Effect of Regression toward the Mean in Evaluating the Efficacy of Kampo Prescriptions. Analysis with AIC / 日本東洋医学雑誌

ABSTRACT

In many clinical studies to evaluate the efficacy of Kampo prescriptions, subjects are selected because they have high or low values of the characteristics of interest, such as blood pressure or serum cholesterol, and the data obtained from the subjects before and after treatment are analyzed. In such analyses, we should take into account the possible changes in the observations due to regression toward the mean, otherwise the results of the analyses will be incorrect.<br>In this paper, we consider the effect of regression toward the mean in analyzing data and propose a method of analysis with Akaike Information Criterion (AIC) to avoid incorrect results.