unimpregnated, polyurethane CVC (Arrows International, Erding, Germany). Patients in the S group received a polyurethane CVC externally coated with S (Siemens AG, Erlangen, Germany). Polyurethane catheters impregnated on the external surface with CH-SS (Arrowgard Blue®; Arrows International) were inserted in patients randomized to the CH-SS group. Although new types of S and CH-SS catheters have been introduced into clinical practice, featuring larger concentrations of antiseptics and enhanced bonding to both the internal and external surface (7), as well as a minocycline-rifampin-coated CVC (8), these CVC were not available in Austria during the time of study planning and approval by the ethical committee. Because of the different colors and packages of the individual catheters, doctors and nurses could not be blinded to the type of study catheter.
If patients received more than one catheter at the same time or consecutively, they were newly entered into the randomization process. However, patient data were analyzed on a per-patient basis only after their first inclusion into the study protocol, whereas CVC were analyzed in a per-catheter analysis after each randomization.
CVC colonization was defined as growth of 15 or more colony-forming units in cultures prepared by the roll-plate method or 103 or more colony-forming units in cultures prepared by the sonication method (9). Clinical signs of systemic inflammatory response syndrome (SIRS) or sepsis were defined according to the American Collage of Chest Physicians/Society of Critical Care Medicine criteria.
After randomization, age, sex, body mass index, admission diagnosis, and the site of CVC insertion, the therapeutic intervention severity score (TISS) 28, and the simplified acute physiologic score (SAPS) II were documented in all patients. TISS 28 and SAPS II were calculated from worst physiological and clinical data during the first 24 h after admission to the intensive care unit. Additionally, incidence of SIRS or sepsis, length of intensive care unit stay, and intensive care unit mortality were reported. After catheter removal, CVC indwelling time and colonization of the CVC were documented. In case of catheter colonization, the microbiological organism was cultivated and specified.
All catheters were removed under aseptic conditions. Two distal segments (2–3 cm) were cut off the catheter tip at the bedside, collected in a sterile container, and subsequently transferred to the microbiological laboratory. In all study catheters, for analysis of extraluminal catheter colonization, the distal catheter tip was processed using the semiquantitative roll-plate method by Maki et al. (10). To avoid missing microorganisms progressing intraluminally, quantitative culturing after sonication of the proximal catheter segment was performed (11).
Differential analysis of cultivated microorganisms was performed using morphological, physiological, and serological criteria. Plasma coagulase negative and positive staphylococci were identified using a commercial test kit (Pastorex Staph Plus®, Sanofi Pasteur Diagnostics, New York, NY) and the classical tube coagulation test. Other Gram-positive cocci, Gram-negative rods, and yeasts were identified with specific test kits (API®, Bio Merieux, Marcy Lètoile, France). Aerobic spore-forming bacteria, microscopically examined as Gram-positive rods, were identified because of catalase production and aerobic endospore formation.
The primary end-point of this study was to detect a possible difference in the rate of CVC colonization between antiseptic-impregnated CVC and standard, unimpregnated catheters. The secondary study end-point was to detect a possible difference in the rate of CVC colonization between S-coated CVC and catheters impregnated with CH-SS.
Sample size was precalculated and based on an assumed colonization difference of 10% (15% standard catheter, 5% antiseptic coated catheter). A two-group 
2 test with a 0.05 two-sided significance level has 80% power to detect this difference when the sample size in each group is 141. To account for dropouts, n = 150 per catheter group was prespecified. Randomization of patients was performed using a random number generating scheme.
Continuous demographic and clinical data were analyzed on a per-patient basis. Data with Gaussian distribution (age, body mass index, and length of intensive care unit stay) were compared with independent sample t-tests; non-Gaussian data (TISS 28 and SAPS II) were compared with the Mann-Whitney U-rank sum test. Categorical data (sex, diagnosis, incidence of SIRS or sepsis, and intensive care unit mortality) were compared with the 2 test. To evaluate differences in CVC colonization among groups, a per-catheter analysis was performed using a 2 test. If group effects were significant, pair-wise comparisons among catheter groups were performed using the same statistical test and applying Bonferroni corrections for multiple comparisons. Additionally, relative risk (RR) ratios, together with 95% confidence intervals (CI), were given to indicate the impact of catheter impregnation on CVC colonization. Kaplan-Meier analysis with a log-rank test was used to describe the incidence of catheter colonization in relation to CVC indwelling time. Bonferroni corrections were applied for multiple comparisons between groups.
