2010 Fiscal Year Final Research Report

Evaluation of validity of a surrogate variable based on the causal effect

Research Project

Project/Area Number	21700303
Research Category	Grant-in-Aid for Young Scientists (B)
Allocation Type	Single-year Grants
Research Field	Statistical science
Research Institution	Kyoto University
Principal Investigator	OBA koji Kyoto University, 北海道大学病院, 助教 (30422926)
Project Period (FY)	2009 – 2010
Keywords	医薬生物 / ゲノム統計解析
Research Abstract	The aim of this study was to propose the new method for evaluation of validity of a surrogate variable using the information of causal direct effect called Natural Direct Effect. In addition, we applied this proposed method to the large scale randomized controlled trial, CASE-J trial, and evaluated the surrogacy of blood pressure on the cardiovascular events. From the statistical point of view, it is meaningful that this study could extend the method of statistical validation for the surrogate endpoint to the case that a surrogate is a continuous and a true variable is repeated measurement. The results showed that the proposed method could evaluate the average natural direct and indirect effects through the achieved systolic blood pressure in the randomized controlled trial and a natural indirect effect through the systolic blood pressure reduction of about 2 mmHg was large for the cerebrovascular events out of cardiovascular events.