687 Outpatient Dialysis Event Surveillance: A Report from the National Healthcare Safety Network, 2007-2008

Saturday, March 20, 2010: 2:15 PM
Regency VI-VII (Hyatt Regency Atlanta)
Priti Patel, MD, MPH , Centers for Disease Control and Prevention, Baltimore, MD
Alexander Kallen, MD, MPH , Centers for Disease Control and Prevention, Atlanta, GA
Jonathan Edwards, MS , Centers for Disease Control and Prevention, Atlanta, GA
Teresa Horan, MPH , Centers for Disease Control and Prevention, Atlanta, GA


More than 340,000 persons in the United States receive outpatient hemodialysis for end-stage renal disease. Bloodstream infections (BSI) cause substantial morbidity and mortality in this population. National quality improvement efforts have focused on increasing fistula use (goal: ≥ 66% prevalence) and reducing long-term central venous catheter (CVC) use (goal: < 10% prevalence) to improve hemodialysis patient outcomes. The National Healthcare Safety Network (NHSN) collects information on BSI and other adverse events among dialysis patients.


Characterize dialysis event rates and vascular access use among chronic hemodialysis patients


Dialysis event surveillance data reported to NHSN from outpatient facilities during 2007-2008 were analyzed. Numerator data are based on event reports made for patients who had a positive blood culture (i.e., BSI), received outpatient intravenous (IV) antibiotics, or had an overnight hospitalization for any cause. The reported number of patients treated in a facility on the first 2 days of the month (monthly patient census) was used as the denominator and the vascular access distribution of these patients was determined. Pooled mean rates and 95% confidence intervals were stratified by vascular access type and calculated per 100 patient-months. Medians were compared using the Wilcoxon rank sum test.


A total of 49 outpatient dialysis facilities reported 10,193 events among 66,694 patient-months during 2007-2008. Twenty-six (53%) facilities were based in, or affiliated with an acute-care hospital; the median average monthly census was 54 (range 15 - 348). The pooled mean rates of BSI, IV antibiotic starts, and hospitalizations were highest among patients with CVCs and lowest for those with fistulas (Table). Median facility vascular access use was 48.1% for fistulas and 34.8% for permanent CVCs, with substantial inter-facility variation. Although duration of CVC use could not be determined, all facilities had > 10% average prevalence of permanent CVCs. Compared to other facilities, hospital-affiliated outpatient facilities had lower median fistula use (45.9% vs. 51.7%; p=0.10) and higher permanent CVC use (37.7% vs. 27.7%; p=0.02). Vancomycin was used in 75.9% of 2989 IV antibiotic starts reported. Of 1593 BSIs reported, 70.4% occurred among patients with a CVC.


Rates of adverse events continue to be highest among dialysis patients with CVCs, and these patients contributed most to the overall burden of BSIs. Dedicated efforts are needed, in both hospital-based and other dialysis facilities, to reduce catheter use and prevent infections among hemodialysis patients who are catheter-dependent.