P values <0.05 were considered to indicate statistical significance. Data are given as mean values ± sd throughout the manuscript if not indicated otherwise.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
|
There was a significant difference in the rate of CVC colonization among study groups (Table 2). CVC colonization and colonization per 1000 catheter days were not different between standard and CH-SS-coated CVC (P = 0.564 and P = 0.24, respectively; RR 0.61, 95% CI, 0.29–1.28). There was also no significant difference in CVC colonization and colonization per 1000 catheter days between standard and S-coated CVC (P= 0.795 and P = 0.639, respectively; RR 1.51, CI 0.8–2.84). CVC colonization and colonization per 1000 catheter days were significantly less with CH-SS-coated CVC when compared to S-impregnated catheters (P = 0.01 and P = 0.003, respectively; RR 0.43, CI 0.21–0.85). CVC indwelling time was significantly longer in the standard group than in the S group (P = 0.009). There was no difference in CVC indwelling times between standard and CH-SS-coated catheters (P = 0.078) or between CH-SS-coated and S-coated catheters (P = 0.434). There was no difference among groups in the number of CVC inserted into different veins and in the type of bacteria cultivated from colonized catheter tips.
|
Figure 1 presents the risk of catheter colonization among groups depending on catheter indwelling time. Colonization rates increased significantly with the time after catheter insertion in all study groups (P < 0.001). Whereas first catheter colonization occurred in standard and S catheters 2 and 3 days after insertion, respectively, no colonization was observed in CH-SS-impregnated CVC during the first 6 days. There was no hypersensitivity or allergic reaction in response to any catheter in any of the study patients throughout the observation period.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The results of this clinical study indicating no difference in catheter colonization rates between antiseptic-coated and standard, unimpregnated CVC are in contrast to published data on this topic (12–14). Whereas some authors could demonstrate a highly significant reduction of CVC colonization rates by the use of CH-SS-coated catheters (up to 90%) (12), our results are in line with studies from Ciresi et al. (15) and Pemberton et al. (16), none of whom found antiseptic catheters beneficial in reducing the incidence of CVC colonization and related bloodstream infections. Moreover, in our study, standard CVC were in place one day longer and could thus be speculated to have been at an even higher risk of colonization than antiseptic catheters.
One explanation for the lack of difference between antiseptic-coated and standard CVC in this study could be the already infrequent catheter colonization in standard, unimpregnated catheters in the present study. Although a highly sensitive combination of two microbiological methods to diagnose catheter colonization was used, the colonization rate of standard catheters (11.8%) was comparable to or even less than colonization rates reported for antiseptic- and antimicrobial-coated CVC in other studies (12,13,17). In addition, with an average indwelling time of 10 days, the CVC in this study were kept in place for a longer period than in most other trials. Despite a mean CVC indwelling time of 9.7 days and a severely ill patient population highly susceptible to CVC colonization and related bloodstream infections, the colonization rate of 7.3% for CH-SS catheters in this study was even less than reported for an antimicrobial catheter coated with minocycline-rifampin left in place for 8.2 days (8). Thus, it is conceivable that in a critical care setting with already infrequent colonization rates, antiseptic-impregnated CVC have less impact on catheter colonization than when colonization rates of standard, unimpregnated catheters are more frequent, as in most previous studies. However, it must also be stated that although the present trial analyzed more catheters than most studies, the number of CVC included might have been too small to detect a significant difference in colonization rates between standard and antiseptic CVC. Although a power analysis was precalculated on the basis of published data before study planning (12,13,18), it assumed a larger reduction of colonization rates than was eventually observed.
Patients in the antiseptic catheter groups received on average more CVC (S, 1.9 catheters per patient; CH-SS, 2.4 catheters per patient) than patients in the standard catheter group (1.3 catheters per patient). Assuming that patients requiring more CVC for more complex hemodynamic monitoring or drug therapy are more severely ill and will stay in the intensive care unit longer, patients receiving antiseptic CVC might have also been more susceptible to catheter colonization than patients in the standard CVC group. This could have contributed to the observation that antiseptic catheter impregnation did not result in less frequent colonization in this trial.
