薬効評価における知っておくべき統計学の知識

抄録

In clinical trials of pharmaceuticals and medical devices, evaluation of efficacy is difficult due to the following problems ; there are variations between and within the therapeutic effects and onset of adverse events, possibility of bias based on study design, and clinical trial data is information from limited patient group and limited sample size in an environment different from usual medical care, etc. In clinical trials, statistical aspects are important in all stages of appropriate study plans, planned study implementation and data collection, analysis, and reporting of results. Therefore, cooperation with medical statistics experts is required from the planning stage. Needless to say, there is a need for more efficient and sensitive communication between statisticians and researchers, project managers, data managers, monitors and CRCs and support staff of clinical trial. In this paper, the 6 points that characterize the clinical trial design that statisticians are paying close attention to as a knowledge of statistics to plan the design are explained.