In this work, Refined Multiscale Entropy (RMSE) was applied to characterize risk of cardiac death in ischemic cardiomyopathy patients, analyzing heart rate variability (HRV) by means of RR series during daytime and nighttime. RMSE approach measures an entropy rate in different time scales of a series, giving a multiscale characterization of complexity of that series. RMSE showed statistically significant differences (p<;0.05) during daytime and nighttime only in middle time scales (τ=4-15 and τ=3-16, respectively). For these scales, RMSE was higher in low risk (SV) than in high risk (CM) group of cardiac death, indicating a reduction of the entropy-based complexity in CM when it was compared with SV. No statistical differences between risk groups were presented at time scale τ=1 (unfiltered original RR series). It can be concluded that the dynamics in middle time scales should be considered to better describe the HRV of patients with cardiac death.