|
|
| |||||||||||||||||||||||||||||||||||||||||
Background The use of central venous catheters impregnated with either minocycline and rifampin or chlorhexidine and silver sulfadiazine reduces the rates of catheter colonization and catheter-related bloodstream infection as compared with the use of unimpregnated catheters. We compared the rates of catheter colonization and catheter-related bloodstream infection associated with these two kinds of antiinfective catheters.
Methods We conducted a prospective, randomized clinical trial in 12 university-affiliated hospitals. High-risk adult patients in whom central venous catheters were expected to remain in place for three or more days were randomly assigned to undergo insertion of polyurethane, triple-lumen catheters impregnated with either minocycline and rifampin (on both the luminal and external surfaces) or chlorhexidine and silver sulfadiazine (on only the external surface). After their removal, the tips and subcutaneous segments of the catheters were cultured by both the roll-plate and the sonication methods. Peripheral-blood cultures were obtained if clinically indicated.
Results Of 865 catheters inserted, 738 (85 percent) produced culture results that could be evaluated. The clinical characteristics of the patients and the risk factors for infection were similar in the two groups. Catheters impregnated with minocycline and rifampin were 1/3 as likely to be colonized as catheters impregnated with chlorhexidine and silver sulfadiazine (28 of 356 catheters [7.9 percent] vs. 87 of 382 [22.8 percent], P<0.001), and catheter-related bloodstream infection was 1/12 as likely in catheters impregnated with minocycline and rifampin (1 of 356 [0.3 percent], vs. 13 of 382 [3.4 percent] for those impregnated with chlorhexidine and silver sulfadiazine; P<0.002).
Conclusions The use of central venous catheters impregnated with minocycline and rifampin is associated with a lower rate of infection than the use of catheters impregnated with chlorhexidine and silver sulfadiazine.
Methods
Patients
The trial was conducted between December 1995 and July 1997 in 12 university-affiliated hospitals. The study was approved by the appropriate institutional review boards. Hospitalized adults who were at high risk for catheter-related infection (such as patients in intensive care units or those who were immunocompromised) and were likely to require a central venous catheter for three or more days were eligible for the study. Pregnant women and patients with a history of allergy to any of the antimicrobial agents used for impregnating the catheters were excluded. All enrolled patients or their legal guardians gave informed consent.
Catheters
Patients were randomly assigned to undergo implantation of 7-French, 20-cm-long, noncuffed, triple-lumen polyurethane central venous catheters impregnated with either minocycline and rifampin (Cook Spectrum, Cook Critical Care, Bloomington, Ind.) or chlorhexidine and silver sulfadiazine (Arrowguard Blue, Arrow International, Reading, Pa.). Both types of catheter are available for clinical use; the retail prices of a catheter tray are $70 and $61, respectively. The catheters impregnated with minocycline and rifampin provided antimicrobial activity on both the external and the internal surfaces. Examination by high-performance liquid chromatography showed that these catheters contained higher amounts of minocycline and rifampin (11.08 and 10.50 mg per catheter, respectively) than those previously studied (2.79 and 0.28 mg per catheter, respectively).2,5 In contrast, only the external surface of catheters impregnated with chlorhexidine and silver sulfadiazine (0.75 and 0.70 mg per catheter, respectively) provided antimicrobial activity.3 All study catheters were sterilized with ethylene oxide before use.
Randomization
A special randomization scheme was used to help match the two study groups closely. Catheter trays wrapped in identical folders were randomly assigned in blinded fashion according to computer-generated identification numbers, in blocks of six (three from each group), so that the catheter trays would be removed from the box one at a time in the prescribed, random order from the top to the bottom. Blocks of six catheters were then shipped to the participating hospitals for assignment to specified patient-care units. In each case, the patients, nurses, physicians, and principal investigators who assessed the outcomes in each hospital were unaware of the type of catheter inserted.
Insertion and Maintenance of Catheters
Attending physicians, house staff, or supervised medical students inserted the catheters into the subclavian, jugular, or femoral vein using maximal sterile-barrier precautions. To avoid the potential confounding effect of the controversial practice of catheter exchange over a guide wire, we determined at the outset of the trial to study only catheters inserted through a new venipuncture. Randomly selected study catheters could be inserted subsequently at new sites in the same patient, so long as that patient had only one study catheter at a time. At the time of catheter insertion and at each dressing change, the insertion site was disinfected with 10 percent povidone–iodine. The dressing was changed and the insertion site was inspected three times a week. Coordinators at each study location evaluated patients daily until the catheter was removed. The decision to remove the catheter was made solely by the patient's physician, who kept the catheter in place until it was no longer needed or until an adverse event, such as catheter-related infection or catheter occlusion, necessitated its removal.
