Abstract
BACKGROUND AND OBJECTIVES: To determine whether warfarin prolongs the time to first mechanical-catheter failure. DESIGN, SETTING, PARTICIPANTS, & MEASUREMENTS: This was a multicenter parallel-group randomized controlled trial with blinding of participants, trial staff, clinical staff, outcome assessors, and data analysts. Randomization was in a 1:1 ratio in blocks of four and was concealed by use of fax to a central pharmacy. Hemodialysis patients with newly-placed catheters received low-intensity monitored-dose warfarin, target international normalized ratio (INR) 1.5 to 1.9, or placebo, adjusted according to schedule of sham INR results. The primary outcome was time to first mechanical-catheter failure (inability to establish a circuit or blood flow less than 200 ml/min). RESULTS: We randomized 174 patients: 87 to warfarin and 87 to placebo. Warfarin was associated with a hazard ratio (HR) of 0.90 (P=0.60; 95% confidence interval [CI], 0.57, 1.38) for time to first mechanical-catheter failure. Secondary analyses were: time to first guidewire exchange or catheter removal for mechanical failure (HR 0.78; 95% CI, 0.37, 1.6); time to catheter removal for mechanical failure (HR 0.67; 95% CI, 0.19, 2.37); and time to catheter removal for any cause (HR 0.89; 95% CI, 0.42, 1.81). Major bleeding occurred in 10 participants assigned to warfarin and seven on placebo (relative risk, 1.43; 95% CI, 0.57, 3.58; P=0.61). CONCLUSIONS: We found no evidence for efficacy of low-intensity, monitored-dose warfarin in preventing mechanical-catheter failure.