Muñoz Navarro S R, Bangdiwala S I
Unidad de Epidemiología Clínica, Departamento de Salud Pública, Facultad de Medicina, Universidad de La Frontera, M. Montt 112, Temuco, Chile.
Rev Med Chil. 2000 Aug;128(8):935-41.
Interim analysis of data accumulated in clinical trials is one aspect of the monitoring of the study progress. It is usually done to assess whether there are significant differences in efficacy between the experimental and control treatment groups, in order to decide whether to stop or no the trial prematurely. Among many reasons for early interruption of a trial is the ethical consideration that subjects should not be exposed to an unsafe, inferior or ineffective treatment. Statistical methods suited for doing interim analysis, that allow to control the probability of incorrectly rejecting the null hypothesis of no treatment differences, are often not well understood by researchers. In this article we present an intuitive, non-mathematical explanation and review of the statistical methods for doing interim analysis in clinical trials along with an illustrative example of the application of the methods on a hypothetical dataset.
对临床试验中积累的数据进行期中分析是研究进展监测的一个方面。这样做通常是为了评估试验组和对照组在疗效上是否存在显著差异，以便决定是否提前终止试验。试验提前中断的诸多原因中，有一个伦理考量是受试者不应暴露于不安全、劣质或无效的治疗中。适合进行期中分析的统计方法，能控制错误拒绝无治疗差异零假设的概率，但研究人员对此往往了解不足。在本文中，我们对临床试验期中分析的统计方法进行直观、非数学的解释和综述，并给出一个在假设数据集上应用这些方法的示例。
