Schmidt G, Barthel P, Ulm K, Goedel-Meinen L, Baedeker W
I. Medizinische Klinik der Technischen Universität München.
Herz. 1990 Feb;15(1):11-20.
The question of the reliability of ambulatory Holter monitoring for assessment of antiarrhythmic treatment has not been adequately resolved. Even though treatment efficacy had been individually assessed with Holter monitoring in the CAST study, during long-term treatment with class IC antiarrhythmic drugs, there were more deaths among patients receiving active drug than in those in the placebo group. Basic biostatistical considerations: Due to the spontaneous variability of frequency and complexity of ventricular arrhythmias, parametric models were developed with the aid of which normal ranges for spontaneous variability of singular ventricular premature complexes (VPC), couplets and salvos can be calculated. We designed a model which enables rapid visual analysis of the results: spontaneous variability = log (EE Day 2 + 0.01/EE Day 1 + 0.01) where EE is the number of ectopic events. For both days, the mean values per hour are applied. The use of parametric models prerequisites normal distribution of the data which can be achieved with logarithmic transformation. A constant is added to all mean values to preclude the mathematically-inadmissible form of log 0. The magnitude of the constant results in some degree of underestimation of the spontaneous variability. We chose the smallest constant, c = 0.01, consistent with a normal distribution of data. Figure 1 shows the normal range of the variability quotients for VPC in patients with cardiac disease and complex ventricular arrhythmias. The contiguous regions above and below the normal range designate active areas indicative of reduction or aggravation. Determinants of spontaneous variability: Frequency of arrhythmias: The number of VPC per unit of time exerts considerable influence on the spontaneous variability. The more infrequent an arrhythmia, the greater is the fluctuation to be anticipated. The differences in the variability of VPC, couplets and salvos are almost exclusively due to their differing frequencies since, in the presence of comparable frequency, they cannot be distinguished statistically from each other (Figure 2). Type and extent of underlying cardiac disease: In our patient population, there were no differences in spontaneous variability of arrhythmias between patients with coronary artery disease and those with dilated cardiomyopathy (Figure 3). Although in patients with coronary artery disease, as compared to those with noncoronary disease, a higher degree of spontaneous variability has been reported for VPC but, due to the inhomogeneity of the latter group, valid comparison is encumbered. The ejection fraction, the left ventricular filling pressure and the end-diastolic volume do not exert meaningful influence on the spontaneous variability (Figures 4 to 6).(ABSTRACT TRUNCATED AT 400 WORDS)
动态心电图监测用于评估抗心律失常治疗的可靠性问题尚未得到充分解决。尽管在心律失常抑制试验（CAST）中已通过动态心电图监测对治疗效果进行了个体评估，但在使用IC类抗心律失常药物进行长期治疗期间，接受活性药物治疗的患者死亡人数多于安慰剂组。基本生物统计学考量：由于室性心律失常的频率和复杂性存在自发变异性，因此开发了参数模型，借助该模型可以计算单个室性早搏（VPC）、成对早搏和连发早搏自发变异性的正常范围。我们设计了一个能够快速直观分析结果的模型：自发变异性=log（第2天异位事件数+0.01/第1天异位事件数+0.01），其中异位事件数即EE。两天均采用每小时的平均值。参数模型的使用前提是数据呈正态分布，这可通过对数转换实现。给所有平均值加上一个常数，以避免出现数学上不可接受的log 0形式。该常数的大小会在一定程度上低估自发变异性。我们选择了与数据正态分布一致的最小常数，c = 0.01。图1显示了患有心脏病和复杂室性心律失常患者VPC变异性商数的正常范围。正常范围上下的相邻区域表示活性区域，分别指示降低或加重。自发变异性的决定因素：心律失常的频率：每单位时间的VPC数量对自发变异性有相当大的影响。心律失常越不频繁，预期的波动就越大。VPC、成对早搏和连发早搏变异性的差异几乎完全归因于它们不同的频率，因为在频率相当的情况下，它们在统计学上无法相互区分（图2）。潜在心脏病的类型和程度：在我们的患者群体中，冠心病患者和扩张型心肌病患者心律失常的自发变异性没有差异（图3）。尽管与非冠心病患者相比，冠心病患者的VPC自发变异性程度更高，但由于后者组的异质性，难以进行有效的比较。射血分数、左心室充盈压和舒张末期容积对自发变异性没有显著影响（图4至6）。（摘要截选至400字）
