Identifying risk factors for treatment selection in NLST

Principal Investigator

Name
Stuart Baker

Institution
NCI

Position Title
Mathematical Statistician

Email
sb16i@nih.gov

About this CDAS Project

Study
NLST (Learn more about this study)

Project ID
NLST-15

Initial CDAS Request Approval
Apr 9, 2013

Title
Identifying risk factors for treatment selection in NLST

Summary
The goal of this project is use NLST data to extend methodology for combining risk factors to improve the identification of a possible promising subgroup with a higher benefit than on average. The method combines three previous methods: the adaptive signature design, benefit prediction via a difference in estimated risks between randomization groups, and the sliding window subpopulation treatment effect pattern plot. Special approaches are considered for survival data.

References:
Baker SG, Kramer BS, Sargent DJ, Bonetti M. Biomarkers, subgroup evaluation, and trial design. Discovery Medicine 2012, 13(70):187-192.

Baker SG. and Kramer BS. Towards precision medicine: a simple method for identifying and evaluating biomarkers for treatment selection using randomized trials. Under submission

Aims

To adapt my software and modify my methodology for use with NLST data. Some challenges include large numbers of controls and missing data.

Collaborators

Ping Hu
Barry Kramer