Rhythm annotation and interobserver reproducibility of measures of heart rate variability

In summary, these data indicate that interobserver reproducibility is high for time domain and power spectral measures of heart rate variability, but that reproducibility is greater for the low-frequency measures, and especially for SDANN, than for the high-frequency measures. Overall interobserver variability of <8% for these measures is largely (50% to 75%) explained by interobserver differences in annotation of supraventricular ectopy and sinus arrhythmia.