Catheter Related Bloodstream Infection in Haematological Patients: A Prospective, Randomized Study Comparing Heparin-Coated with Chlorhexidine and Silver Sulfadiazine Impregnated Central Venous Catheters

Central venous catheters (CVCs) are widely used in patients with hemato-oncological disease for indications such as administration of chemotherapy, parenteral nutrition, blood products and monitoring of hemodynamics. CVCs are a major source of hospital-acquired infection. The National Nosocomial Infections Surveillance System reported a rate of catheter-related bloodstream infection (CRBI) of five per 1000 central catheter-days. This study was a randomized trial in which patients with nontunneled catheters were randomly assigned to heparin-coated (7 French, polyurethane, double lumen; Arrow^®, USA) or chlorhexidine and silver sulfadiazine impregnated CVCs (7 French, polyurethane, double lumen; Arrow^®, USA). All CVCs were placed in the subclavian vein in the operating room. CRBI was defined according to Infectious Disease Society of America guidelines. The principal investigator determined whether infections were catheter related and had no knowledge of “the assigned arm” at the time of adjudication of the reference standard definition. Before catheter insertion, laboratory prothrombotic markers included factor V Leiden, the prothrombin gene Gly20210A mutation, plasma antithrombin levels, and protein C and S activity. In our study, all patients were systematically examined by ultrasonography just before, or less than 24 hours after, catheter removal and in case of clinical signs of thrombosis. Two radiologists performed the ultrasonography and were unaware of the allocation of the patients. Two hundred and twenty patients were randomly assigned. Eight patients were excluded after assignment. Ultimately, 212 patients were analyzed [median age: 28 years (7–60); 97 female and 115 male]. CRBI occurred in 6.7% (7/104) of those in the heparin-coated group (2.7 events per 1000 days) and in 4.6% (5/108) in the chlorhexidine and silver sulfadiazine group (2 events per 1000 days) (P=0.5). The microorganisms involved in CRBIs were coagulase-negative Staphylococcus (five cases in the heparin group and 3 cases in the antiseptic group), Staphylococcus aureus (one case in the heparin group), Candida parapsilosis (one case in the antiseptic group), Pseudomonas aeruginosa (one case in the heparin group and one case in the antiseptic group). We observed a CVC-related thrombosis in 11 patients (5.2%; 5 in the heparin group and 6 in the antiseptic group P=0.8). Catheter-related thrombosis and CRBI coincided in two patients and were not significantly correlated (P=0.6). Four (1.9%) and three (1.4%) patients had evidence of protein C and protein S deficiency, respectively. Only one patient had an antithrombin deficiency (0.5%). In total, 9 patients (4.2%) were heterozygous for the factor V Leiden mutation. Thrombosis was diagnosed in three out of 17 patients (17.6%) with a inherited prothrombotic abnormality compared to 8 out of 195 patients (4.1%) who did not have a thrombophilic marker (relative risk 4.3 CI 95% 1.2–14.7). Six and five patients experienced severe bleeding in the heparin and antiseptic groups, respectively. We did not observe heparin-induced thrombocytopenia and anaphylaxis to chlorhexidine was not reported. To our knowledge, this is the first randomized study comparing in haematological patients heparin-coated with antiseptic impregnated CVCs. Furthermore, our results suggest that inherited prothrombotic abnormalities contribute substantially to CVC-related thrombosis in hemato-oncological patients. Presently, the best options for reducing CRBI are heparin-coated and antibiotic-impregnated CVCs. A large, well designed randomized controlled trial is required to determine which of these is most effective.