Estimation and Model-Checking Methods for Recurrent Gap Time Data

Chiung-Yu Huang
Biostatistics Research Branch
National Institute of Allergy and Infectious Diseases
National Institutes of Health

Wednesday, November 4, 2009
3:30pm
Weaver-Densford Hall 2-120
Minneapolis Campus

Abstract:

In longitudinal studies, a subject may repeatedly experience recurrent events of the same type during follow-up. The Cox proportional hazards model is frequently used to assess covariate effects on the gaps between consecutive recurrent events. While various methods have been suggested for checking the appropriateness of the Cox model with univariate survival time data or clustered survival time data, these methods cannot be directly applied to recurrent gap time data. In this talk, I will review estimation procedures for recurrent gap time models. I will also describe a class of new graphical and numerical testing techniques, for which the basic building block is a class of averaged residuals processes derived by averaging the martingale-like residual processes from gap times within a subject. This maneuver is very general and suitable for various purposes of model fitting assessment. Given the dearth of statistical software, I will present a modified within-cluster resampling method that can be used to approximate existing and proposed methods. Numerical simulations will be presented to examine the finite-sample performance, and the proposed model-checking techniques are illustrated with data from the Danish Central Psychiatric Register (DCPR). (This is joint work with Xianghua Luo and Dean Follmann.)

A social tea will be held at 3:00 P.M. in A434 Mayo. All are Welcome.
For more details contact 612-624-4655 or see http://www.biostat.umn.edu/seminar_academic.html