OBJECTIVE: This analysis evaluated the medical cost savings of delaying time to dialysis in elderly CKD patients receiving EPO. METHODS: Using health claims and laboratory data from >35 health plans between January 1999 and April 2004, dialysis patients (≥65 years) who had ≥1 hemoglobin (Hb) value and ≥1 glomerular filtration rate (GFR) value <60 mL/min prior to dialysis were identified. Patients were excluded if they had an organ transplant, had received blood transfusions or darbepoetin alfa, or had received dialysis for non-CKD reasons. Each CKD patient in the EPO group was matched by Hb and GFR to one control CKD patient not receiving EPO. The time from when the EPO and matched control patients had the same GFR value to dialysis was compared. The economic impact of EPO on delaying dialysis was monetized using published Medicare expenditures in 2001, adjusted to 2005, at $1204 per member per month (PMPM) prior to end-stage renal disease (ESRD) and $6020 PMPM for the post-ESRD period, and was calculated as the PMPM cost difference between the post- and pre-ESRD periods multiplied by the number of months of delay to dialysis by EPO. RESULTS: Sixty-eight patients (34 EPO and 34 controls) formed the study population. Mean age was similar between the two groups (years; EPO: 75.0 vs. control: 75.3), with a higher proportion of men in the EPO group (68% vs. 44%). The average time to dialysis was 156 days longer for the EPO group (319±252 vs. 163±151 days, P=0.003), resulting in cost savings of $25,044 per patient relative to the control group. CONCLUSIONS: This matched cohort analysis suggested that elderly CKD patients receiving EPO had a longer time to dialysis, thus reducing medical costs. These findings need to be further validated.