Cultures
Four-centimeter segments from the tips and subcutaneous sections of the aseptically removed catheters were cultured by the roll-plate method,6 then cultured by the sonication method.7 To help identify the sources of organisms that colonize catheters, swab cultures of surrounding skin were obtained at the times of catheter insertion and catheter removal in four participating hospitals. In patients in whom catheter-related infection was suspected on clinical grounds, one or more peripheral-blood samples for culture were collected before or immediately after catheter removal. Recovered organisms were identified by standard microbiologic methods.
Molecular Typing
Bacterial isolates from cultures of blood, catheters, and, when available, skin of patients in whom catheter-related bloodstream infection was diagnosed were typed by genomic fingerprinting with the use of the repetitive-element polymerase chain reaction.8 If the same bacterial species was isolated from different sites in a single patient, DNA-fingerprint patterns were compared for similarity by visual inspection of band patterns and by computer-assisted analysis (RFLPscan Plus, Scanalytics, Billerica, Mass.). Bacterial isolates were considered similar if fingerprint patterns differed by no more than one amplification band.
Antimicrobial Susceptibility
To help determine whether these antimicrobial-impregnated catheters increase the likelihood of the emergence of antibiotic resistance, we compared the minimal inhibitory concentrations and minimal bactericidal concentrations for bacteria isolated from the two kinds of catheters by a standard broth-microdilution assay.9
Definitions
We adopted the definitions of catheter colonization and infection proposed by the Centers for Disease Control and Prevention10,11 and used in previous clinical trials.2 Catheter colonization was defined as the growth of 15 or more colony-forming units in culture of catheter segments prepared by the roll-plate method or 1000 or more colony-forming units in cultures prepared by the sonication method from either the tip or a subcutaneous segment of the catheter. Catheter-related bloodstream infection was defined as the isolation of the same organism (i.e., the same species with identical antimicrobial susceptibility) from the colonized catheter and from peripheral blood in a patient with clinical manifestations of sepsis and no other apparent source of bloodstream infection.
Statistical Analysis
Before undertaking this study, we estimated the number of catheters that would be required for an adequate examination of the hypothesis that catheters impregnated with minocycline and rifampin are significantly less likely to be colonized than catheters impregnated with chlorhexidine and silver sulfadiazine. On the basis of previous reports, we estimated that 7 percent of catheters impregnated with minocycline and rifampin2 and 13.6 percent of catheters impregnated with chlorhexidine and silver sulfadiazine3 would be colonized. Randomly assigning approximately 362 catheters that could be evaluated to each group would have allowed us to detect with 80 percent power a significant difference in the rates of colonization between the two types of catheters at a two-tailed significance level of 5 percent.
The significance of the differences between the two study groups was determined with use of Student's t-test or the Wilcoxon rank-sum test for continuous variables and Fisher's exact test or the chi-square test for categorical variables. All P values were based on two-tailed tests of significance. The proportions of catheters that were free of colonization and not associated with bloodstream infection as a function of the length of time they had been in place were compared between the groups with use of a log-rank test on Kaplan–Meier estimates. A multivariate logistic-regression model was used to estimate the simultaneous effects of multiple variables on the incidence of catheter colonization and catheter-related bloodstream infection. To avoid rejecting variables that might have influenced the risk of catheter colonization or catheter-related bloodstream infection, variables that were significant at a P value of 0.25 or less in the univariate analysis were entered in stepwise fashion into logistic-regression analyses and tested for an independent effect. The limit for entering or removing variables in the logistic-regression models was a P value of 0.05 or less. All computations were performed with SAS/STAT software.12 An independent monitoring board composed of experts on infectious diseases reviewed and helped interpret the findings of the study. An interim analysis of the data was not performed.
Results
Characteristics of Patients and Catheters
A total of 865 study catheters (414 impregnated with minocycline and rifampin and 451 impregnated with chlorhexidine and silver sulfadiazine) were inserted into 817 patients. Complete data could be evaluated for 738 catheters (85 percent): 356 impregnated with minocycline and rifampin and 382 impregnated with chlorhexidine and silver sulfadiazine, inserted in 698 patients. The remaining 127 catheters (58 impregnated with minocycline and rifampin and 69 impregnated with chlorhexidine and silver sulfadiazine, with similar patient and catheter characteristics) were not cultured (84 were removed without notification of study coordinators, 19 were grossly contaminated during removal, and 24 were not available for other reasons) and therefore were excluded from further analysis. The two groups of catheters that could be evaluated were similar with respect to characteristics of patients and catheters (Table 1).
