PLACE:  319 Snedecor

SPEAKER:
Terry M. Therneau
Mayo Clinic

TITLE:
Random Effects Survival Models for Familial Data

ABSTRACT:

We are currently dealing with a large study investigating both genetic and environmental risk factors for breast cancer. The principle study data consists of the descendents and relatives of 426 incident cancer cases identified between 1940 and 1952. Full family pedigrees have been evaluated through 1992, currently there are data on 11848 women of which 5185 are marry-ins. Adjustment for familial genetic effects is an important part of the investigation of other risk factors such as oral contraceptive use, obesity, or smoking. The size of
the data set can make this challenging, however. We have relied primarily on our S-plus implementation of random effects Cox and accelerated failure time models.  The talk will focus on some of the computational and algorithmic challenges in implementing these, along with the results and benefits/problems of this type of model for such data. Liberal use will be made of examples from the above and selected other data sets.
 

COFFEE: 3:45 p.m., 104 Snedecor