Hemoglobin loevel variability: Anemia management among variability groups

David T. Gilbertson, Yi Peng, Brian Bradbury, James P. Ebben, Allan J. Collins

Research output: Contribution to journalReview articlepeer-review

18 Scopus citations


Background/Aims: Hemoglobin level variability in hemodialysis patients is common, and has been associated with comorbidity, intercurrent illness, and mortality risk. We aimed to describe the influence of anemia management interventions (erythropoiesis-stimulating agents [ESAs], intravenous iron, and transfusions) on hemoglobin variability. Methods:We studied all Medicare primary payer hemodialysis patients who survived and had ESA claims in the first 6 months of 2004 (n = 159,720). Monthly hemoglobin values were categorized as low (<11 g/dl), intermediate (11-12.5 g/dl), and high (>12.5 g/dl). Variability groups were classified based on lowest and highest hemoglobin categories during a 6-month observation period. ESA, intravenous iron, and transfusion use were characterized by variability group. Results:Patients with consistently low or low and intermediate hemoglobin received the highest ESA doses and the most frequent transfusions, while patients with consistently or intermittently intermediate or high hemoglobin received lower ESA doses and fewer transfusions. Intravenous iron doses were highest initially for patients with consistently high hemoglobin; these doses subsequently declined. Iron doses were lowest for patients with consistently intermediate hemoglobin. Conclusions: Anemia management protocols describing coordinated administration of ESAs and iron may help to increase the number of patients achieving target hemoglobin levels.

Original languageEnglish (US)
Pages (from-to)491-498
Number of pages8
JournalAmerican Journal of Nephrology
Issue number6
StatePublished - Dec 2009


  • Anemia
  • Epoetin alfa
  • Erythropoietin
  • Hemoglobin
  • Iron, intravenous
  • Mortality


Dive into the research topics of 'Hemoglobin loevel variability: Anemia management among variability groups'. Together they form a unique fingerprint.

Cite this