|
Eighty-seven of 382 catheters impregnated with chlorhexidine and silver sulfadiazine (22.8 percent) and 28 of 356 catheters impregnated with minocycline and rifampin (7.9 percent) were colonized according to at least one method of assessment (relative risk, 2.90; 95 percent confidence interval, 1.94 to 4.33; P<0.001). Catheters impregnated with minocycline and rifampin were less likely to be colonized than those impregnated with chlorhexidine and silver sulfadiazine, whether the catheter remained in place for seven days or less (13 of 217 catheters [6.0 percent] vs. 45 of 210 [21.4 percent], P<0.001) or for more than seven days (15 of 139 [10.8 percent] vs. 42 of 172 [24.4 percent], P<0.002). Analysis of the Kaplan–Meier estimates of the risk of catheter colonization according to the length of time the catheters were in place in each group showed that catheters impregnated with minocycline and rifampin were significantly less likely to be colonized (P<0.001 by the log-rank test). The overall beneficial effect of the use of catheters impregnated with minocycline and rifampin was seen in all hospitals that contributed more than 32 catheters that could be evaluated. Catheters impregnated with minocycline and rifampin were also significantly less likely to be colonized than catheters impregnated with chlorhexidine and silver sulfadiazine (P<0.001) according to each of the four combinations of catheter segment and culture method (tip–roll plate, tip–sonication, subcutaneous segment–roll plate, and subcutaneous segment–sonication) or any combination of these assessment methods.
Catheters impregnated with chlorhexidine and silver sulfadiazine were significantly more likely than those impregnated with minocycline and rifampin to be colonized with coagulase-negative staphylococci (18 percent vs. 4 percent; relative risk, 4.16; 95 percent confidence interval, 2.42 to 7.14; P<0.001), gram-positive bacilli (2 percent vs. 0.3 percent; relative risk, 7.46; 95 percent confidence interval, 0.94 to 58.8; P=0.04), or gram-negative bacilli (4 percent vs. 1 percent; relative risk, 3.96; 95 percent confidence interval, 1.35 to 11.63; P=0.007). However, the rates of colonization of catheters with Staphylococcus aureus (1 percent vs. 0), enterococci (2 percent vs. 2 percent), and yeast (2 percent vs. 3 percent) did not differ significantly between the two groups.
Factors that may have increased the likelihood of catheter colonization (detected by any of the assessment methods) in the univariate analysis (P
0.25) were entered into a multivariate logistic-regression model, which identified the following predisposing factors as significant (P0.05): insertion of the catheter into the femoral or jugular vein (odds ratio as compared with other locations, 3.05; 95 percent confidence interval, 1.86 to 5.01; P<0.001), use of a catheter impregnated with chlorhexidine and silver sulfadiazine (odds ratio as compared with minocycline and rifampin, 2.80; 95 percent confidence interval, 1.68 to 4.66; P<0.001), hospitalization in the intensive care unit (odds ratio as compared with other wards, 2.60; 95 percent confidence interval, 1.47 to 4.62; P=0.001), male sex (odds ratio, 2.45; 95 percent confidence interval, 1.43 to 4.20; P=0.001), and mechanical ventilation (odds ratio, 1.97; 95 percent confidence interval, 1.14 to 3.41; P=0.01).
Catheter-Related Bloodstream Infection
In 14 cases, bloodstream infection was attributed to an indwelling study catheter. These catheters had been in place for a median of 11 days. Thirteen cases of catheter-related bloodstream infection occurred among the catheters impregnated with chlorhexidine and silver sulfadiazine (3.4 percent), as compared with one case among the catheters impregnated with minocycline and rifampin (0.3 percent; relative risk, 12.05; 95 percent confidence interval, 1.59 to 90.9; P<0.002). Two patients died as a result of bloodstream infections associated with catheters impregnated with chlorhexidine and silver sulfadiazine. Among the catheters that remained in place for more than seven days, the rate of associated bloodstream infection was significantly higher for catheters impregnated with chlorhexidine and silver sulfadiazine than for catheters impregnated with minocycline and rifampin (11 of 172 catheters [6.4 percent] vs. 1 of 139 [0.7 percent], P=0.01). The rates of catheter-related bloodstream infection per 1000 catheter-days were 0.3 (95 percent confidence interval, 0.01 to 1.85) for catheters impregnated with minocycline and rifampin and 4.1 (95 percent confidence interval, 2.22 to 6.99) for catheters impregnated with chlorhexidine and silver sulfadiazine (P<0.001). Figure 1 shows the Kaplan–Meier estimates of the risk of catheter-related bloodstream infection according to duration of catheterization in each group and shows that catheters impregnated with minocycline and rifampin were superior (P=0.001 by log-rank test). The same conclusion was reached when we considered only the results from culture of the catheter tip (1 infection among 356 catheters [0.3 percent] vs. 11 among 382 [2.9 percent], P=0.006) or the subcutaneous segment (1 of 356 [0.3 percent] vs. 12 of 382 [3.1 percent], P=0.003).
