Use and cost of erythropoiesis-stimulating agents in patients with cancer

To compare the baseline characteristics, episodes of care, and cost of erythropoiesis-stimulating agents among cancer patients in a US managed-care population.

Retrospective analysis of administrative claims data. Episodes of care for patients with cancer receiving erythropoiesis-stimulating agents between January 1, 2004 and January 17, 2006 included all claims for erythropoiesis-stimulating agents with < or = 42 days' gap between claims, plus the duration of therapeutic benefit based on median days between consecutive doses.

Main outcome measures were average weekly dose of erythropoiesis-stimulating agents and costs of therapy.

A total of 15,007 eligible episodes of care (darbepoetin alfa, 7769 episodes [5587 patients]; epoetin alfa 7238 episodes [5157 patients]) were identified. Fewer claims were observed per episode of care for darbepoetin alfa than for epoetin alfa (mean [SD] 3.7 [4.1] vs. 5.3 [6.4]). The median time between consecutive claims was 15 days (darbepoetin alfa) and 8 days (epoetin alfa). The mean (SD) weekly doses were 105 (56) microg (darbepoetin alfa) and 34,242 (28173) U (epoetin alfa), a dose-comparison ratio of 326 : 1. Dose-comparison ratios were sensitive to assumptions about duration of clinical benefit. The mean (95% CI) weekly costs were significantly lower for darbepoetin alfa ($560 [553-567]) than for epoetin alfa ($645 [630-659], p < 0.0001) when duration of clinical benefit was considered.

Significant differences characterize patterns of use of erythropoiesis-stimulating agents. Duration of therapeutic benefit is an important variable in comparing darbepoetin alfa with epoetin alfa; incorporation of this variable in analyses of costs of therapy may have notable effects on calculated treatment costs. Limitations of the study include the potential for database errors or omissions, lack of detailed disease data, and lack of adjustment for differences in the ages and comorbidities of patients.