Supplementary Material for: Extremely Low Dose of Erythropoiesis-Stimulating Agent May be Associated with Increased Mortality in Hemodialysis Patients

Introduction: Although high dose erythropoiesis-stimulating agent (ESA) has been shown to increase mortality risk and adverse cardiovascular events in hemodialysis patients, the safety of extremely low dose ESA is unclear. Methods: We retrospectively analyzed the association between ESA dose and mortality in the monthly dosing range of 0-43,000 U of equivalent epoetin alpha in 304 Taiwan hemodialysis patients by using Cox proportional hazard model and cubic spline model. Results: Compared with mean monthly ESA dose of 15,000-25,000 U (mean ± standard deviation 20,609 ± 2,662 U), monthly ESA dose of less than 15,000 U (mean ± standard deviation 7,413 ± 4,510 U) is associated with increased mortality. Monthly ESA dose of 25,001-43,000 U (mean ± standard deviation 31,160 ± 4,304 U) is not associated with higher mortality risk than monthly ESA dose of 15,000-25,000 U. The results were consistent in Cox proportional hazard models and cubic spline models. Subgroup analyses showed no significant heterogeneities among prespecified subgroups. Conclusions: Extremely low dose of ESA in hemodialysis patients may be associated with increased mortality risk. Future studies are warranted to prove this association.