|
The clonal relation of isolates from blood and catheter cultures was confirmed by DNA typing in 13 of 14 cases (Figure 2). Skin swabs were cultured at the time of the removal of the catheter from seven patients with catheter-related bloodstream infection. In five of these cases (71 percent), the culture yielded bacteria of the same species with a DNA-fingerprint pattern similar to that of the isolates from the catheter and the blood; a different organism grew from skin cultures in the other two patients (29 percent), suggesting that catheter infection may have originated from contamination of the catheter hub.
|
0.25 as the criterion) were entered into a multivariate logistic-regression model, which identified the following predisposing factors as significant (P0.05): catheterization for more than seven days, use of a catheter impregnated with chlorhexidine and silver sulfadiazine, and male sex (Table 2).
|
Antimicrobial Susceptibility
The ranges of the minimal inhibitory concentrations and minimal bactericidal concentrations of minocycline and rifampin for S. epidermidis and enterococci were similar for isolates cultured from the two types of catheters (Table 3). Moreover, in the two cases in which the same organism was isolated from paired cultures of skin obtained before insertion and at the time of removal of a catheter impregnated with minocycline and rifampin (S. epidermidis in one case and enterococcus in the other), the minimal inhibitory concentrations and minimal bactericidal concentrations of minocycline and rifampin for the corresponding paired isolates were similar.
|
There were no local or systemic hypersensitivity reactions associated with the use of either catheter.
Discussion
Recent comparative studies have shown that the use of central venous catheters impregnated either with minocycline and rifampin2 or with chlorhexidine and silver sulfadiazine3 is associated with lower rates of catheter colonization and bloodstream infection than the use of unimpregnated catheters. Although three smaller clinical trials (which studied 72, 282, and 308 catheters that could be evaluated)13,14,15 showed a nonsignificant trend toward lower rates of bloodstream infection with catheters impregnated with chlorhexidine and silver sulfadiazine than with unimpregnated catheters, none had sufficient power to determine that there were no differences. We compared two very differently prepared antiinfective catheters. As we hypothesized, our findings indicated that the antiinfective efficacy of catheters impregnated with minocycline and rifampin was superior to that of catheters impregnated with chlorhexidine and silver sulfadiazine.
The majority of cases of catheter-related bloodstream infection are associated with the short-term use of noncuffed central venous catheters.16 On average, 5 percent2,3,14,15 of the 3 million short-term, unimpregnated central venous catheters that are inserted annually in the United States lead to bloodstream infection,16 resulting in about 150,000 cases of catheter-related bloodstream infection a year. Our findings of remarkably low rates of catheter-related bloodstream infection (0.3 percent) and catheter colonization (7.9 percent) associated with the use of catheters impregnated with minocycline and rifampin are similar to previously reported rates.2 However, we found rates of catheter colonization (22.8 percent) and bloodstream infection (3.4 percent) associated with the use of catheters impregnated with chlorhexidine and silver sulfadiazine that were higher than those reported by Maki and colleagues (13.5 percent and 1 percent, respectively).3 The differences in rates of colonization of catheters could be attributed, at least in part, to our use of roll-plate and sonication cultures of both the tips and subcutaneous segments, as compared with the use by Maki et al. of only roll-plate culture of the catheter tips alone.3 As in other reports,7,17 the roll-plate culture had a limited sensitivity for the diagnosis of catheter colonization (78 of 115 catheters [68 percent]) and catheter-related bloodstream infection (12 of 14 [86 percent]) in our study.
Unlike catheters impregnated with minocycline and rifampin, in which antimicrobial activity is present on both the external and the internal surfaces of the catheter, the antimicrobial activity of catheters impregnated with chlorhexidine and silver sulfadiazine is limited to the external surface. The difference might be an important determinant of the difference in efficacy between these two antimicrobial-impregnated catheters. For instance, catheters impregnated with chlorhexidine and silver sulfadiazine reduced colonization of the external surface as compared with uncoated catheters in studies in which the roll-plate method alone was used to culture only the catheter tips2 or both the catheter tips and the subcutaneous segments.14,15 Our use of sonication cultures that retrieve organisms from both the external and internal surfaces is justified by the role of luminal colonization in causing catheter-related bloodstream infection.18 Other factors that may have contributed to the superior efficacy of the catheters impregnated with minocycline and rifampin include the particular method used to incorporate the antimicrobial agents into the catheter material and the resulting concentration and availability of those agents on the catheter surface.
