The association between
hemodialysis vascular access type, costs, and outcome of Staphylococcus aureus
bacteremia (SAB) among patients with
ESRD remains incompletely characterized. This study was undertaken to compare resource utilization, costs, and clinical outcomes among SAB-infected patients with
ESRD by
hemodialysis access type. Adjusted comparisons of costs and outcomes were based on multivariable linear regression and multivariable logistic regression models, respectively. A total of 143 hospitalized
hemodialysis-dependent patients had SAB at Duke University Medical Center between July 1996 and August 2001. A total of 111 (77.6%) patients were hospitalized as a result of suspected
bacteremia; 32 (22.4%) were hospitalized for other reasons. Of the 111 patients, 59.5% (n = 66) had
catheters as their primary access type, 36% (n = 40) had arteriovenous (AV) grafts, and 4.5% (n = 5) had AV
fistulas. Patients with
fistulas were excluded from analyses because of small numbers. Patients with
catheters were more likely to be white, had shorter dialysis vintage, and had higher Acute Physiology and Chronic Health Evaluation II scores compared with patients with grafts. Unadjusted 12-wk mortality did not significantly differ between patients with
catheters compared with patients with grafts (22.7 versus 10.0%; P = 0.098); neither did 12-wk costs differ by access type ($22,944 +/- 18,278 versus $23,969 +/- 13,731,
catheter versus graft; P > 0.05). In adjusted analyses, there was no difference in 12-wk mortality (odds ratio 1.63; 95% confidence interval 0.29 to 9.02;
catheter versus graft) or 12-wk costs (means ratio 0.84; 95% confidence interval 0.60 to 1.17;
catheter versus graft) among SAB-infected patients with
ESRD on the basis of
hemodialysis access type. Twelve-week mortality and costs that are associated with an episode of SAB are high in
hemodialysis patients, regardless of vascular access type. Efforts should focus on the prevention of SAB in this high-risk group.