Non‐parametric analysis of covariance for confirmatory randomized clinical trials to evaluate dose–response relationships
Non‐parametric analysis of covariance for confirmatory randomized clinical trials to evaluate dose–response relationships
复制标题
用于评估剂量反应关系的验证性随机临床试验的协方差非参数分析
作者：
C. Tangen;G. Koch
在旨在将一种测试治疗的多个剂量与对照组以及彼此之间进行比较的验证性随机临床试验中，通常存在需要考虑的关于复合假设和多重比较的统计问题。在大多数情况下，分析计划需要明确规范进行统计检验（或管理总体显着性水平）的建议顺序、将使用哪些统计方法以及是否将对协变量进行调整。指定非参数协方差分析 (ANCOVA) 来执行主要验证性分析有几个好处。在研究设计中，除了随机化之外，只需要最少的假设，而基于回归模型的方法对模型拟合有假设，如果偏离，可能需要与完全预先指定的分析计划不兼容的修改。非参数方法提供了传统上预期的 ANCOVA 结果；也就是说，当协变量与感兴趣的响应强相关时，通常会对治疗比较的估计进行小幅调整（以考虑治疗组之间协变量的随机不平衡）和该估计的方差减少。两项随机临床试验说明了非参数 ANCOVA 的应用。第一个采用 (3×4) 阶乘响应面设计，用于比较 12 种治疗方法（即三种剂量的一种药物和四种剂量的另一种药物的组合）的血压变化；第二个例子比较了三种剂量的测试治疗和安慰剂的疾病进展时间。该临床试验对二分标准、Wilcoxon 等级评分和累积生存率平均值的治疗进行了比较。在每个示例中，非参数协方差方法相对于其未调整的对应方法提供了方差减少。版权所有 © 2001 约翰·威利父子有限公司
In confirmatory randomized clinical trials that are designed to compare multiple doses of a test treatment with a control group and with one another, there are often statistical issues regarding compound hypotheses and multiple comparisons which need to be considered. In most cases the analysis plan needs a clear specification for the proposed order for conducting statistical tests (or for managing the overall significance level), which statistical methods will be used, and whether adjustment for covariates will be performed. There are several benefits of specifying non‐parametric analysis of covariance (ANCOVA) for performing the primary confirmatory analyses. Only minimal assumptions are needed beyond randomization in the study design, whereas regression model based methods have assumptions about model fit for which departures may require modifications that are incompatible with a fully prespecified analysis plan. Non‐parametric methods provide traditionally expected results of ANCOVA; namely, a typically small adjustment to the estimate for a treatment comparison (so as to account for random imbalance of covariates between treatment groups) and variance reduction for this estimate when covariates are strongly correlated with the response of interest. The application of non‐parametric ANCOVA is illustrated for two randomized clinical trials. The first has a (3×4) factorial response surface design for the comparison of 12 treatments (that is, combinations of three doses of one drug and four doses of a second drug) for change in blood pressure; and the second example addresses the comparison of three doses of test treatment and placebo for time‐to‐disease progression. This clinical trial has comparisons among treatments made for a dichotomous criterion, Wilcoxon rank scores and averages of cumulative survival rates. In each example, the non‐parametric covariance method provides variance reduction relative to its unadjusted counterpart. Copyright © 2001 John Wiley & Sons, Ltd.
DOI：
10.1016/s0197-2456(97)00147-5
发表时间：
1998-06-01
期刊：
CONTROLLED CLINICAL TRIALS
影响因子：
--
作者：
Hauck, WW;Anderson, S;Marcus, SM
通讯作者：
Marcus, SM