Bootstrap for the Case-Cohort Design in Survival Studies

• Activity Number: 479
• Type: Invited
• Date/Time: Wednesday, August 6, 2014 : 10:30 AM to 12:20 PM
• Sponsor: Biometrics Section
• Author(s): Yijian Huang*+
• Companies: Emory University
• Keywords: Confidence band ; Interval estimation ; Simple random sampling without replacement ; Two-stage sampling

Abstract

The case-cohort design facilitates economical investigation of risk factors in a large cohort study, with covariate data collected only from the cases and a simple random subcohort. Methods that accommodate the design have been developed for various semiparametric models, but most inference procedures are based on asymptotic distribution theory. Such inference can be cumbersome to derive and implement, and does not permit confidence band construction. While resampling is an obvious alternative, how to bootstrap is unclear with the two-stage sampling. We propose a novel and versatile nonparametric bootstrap to circumvent the complication. Theoretical justification and numerical assessment are provided for a number of procedures under the proportional hazards model.