Power Calculation and Simulation for Studies of Estimating Chronic Health Effect of PM2.5 Using Three-Level Hierarchical Random Effect Model

• Activity Number: 420
• Type: Contributed
• Date/Time: Tuesday, August 2, 2011 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistics and the Environment
• Author(s): Ayano Takeuchi*+
• Companies: The University of Tokyo
• Keywords: power calculation ; PM2.5 ; chronic effect ; cohort study ; three-level hierarchical model

Abstract

Last decade, it has been known that fine particulate matter (PM2.5) has some acute health effect from the result of various time series studies. In this time, we have decided to evaluate chronic influence of PM2.5 on the development of child pulmonary function. We plan to conduct cohort study in more than 10 regions of Japan, and measure lung function and any other covariates for elementary school student for four years. We plan to use three-level hierarchical random effect model for adjusting individual development of lung function, time-related covariates, subject-specific covariate, and community-specific covariate to estimate the chronic effect of PM2.5. It is necessary to calculate power and sample size, especially the balance of the number of community and the number of subjects within one community. I did power calculation and simulation study for these purpose. In conclusion, when I fixed the number of subjects, the power increases according as the number of regions increasing, and we have enough power to follow up more than 10 regions and total 1000 subjects.