Changes in the distribution of ventricular ectopic beats in long-term electrocardiograms.

Only simple methods have been used to assess antiarrhythmic and proarrhythmic effects when comparing baseline and on-therapy Holter recordings taken from the same patient. The paper suggests a new method for definition of these effects based upon comparisons of statistical distributions of ectopic beats. The method is based on multiple random sampling of the recordings and on application of the Smirnov test to compare the samplings. The results of these multiple comparisons are subsequently evaluated using the chi-squared test. An analysis is reported of Holter recordings made on seven patients suffering from ventricular arrhythmia but with anatomically normal heart. In each patient, one baseline recording and one recording on each of three different drugs were made. The results show that the definition of proarrhythmic effects can be partly addressed in a precise mathematical way. The method can also detect a significant change in the character of arrhythmia which can neither be classified as antiarrhythmic nor as proarrhythmic.