Impact of arrhythmias on heart rate variability-strategies to deal with imperfect clinical data

By introducing differing degrees and quality of ectopic beats into clinical data files, the considerable impact of arrhythmias on heart rate variability is demonstrated. Arrhythmias can be removed by either of three approaches proposed, each having variable fidelity with respect to restoring time and frequency domain indices of heart rate variability. Methods devised for the correction of the RR-interval time series substitute the ectopy associated RR-intervals by either a constant RR-interval, identical to the last normal value prior to the ectopic beat (Method I), by the mean value of the total sum of ectopy-associated RR-intervals (Method II) or by values generated by cubic splining and interpolation over the duration of the arrhythmia (Method III). It is shown that splining with interpolation is the overall best technique of overcoming undesired segments in data files with complex arrhythmias. The other techniques, however, have their own merits, particularly when only little ectopy is present