Another interesting finding of this study was the strong time dependence of catheter colonization among study groups. Although catheter colonization significantly increased over time in all study groups, CH-SS-impregnated catheters were colonized for the first time after six days, whereas standard and S-coated CVC had already been colonized after two and three days, respectively. Therefore, it may be speculated that release of the antiseptic substances CH and SS from the catheter impregnation was only large enough to prevent catheter colonization and bacterial biofilm colonization during the first week after CH-SS catheter insertion.
However, these results are, once again, not in line with earlier reports demonstrating release of antiseptic impregnation over approximately 15 days in vivo (19). Bach et al. (20) found that the release of CH-SS followed an exponential curve, with detectable antibacterial activity being present, even 520 hours after catheter insertion. When considering comparable clinical studies investigating catheter colonization rates as main outcome variables, our study results are in accordance with previous reports demonstrating superiority of CH-SS-impregnated CVC when compared with unimpregnated catheters within the first week after insertion (12–14). Accordingly, a systematic review of randomized, controlled trials comparing antiinfective with unimpregnated standard CVC suggested that CH-SS coating is antiinfective for only a short period after insertion (21).
Another explanation for effective antibacterial activity during the first six days after CVC placement is that CH-SS catheters used in this protocol were impregnated only on the external surface. The external surface plays an important pathogenetic role in the development of CVC colonization during the first week, whereas after seven days, colonization is suggested to occur mainly via the intraluminal route (3). This could explain why externally impregnated CH-SS CVC may only prevent catheter colonization during the first week after insertion. Thus, antiseptic catheters might have been in place for too long a time in this study, exceeding their potential to effectively protect against CVC colonization. Accordingly, in a recent study, Brun-Boisson et al. (7) observed that the new generation of CH-SS CVC, which are impregnated with larger concentrations of the antiseptic on both the internal and external surfaces, could significantly reduce the CVC colonization rate, even when CVC were left in place for >10 days.
When comparing the two antiseptic CVC in the present study, the use of S-coated CVC was inferior to CH-SS CVC with regard to catheter colonization. Synergistic effects of the broad-spectrum germicide CH and the antiseptic SS against nearly all nosocomially transmitted bacteria and yeasts have been repeatedly demonstrated in vitro and in vivo (6). Data on the use of S-coated catheters are conflicting. Several studies, also including critically ill patients, failed to demonstrate a reduction in both catheter colonization and catheter-related bloodstream infection rates with S-coated catheters when compared with standard, uncoated CVC (22). Only in selected patient populations (e.g., oncologic and pediatric patients) (23) and after special processing using a new technique to incorporate S into the polyurethane structure of the CVC are there positive reports on the decrease of the magnitude of CVC colonization and catheter-related bloodstream infections. In an in vitro study, Gaonkar and Modak (24) found that S impregnation was ineffective in preventing catheter colonization with different staphylococci specimens. The significant portion of staphylococci isolates in this analysis, together with the fact that external coating with S cannot prevent intraluminal catheter infection, could explain the inefficacy of S-coated CVC to reduce catheter colonization when compared with CH-SS-coated CVC or standard catheters.
When interpreting the results of this study, an important limitation must be considered. Whereas 275 patients were included into the study, 485 CVC were inserted, meaning that some patients received more than one CVC and also more than one type of CVC. Thus, carryover effects from one CVC to the other cannot be excluded. However, according to the study end-point to detect a possible difference in CVC colonization, this should not be a limiting factor. However, together with the small sample size of study patients, it precludes meaningful interpretation of the results with regard to CVC-related bloodstream infections. There are data in the literature showing a correlation between CVC colonization and catheter-related bloodstream infections (r = 0.69; r2 = 0.48; P < 0.001) (25). Rijinders et al. (25) suggest that the rate of catheter-tip colonization is a good surrogate end-point for the incidence of catheter-related bloodstream infection in the clinical setting.
In conclusion, antiseptic-impregnated CVC could not prevent catheter colonization when compared with standard polyurethane catheters in a critical care setting with infrequent catheter colonization rates and CVC left in place for >10 days.
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
