Predictors of ESA use in the non-dialysis chronic kidney disease population with anemia

Allan J. Collins, Haifeng Guo, David T. Gilbertson, Brian D. Bradbury

Research output: Contribution to journalArticlepeer-review

5 Scopus citations


Background: Anemia is a common complication of chronic kidney disease (CKD), but anemia treatment may be less comprehensive than guidelines suggest. Methods: The study population (n = 11,754) included general Medicare recipients with Parts A and B coverage before January 1, 2001, aged ≥65 years on January 1, 2001, and alive with Medicare as primary payer through December 31, 2001. Time-dependent proportional hazards models were used to investigate predictors of erythropoiesis-stimulating agent (ESA) use, adjusted for comorbid conditions and severity-of-disease variables as time-dependent, and age, sex, and race as fixed variables. ESA use was defined during 2002 and time-dependent variables during 2001-2002. Results: Only 839 patients (7%) received ESAs. Characteristics significantly predictive of ESA use (p < 0.05) were: outpatient specialty services, nephrology and hematology/oncology/medical oncology (RR 6.92); outpatient specialty services, hematology/oncology/medical oncology (RR 6.02); outpatient specialty services, nephrology (RR 3.44); inpatient principle procedure, other operations on vessels (RR 1.68); transfusions (RR 1.54), hypertension (RR 1.50); congestive heart failure (RR 1.34); home oxygen (RR 1.28). Conclusions: Access to anemia treatment may be an important marker for access to CKD care. Clinical trials are needed to assess effects of early referral and more comprehensive anemia treatment.

Original languageEnglish (US)
Pages (from-to)c141-c148
JournalNephron - Clinical Practice
Issue number2
StatePublished - Feb 2009


  • Anemia
  • Chronic kidney disease
  • Erythropoiesis-stimulating agent


Dive into the research topics of 'Predictors of ESA use in the non-dialysis chronic kidney disease population with anemia'. Together they form a unique fingerprint.

Cite this