Supplementary Material for: Mortality Associated with Dose Response of Erythropoiesis-Stimulating Agents in Hemodialysis versus Peritoneal Dialysis Patients

Background: Several studies have shown an association between erythropoietin-stimulating agent (ESA) responsiveness and mortality in chronic kidney disease (CKD) patients. In our present study, we examined the association between prescribed ESA dose and mortality in peritoneal dialysis (PD) and hemodialysis (HD) patients. We hypothesized that PD patients received lower ESA dose for the same achieved hemoglobin compared to HD patients and that ESA dose-mortality associations were different between PD and HD patients. Methods: We compared the prescribed doses of ESA between 139,103 HD and 10,527 PD patients treated in DaVita dialysis clinics from 7/2001 through 6/2006 using adjusted Poisson regression and examined mortality-predictability of prescribed ESA dose and ESA responsiveness index (ESA/hemoglobin) in PD and HD with follow-up through 6/2007 using Cox regression models. Results: Poisson adjusted ratio of ESA dose of HD to PD was 3.6 (95% CI 3.5–3.7). In PD patients, adjusted all-cause death hazard ratios (HR) for ESA doses of 3,000–5,999, 6,000–8,999 and ≧9,000 U/week (reference <3,000 U/week) were 0.97 (0.87–1.07), 0.85 (0.76–0.95) and 1.08 (0.98–1.18), respectively; whereas in HD patients across commensurate ESA dose increments of 10,000–19,999, 20,000–29,999 and ≧30,000 U/week (reference <10,000 U/week) were 1.14 (1.11–1.17), 1.54 (1.50–1.58) and 2.15 (2.10–2.21), respectively. In PD and HD patients, the adjusted death HR of the 4th to 1st quartile of ESA responsiveness index were 1.14 (1.04–1.26) and 2.37 (2.31–2.43), respectively. Conclusions: Between 2001 and 2006, most PD patients received substantially lower ESA dose for same achieved hemoglobin levels, and low ESA responsiveness was associated with higher mortality in both HD and PD patients.