Although antimicrobial resistance is an issue of potential concern, we and others2 have, so far, found a very low likelihood that antibiotic resistance will result from the use of antimicrobial-impregnated catheters. However, continued surveillance for resistance is required as part of the further clinical use of such catheters. Although it is possible that the use of any antiinfective catheter that reduces ultrastructural colonization may decrease the likelihood of the development of resistance to systemically administered antibiotics such as vancomycin,19 the actual effects of the use of antimicrobial-impregnated catheters on infection-control measures require further evaluation.
In conclusion, our results demonstrate that the capacity of catheters impregnated with minocycline and rifampin to resist infection is superior to that of catheters impregnated with chlorhexidine and silver sulfadiazine, particularly in patients who require vascular access for seven or more days. Despite their proven efficacy, antimicrobial-impregnated catheters should complement rather than replace adequate aseptic practices.10,11
Supported by funds from Cook Critical Care, Bloomington, Indiana; the Department of Veterans Affairs, Washington, D.C.; and the University Cancer Foundation at the University of Texas M.D. Anderson Cancer Center, Houston.
Impregnation of catheters with minocycline and rifampin is described in two patents that are the property of Baylor College of Medicine and the University of Texas M.D. Anderson Cancer Center, Houston. Dr. Darouiche (an employee of Baylor College of Medicine) and Dr. Raad (an employee of the University of Texas M.D. Anderson Cancer Center) are coinventors of the two patented methods. Both patents were licensed by Cook Critical Care, Bloomington, Indiana, with royalty rights to Baylor College of Medicine and the University of Texas M.D. Anderson Cancer Center. The inventors receive a percentage of the royalties according to the official policies of each academic institution. None of the authors, including Dr. Darouiche and Dr. Raad, have other financial links to Cook Critical Care or other catheter-manufacturing companies.
We are indebted to Daniel M. Musher, M.D., and Gerald P. Bodey, M.D., for serving on the study monitoring board and for their critical review of the manuscript.
Source Information
From the Departments of Medicine (R.O.D., R.L.H.), Physical Medicine and Rehabilitation (R.O.D.), Family and Community Medicine (J.I.T.), and Anesthesiology (O.C.W.), Baylor College of Medicine and Veterans Affairs Medical Center, Houston; the Department of Medical Subspecialties, University of Texas M.D. Anderson Cancer Center, Houston (I.I.R., H.H., R.H.); the Department of Anesthesiology, University of Massachusetts Medical Center, Worcester (S.O.H.); the Departments of Surgery (A.G.) and Medicine (J.B.), University of Florida College of Medicine, Gainesville; the Department of Medicine, Southern Illinois University School of Medicine, Springfield (N.K.); and the Department of Medicine, University of Texas Medical Branch, Galveston (G.M.). Presented in part as an abstract (LB-22) at the 37th Interscience Conference on Antimicrobial Agents and Chemotherapy, Toronto, September 28–October 1, 1997.
Address reprint requests to Dr. Darouiche at the Veterans Affairs Medical Center, Infectious Disease Section (Rm. 4B-370), 2002 Holcombe Blvd., Houston, TX 77030, or at darouiche.rabih.o{at}houston.va.gov.
References
In addition to the authors, the following members of the Catheter Study Group participated in the clinical trial: C. Robertson, M. Wall, J. Jones, M. Mansouri, C. Stewart, and S. Dunbar (Baylor College of Medicine, Houston); J. Dupuis, A. Buzaid, K. Price, A. El-Rahwan, J. Abbas, and S. Sidarous (University of Texas M.D. Anderson Cancer Center, Houston); I. Toth, K. Longtine, and A. Breuggemann (University of Massachusetts Medical Center, Worcester); K. Rand (University of Florida College of Medicine, Gainesville); S. Bjornson (University of Cincinnati Medical Center, Cincinnati); and P. Falk (University of Texas Medical Branch, Galveston).
| |||||||||||||||||||||||||||||||||||||||||
Related Letters:
Antimicrobial-Impregnated Central Venous Catheters
Paterson D. L., Bach A., Maury E., Offenstadt G., Yasukawa T., Fujita Y., Sari A., Darouiche R. O., Raad I. I.
Extract |
Full Text
N Engl J Med 1999;
340:1761-1762, Jun 3, 1999.
Correspondence
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2008 Massachusetts Medical Society. All rights reserved. |
