Estimating Incremental Value in Marker Combinations Using a Two-Stage Study Design

• Activity Number: 253
• Type: Contributed
• Date/Time: Monday, August 2, 2010 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Aasthaa Bansal*+ and Margaret Sullivan Pepe
• Companies: University of Washington and Fred Hutchinson Cancer Research Center
• Keywords: two-stage design ; diagnostic tests ; classification ; biomarkers ; ROC

Abstract

BACKGROUND: When an existing marker, X, does not have sufficient diagnostic accuracy on its own, a new marker, Y, may in combination with X improve performance. We investigate estimating the increment in ROC performance due to Y from a two-stage study with X measured for all subjects in a cohort and Y measured for a case-control subset. METHODS: We develop non-parametric and semi-parametric methods for estimation and evaluate the methodology using simulation studies. RESULTS: The methods are easy to implement and provide valid inference. Matching controls to cases can compromise estimation of incremental value in certain settings. Using a combination of matched and unmatched controls may provide a more efficient approach to estimation in such instances.