Ørnulf Borgan: Using cumulative sums of martingale residuals for model checking in nested case-control studies

Ørnulf Borgan (Department of Mathematics,UiO) gives a seminar

in room 107, 1st floor N.H. Abels House at 14:15 Tuesday September 23rd:

Using cumulative sums of martingale residuals for model checking in nested case-control studies  

Abstract:
Standard use of Cox's regression model for censored survival data requires collection of covariate information on all individuals in a cohort even when only a small fraction of them fail (i.e. die or get diseased). This may be very expensive, or even logistically impossible. Further, in biomarker studies, it will imply a waste of valuable biological material that one may want to save for future studies. For such situations, the nested case-control design offers a useful alternative. For this design, covariate information is collected only for the failing individuals (cases) and a small number of controls selected at random from those at risk at the cases' failure times. For cohort data, methods based on martingale residuals are useful for checking the fit of a Cox model. Similar methods have not been available for nested case-control data. In the talk, I will discuss how one may define martingale residuals for nested case-control data, and I will show how plots of cumulative sums of the martingale residuals can be used to check the fit of a Cox model. The plots may be obtained using available software.
 

Published Sep. 10, 2014 10:38 AM - Last modified Sep. 10, 2014 10:41 AM