Prevention of Tunneled Hemodialysis Catheter-Related Infections Using Catheter-Restricted Filling with Gentamicin and Citrate: A Randomized Controlled Study GURSHARAN K. DOGRA,* HELEN HERSON, † BRIAN HUTCHISON, † ASHLEY B. IRISH, ‡ CHRISTOPHER H. HEATH, § CLAY GOLLEDGE, ¶ GRANT LUXTON, † and HARRY MOODY † *Department of Medicine, University of Western Australia, Perth, Western Australia; † Department of Nephrology and ¶ Department of Microbiology, Sir Charles Gairdner Hospital, Perth, Western Australia; and ‡ Department of Nephrology and § Department of Microbiology, Royal Perth Hospital, Perth, Western Australia. Abstract. Tunneled catheters are widely used for the provision of hemodialysis. Long-term catheter survival is limited by tunneled catheter-related infections (CRI). This study assesses the efficacy of catheter-restricted filling with gentamicin and citrate in preventing CRI in hemodialysis patients. A double- blind randomized study was conducted to compare heparin (5000 U/ml) with gentamicin/citrate (40 mg/ml and 3.13% citrate; ratio 2:1) as catheter-lock solutions. A total of 112 tunneled catheters in 83 patients were enrolled at the time of catheter insertion for commencement or maintenance of hemo- dialysis. The primary end point was CRI. Catheter malfunc- tion, defined as blood flow rate of 200 ml/min for three consecutive dialyses and/or the use of urokinase, was also assessed as a secondary end point. Infection rates per 100 catheter-days were 0.03 in the gentamicin group versus 0.42 in the heparin group (P = 0.003). Kaplan-Meier survival analyses showed mean infection-free catheter survival of 282 d (95% CI, 272 to 293 d) in the gentamicin group versus 181 d (95% CI, 124 to 237 d) in the heparin group (log rank, 9.58; P = 0.002). Cox regression analyses showed a relative risk for infection-free catheter survival of 0.10 (95% CI, 0.01 to 0.92) in the gentamicin group when adjusted for gender, race, dia- betes mellitus, catheter malfunction, and hemoglobin (P = 0.042). The incidence of catheter malfunction was not signif- icantly different between groups. Predialysis gentamicin levels were significantly higher in patients randomized to gentamicin (gentamicin/citrate: median 2.8 mg/L [range, 0.6 to 3.5 mg/L], n = 5; heparin: median 0.2 mg/L [range 0.2 to 0.2 mg/L], n = 5; P = 0.008). Tunneled hemodialysis catheter-restricted filling with gentamicin and citrate is a highly effective strategy for prevention of CRI. Although citrate as a catheter-lock solution provides adequate anticoagulation for the interdialytic period, gentamicin levels suggest significant risk for chronic aminoglycoside exposure and associated ototoxicity. Before this technique is adopted, these preliminary observations war- rant replication in future studies that will examine the efficacy and safety of lower doses of gentamicin or alternative agents with a reduced potential for toxicity. In the absence of a functioning arteriovenous fistula or syn- thetic graft, tunneled hemodialysis catheters are essential for the provision of hemodialysis. Approximately 19% of new hemodialysis patients in the United States rely on tunneled catheters for the initiation of hemodialysis, and nearly 70% of these patients are still using tunneled catheters after 60 d (1). The most important factors limiting long-term survival of tunneled catheters are poor blood flow and catheter-related infections, each of which can predispose to the other (2). The infection rate for tunneled catheters averages 0.08 to 0.7 per 100 catheter-days (1,3,4). Catheter-related infections (CRI) are associated with a substantial morbidity, mortality, and addi- tional cost per infective episode. Data from nontunneled cath- eters used in intensive care units indicate an average 3% per annum mortality rate and a cost-per-infective-episode of be- tween US$3700 and US$29,000 (5). CRI are a consequence of colonization of the catheter hub or surrounding skin followed by intraluminal or extraluminal spread (4). Prevention strategies are directed at decreasing growth and/or adherence of pathogens to the catheter hub and surface (1,5). Catheter-restricted filling with antibiotics as a prophylaxis against infection has recently emerged as a prom- ising option, but evidence in the form of randomized controlled trials is lacking. Animal studies comparing catheters filled with gentamicin and chymotrypsin with heparin-filled catheters showed an absence of infection in the antibiotic treated group (6). As gentamicin is incompatible with heparin in solution (7), human studies have used catheter-restricted filling with genta- micin in combination with citrate, but they have only been published in abstract form (8 –10). Two of theses studies were Received February 18, 2002. Accepted May 1, 2002. Correspondence to Dr. Gursharan Dogra, MRF Building, University Depart- ment of Medicine, Royal Perth Hospital, GPO BOX X2213, Perth, Western Australia 6847. Phone: 61-8-9224-0312; Fax: 61-8-9224-0246; E-mail: sdogra@cyllene.uwa.edu.au 1046-6673/1308-2133 Journal of the American Society of Nephrology Copyright © 2002 by the American Society of Nephrology DOI: 10.1097/01.ASN.0000022890.29656.22 J Am Soc Nephrol 13: 2133–2139, 2002