Brown M B
Department of Biostatistics, School of Public Health, University of Michigan, Ann Arbor 48109-2029.
Control Clin Trials. 1992 Jun;13(3):213-25. doi: 10.1016/0197-2456(92)90004-j.
Consider a randomized clinical trial that is designed to compare two treatments in which the treatment continues during the entire period of the study. Some subjects may refuse to complete the protocol and will not return for the final evaluation. Since the reason for dropping out may be related to the subject's self-assessed evaluation of the usefulness of the treatment or to undesirable side effects of the treatment, subjects who drop out cannot be treated as a random sample of those who entered the trial. We consider the situation where the measure of efficacy of the treatment is continuous. Under the assumption that the expected value of the measure for those who drop out is not better (the direction depends on the measure) than that for those who complete the study, we propose an adjustment to the usual test for a difference between treatments that allows for the inclusion of the probable effect of the dropouts; this provides a bound on the test for efficacy of the treatment. First, we estimate a predetermined percentile, such as the median score, of the control, or placebo, group and assign this score to all those who dropped out from both groups and to all subjects in both groups with scores that are worse than the assigned score. A Mann-Whitney statistic is then used to test the equality of the distributions of the two groups. We show by simulation that this modified test is equivalent to a test using the complete data and has greater power than that obtained when including the dropouts by assigning the worst observed score to them. This test will be less sensitive to bias that is induced by exclusion of dropouts from the final evaluation.
考虑一项随机临床试验，其旨在比较两种治疗方法，且治疗在整个研究期间持续进行。一些受试者可能拒绝完成方案，且不会返回进行最终评估。由于退出的原因可能与受试者对治疗有效性的自我评估或治疗的不良副作用有关，因此退出的受试者不能被视为进入试验的受试者的随机样本。我们考虑治疗效果度量为连续变量的情况。在退出者的度量期望值不比完成研究的受试者更好（方向取决于度量）这一假设下，我们提出对治疗组间差异的常规检验进行调整，以考虑退出者可能产生的影响；这为治疗效果检验提供了一个界限。首先，我们估计对照组或安慰剂组的一个预定百分位数，比如中位数得分，并将该得分赋予两组中所有退出的受试者以及两组中所有得分比所赋得分差的受试者。然后使用曼 - 惠特尼统计量来检验两组分布的相等性。我们通过模拟表明，这种改进后的检验等同于使用完整数据的检验，并且比将最差观察得分赋予退出者并将其纳入分析时所获得的检验具有更大的功效。该检验对因在最终评估中排除退出者而导致的偏差不太敏感。
