|
Activity Number:
|
116
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, August 3, 2009 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Statistics in Epidemiology
|
| Abstract - #304863 |
|
Title:
|
Semiparametric Regression Splines Models for Detecting Effect Modification in Matched Case-Crossover Studies
|
|
Author(s):
|
Inyoung Kim*+ and Ho Kim
|
|
Companies:
|
Virginia Polytechnic Institute and State University and Seoul National University
|
|
Address:
|
Department of Statistics, Blacksburg, VA, 24061,
|
|
Keywords:
|
Bi-directional case-crossover ; Effect modification ; Regression splines ; Semiparametric model ; Varying coefficients model
|
|
Abstract:
|
Assessing and characterizing effect modification by stratum variable in matched case-crossover studies are of often interested to epidemiology studies. In this talk, we develop two semiparametric methods to understand both parametric and nonparametric behavior of effect modification by stratum variable. These methods are based on two semiparametric models: 1) regression splines varying coefficients model and 2) regression splines multiple interaction model. Simulation result shows that two approaches are comparable to each other. These methods can be used in any matched case-control studies and extended to multilevel effect modification case. We demonstrate the advantage of our approaches using an epidemiology example of 1-4 bi-directional case-crossover study of childhood aseptic meningitis associated with drinking water turbidity.
|
