Original Contribution Erythropoiesis-Stimulating Agent Use After Changes in Medicare Reimbursement Policies

Purpose: Since 2004, concerns about the safety of erythro-poiesis-stimulating agents (ESAs) have resulted in label changes and restrictions on their use. We examined changes in ESA use and blood transfusions over time. Methods: The SEER-Medicare database was used to identify patients age 65 years with breast, lung, prostate, ovary, or colon cancer, diagnosed between 2000 and 2007, who had a chemotherapy claim after their cancer diagnosis. We calculated the mean number of ESA claims per patient per year. Follow-up claims were available through 2008. We used multivariable logis-tic regression models to analyze the association of ESA use and extended ESA use with clinical and demographic variables. Results: Among 121,169 patients identified, 46,063 (38%) received an ESA. ESA use increased from 12.4 % to 16.2 % by 2006 and then decreased to 7.9 % by 2008. Similarly, the mean number of ESA claims per patient decreased steadily over the entire timeframe. The annual percentage of patients undergoing transfusion remained relatively constant (9 % to 10%). In a Cox proportional hazards time-dependent model, ESA use was pos-itively associated with black race (odds ratio [OR], 1.11; 95 % CI, 1.07 to 1.15), metropolitan location (OR, 1.17; 95 % CI, 1.13 to 1.21), metastatic disease (OR, 1.39; 95 % CI, 1.35 to 1.41), fe-male sex (OR, 1.17; 95 % CI, 1.14 to 1.20), one comorbidit