Current issues of ACP Journal Club are published in Annals of Internal Medicine


A catheter coated with minocycline and rifampin reduced bloodstream infections

ACP J Club. 1998 Mar-April; 128:41. doi:10.7326/ACPJC-1998-128-2-041

Related Content in this Issue
• Companion Abstract and Commentary: Antiseptic catheters reduced catheter-related infections and were well tolerated

Source Citation

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 Aug 15;127: 267-74. [PubMed ID: 9265425]



To determine the efficacy of a catheter coated with minocycline and rifampin in preventing catheter colonization and catheter-related bacteremia.


Randomized, double-blind, controlled trial.


5 university hospitals in the United States.


281 hospitalized patients ≥ 18 years (mean age 57 y, 60% men) who required a central venous catheter. Patients who were pregnant, allergic to rifampin or tetracycline, had dermatitis or a burn at the insertion site, or were anticipated to be catheterized for < 3 days were excluded.


298 triple-lumen, polyurethane, central venous catheters (147 pretreated with tridodecylmethyl-ammonium chloride and coated with minocycline and rifampin, and 151 untreated and uncoated) were allocated to patients. Catheters were inserted at a new site (subclavian, internal jugular, or femoral vein) after skin disinfection with povidone-iodine or chlorhexidine gluconate. Maximal barrier precautions were used without guidewire exchange. The insertion site was disinfected every 72 hours and covered with sterile gauze. Catheters that were removed without subsequent culture (n = 32) were excluded.

Main outcome measures

Catheter colonization measured by quantitative cultures and bacteremias.

Main results

Colonization occurred in fewer of the coated catheters (n = 11) than in the uncoated catheters (n = 36; 8% vs 26%, P < 0.001). Staphylococcus epidermidis alone or with other organisms colonized fewer coated catheters (n = 3) than uncoated catheters (n = 25; 2% vs 18%, P < 0.001), as did gram-positive organisms alone or with other organisms (n = 4 vs n = 31; 3% vs 23%, P < 0.001). Bacteremia was not associated with any of the coated catheters compared with 7 uncoated catheters (0% vs 5%, P < 0.01). 6 of the 7 cases of bacteremia were caused by Staphylococcus epidermidis, and 1 was caused by Enterococcus faecalis.


Catheters coated with minocycline and rifampin reduced the risk for catheter-related colonization and bloodstream infections.

Sources of funding: The University Cancer Foundation; The University of Texas M.D. Anderson Cancer Center; Cook Critical Care (catheters).

For article reprint: Dr. I. Raad, Section of Infection Control, Box 47, The University of Texas, M.D. Anderson Cancer Center, 151 Holcombe Boulevard, Houston, TX 77030, USA. FAX 713-792-8233.

Table. Catheters coated with minocycline and rifampin (coated catheters) vs untreated catheters control catheters)*

Outcomes Coated catheters Control catheters RRR (95% CI) NNT (CI)
All catheter colonization 8% 18% 68% (41 to 83) 6 (4 to 11)
Colonization with Staphylococcus epidermidis alone or with other organisms 2% 18% 87%(62 to 96) 6 (4 to 11)
Colonization with gram-positive organisms alone or with other organisms 3% 23% 87% (65 to 95) 5 (4 to 8)
Bacteremia 0% 52% 100% (44 to 100) 19 (10 to 46)

*Abbreviations defined in Glossary; RRR, NNT, and CI calculated from data in article.


The incidence of central venous catheter infection and associated bacteremia is high in patients who are critically ill, resulting in substantial morbidity and mortality. Aside from strict adherence to aseptic practices during catheter insertion (e.g., “maximal barrier precautions,” including gloves, gowns, and large sterile drapes) and during subsequent care, few interventions have been shown to effectively reduce infection rates. Innovative strategies to minimize the risk for infection that are being actively pursued include new biomaterials to improve the resistance of catheters to microbial colonization, as illustrated by these 2 trials of antiseptic-impregnated or antibiotic-coated catheters.

Maki and colleagues found a nearly 5-fold reduction of catheter-related bacteremia with antiseptic-impregnated catheters and a 2-fold reduction in the rate of catheter colonization. In this study, catheters, rather than patients, were used as the unit of analysis because patients were allocated to receive an antiseptic-impregnated catheter or a standard catheter on repeated insertion, either at a new site or with exchange over a guidewire. A trend was shown for fewer cases of bacteremia with antiseptic-impregnated catheters, whether it was a first catheter in a new site (1.4% vs 4.6%, P = 0.14) or a catheter inserted using guidewire exchange (2.0% vs 6.1%, P = 0.14). Although more local erythema was noted with impregnated catheters (54% vs 35%, P < 0.01), no patient experienced local or systemic hypersensitivity. Small amounts of chlorhexidine or silver sulfadiazine were detected in the blood of 1 of the 12 patients tested; however, 6 of the 12 patients had detectable serum levels of sulfadiazine (0.3 to 8.9 mg/L).

Raad and colleagues found a 3-fold reduction in the rate of catheter colonization with antibiotic-coated catheters and a markedly decreased rate of catheter-related bacteremia. As expected, these differences were entirely accounted for by a lower rate of colonization with gram-positive cocci (present in 31 uncoated vs 4 coated catheters, P < 0.001), whereas the rate of colonization with gram-negative bacilli or Candida was similar (6 uncoated vs 4 coated catheters). Staphylococci isolates recovered from catheters or skin insertion sites were not resistant to the antibiotics coating the catheters (minimal inhibitory concentration of 1 mg/L for minocycline and 2 mg/L for rifampin). However, it is not known whether the minimal inhibitory concentration of these strains differs from those of staphylococci isolated at the participating institutions and whether tetracycline resistance (endemic worldwide among staphylococci detected in hospitals) would affect the preventive efficacy of antibiotic-coated catheters. Concern remains about the selection of resistant strains with the use of antibiotic-coated catheters.

Both studies used molecular typing to compare isolates from blood and catheters and to confirm the origin of infection. The results of both trials are remarkably consistent in terms of the relatively high rates of infection in control catheters despite maximal barrier precautions used on insertion (about 25% colonization rate of catheters and an incidence of bacteremia of 7/1000 catheter-days or in 5% of catheters) and in the protective effect from coating catheters. Both trials studied triple-lumen central venous polyurethane catheters inserted for a relatively short period (mean of 6 d in both studies) in patients who were critically ill. It is unknown whether a similar benefit would result from using either antiseptic- or antibiotic-coated catheters in patients at lower risk for infection or different types of catheters (i.e., single- or double-lumen catheters). Moreover, it is unclear whether a similar benefit would allow longer-term catheterization because the antiseptics or antibiotics are gradually released and anti-microbial activities may persist for about 10 to 15 days.

For now, this new approach to prevention appears to be beneficial in high-risk patients who are critically ill and require relatively short-term central venous access (e.g., for 3 to 10 d). It may also decrease costs in institutions where the rates of infection remain high despite adequate infection control practices.

Christian Brun-Buisson, MD
Université Paris XIICréteil, France

Christian Brun-Buisson, MD
Université Paris XII
Créteil, France