[Skip to Content]
[Skip to Content Landing]
Citations 0
July 7, 2010

Anemia Management and Mortality Risk in Incident Hemodialysis Patients—Reply

JAMA. 2010;304(1):41-43. doi:10.1001/jama.2010.891

In Reply: Dr Zhang and colleagues are incorrect that we found that “greater ESA dosage was associated with decreased mortality risk.” We observed decreased mortality rates in centers using larger ESA doses in patients with low hematocrit. And we observed increased mortality rates in centers using greater ESA doses in patients with high hematocrit. These results suggest that the risk/benefit trade-off of high-dose ESA depends on the severity of anemia. To address their concern whether our results may be sensitive to excluding ESA dosing data from months in which patients spent more than 5 days in the hospital, we re-ran our main analysis including data from months in which patients spent up to 10 days in the hospital. Our findings were very similar to those originally reported (Table). We think it is highly improbable that the 1- to 2-day gaps between hospitalization and outpatient dialysis would lead to any meaningful misclassification of a dialysis unit's anemia management practice. We disagree with Zhang et al that our findings are inconsistent with a report that the dialysis chain using the smallest doses of ESAs also had the lowest mortality rates.1 Our model suggests that centers using ESAs the most aggressively across all hematocrit categories would have increased mortality rates relative to the most conservative centers. Therefore, our results are quite compatible with the cited report.

First Page Preview View Large
First page PDF preview
First page PDF preview