Online Program Home
My Program

Abstract Details

Activity Number: 470 - Biomarker Evaluation and Winning Student Papers on Medical Devices and Diagnostics
Type: Contributed
Date/Time: Wednesday, July 31, 2019 : 8:30 AM to 10:20 AM
Sponsor: Section on Medical Devices and Diagnostics
Abstract #304583
Title: Constructed Composite Response: a Framework for Constructing Targeted Latent Variables
Author(s): Christopher Barbour* and Mark Greenwood and Dominique Zosso and Bibiana Bielekova
Companies: National Institutes of Health and Montana State University and Montana State University and National Institute of Allergy and Infectious Diseses
Keywords: longitudinal latent factor development; clinical endpoint development; multiple sclerosis; dimension reduction

Clinical scales are diagnostic measurements used by clinicians to predict the presence of a disease or level of severity for a disease for individual patients. Scales such as these are often constructed from multiple outcome measures to create a combined metric that better measures the true underlying trait of interest than any of the original outcome measures. These methods typically focus on explaining cross-sectional variation in the outcomes using projections into one or more dimensions to define the combinations of variables. When the interest is in creating a scale that is sensitive to changes over time, developing it using cross-sectional data may not tune the projection to detect changes over time optimally. This research develops methodology for scale creation that is optimized to detect changes over time in longitudinal data. The proposed method, Constructed Composite Response (CCR), applied to a motivating dataset of multiple sclerosis (MS) patients emphasizes its strengths and highlights potential improvement. Future extensions are also discussed, such as inclusion of sparse-learning strategies.

Authors who are presenting talks have a * after their name.

Back to the full JSM 2019 program