Vascular Access Outcomes and Medication Use: A USRDS Study Alexander S. Yevzlin,* Emily L. Conley,* Robert J. Sanchez,†‡ Henry N. Young,† and Bryan N. Becker* *School of Medicine and Public Health, University of Wisconsin; †School of Pharmacy, University of Wisconsin, Madison, Wisconsin; and ‡Pfizer, Inc., New York, New York ABSTRACT Several medications have been proposed to improve hemodial- ysis (HD) vascular access outcomes based on potentially favo- rable anticoagulant, antiplatelet, or pleiotropic properties. The purpose of this study was to evaluate the relationship between medication use and vascular access patency in a group of HD patients. We conducted a historical cohort study of the US Renal Data System Dialysis Mortality and Morbidity Wave II study to identify patients with an arteriovenous fistula (AVF), polytetrafluoroethylene (PTFE) graft, or a permanent catheter for vascular access. Cox regression analysis, adjusted for age, gender, race, history of coronary artery disease, peripheral vas- cular disease, or coronary artery bypass graft, was used to model the hazard ratio (HR) of permanent vascular access fail- ure. Of the 2001 HD patients in the Wave II study, 901 (45%) were included in the analysis. PTFE graft patency was greater for males (HR, 0.73; 95% CI 0.53–1.00, p ¼ 0.05) and for older individuals (HR, 0.99; 95% CI 0.98–1.00, p ¼ 0.02). Treatment with antiplatelet medications, ticlopidine and dipy- ridamole (HR, 3.54; 95% CI 1.07–11.76; p ¼ 0.04), or aspirin (HR, 2.49; 95% CI 1.31–4.73; p ¼ 0.005) was associated with significantly worse AVF patency. Antiplatelet agents had a sig- nificant negative association with access patency in this cohort. In contrast to other published data, it was difficult to identify any beneficial effect of specific medications on access patency. Vascular access failure (VAF) is the most common reason for hospitalization among hemodialysis (HD) patients (1,2). The economic burden of VAF is estimated at greater than 1 billion dollars per year (3) and contin- ues to grow. The most common cause of VAF in arterio- venous fistulae (AVF) and polytetrafluoroethylene (PTFE) arteriovenous grafts is stenosis at the venous an- astomosis (4), where inflammation and oxidative stress lead to neointimal proliferation and, eventually, vascular access thrombosis (5). Various medications have been shown to reduce neo- intimal proliferation and vascular inflammation based on in vitro and in vivo studies. These include HMG-CoA reductase inhibitors (statins) (6–8), angiotensin convert- ing enzyme (ACE) inhibitors (9), angiotensin II receptor blockers (ARB) (10), calcium channel blockers (CCB) (11,12), aspirin (13), and combinations of these drugs (9,11). HD patients typically are prescribed many of these medications for cardiovascular benefit. Interest- ingly, only a few of these medications have been studied for their impact on vascular access outcomes. Saran et al. (14) used the Dialysis Outcomes and Practice Patterns Study (DOPPS) database to analyze the relationship between several cardioprotective and antithrombotic medications and vascular access patency. Their findings showed that CCBs and aspirin were associated with improved graft patency, while ACE inhibitors were asso- ciated with improved fistula patency, and warfarin appeared to have a negative effect on graft patency. To further define associations that may exist between vasoactive medications and vascular access outcomes in the United States, we performed a retrospective cohort analysis using the United States Renal Data System (USRDS) database. We hypothesized that the afore- mentioned vasoactive medications would improve access outcomes. Methods We conducted a retrospective analysis of data from the USRDS Dialysis Morbidity and Mortality Study (DMMS) Wave II. Full details of data compilation and studies have been described elsewhere (15). Briefly, DMMS Wave II is a prospective longitudinal study of adult HD patients sampled from 25% of the dialysis units in the United States. All incident (new) end-stage renal disease (ESRD) patients as of January 1, 1996, over the age of 18, receiving either HD or peritoneal dialysis were eligible for inclusion. The present study uses data compiled from the HD patients in Wave II, with incident patients from 1996 to 1997. Address correspondence to: Alex Yevzlin, School of Medicine and Public Health, University of Wisconsin, 3034 Fish Hatchery Road, Suite B, Madison, WI 53713, or e-mail: asy@medicine.wisc.edu. Seminars in Dialysis—Vol 19, No 6 (November–December) 2006 pp. 535–539 535