Statistische Methoden zur Bewertung der Individuellen Wirkung Einer Anti-Arrhythmischen Substanz

Summary

In a controlled clinical trial the efficacy of a certain drug can be tested. The question is whether this drug is also effective within an individual patient. For the quantitative recording of ventricular arrhythmias a long term ECG monitoring is used. The evaluation of the ECG findings with respect to an effect of an anti-arrhythmic drug is a problem especially due to the high spontaneous variation in arrhythmic freqencies.

Previously a 50 to 70 per cent reduction of ventricular premature depolarization (=VPD) over 24 hours has been used as an index of efficacy.

In the last years some authors tried to establish objective criteria for assessing such an index based upon biostatistical methods. The different methods lead to different results in the minimal percantage reduction in VPD frequency required to demonstrate an effect due to the intervention rather than spontaneous variation at a certain level of confidence. The range for the reduction ist from 65 to 100 per cent for a period of 24 hours or longer. This is not feasible in reality, for one has to select an appropriate drug under several possible drugs. Therefore we developed a method which allows, within a period of 5 hours after injection, to exclude drugs with no effect. The reduction must be 50 to 60 per cent. The comparison with the existing methods confirms our findings.