JSM 2015 Preliminary Program

Online Program Home
My Program

Abstract Details

Activity Number: 653
Type: Contributed
Date/Time: Thursday, August 13, 2015 : 8:30 AM to 10:20 AM
Sponsor: Biometrics Section
Abstract #317456 View Presentation
Title: Classification by Longitudinal Data with Latent Class Models
Author(s): Huijing Wang* and X. Joan Hu
Companies: Simon Fraser University and Simon Fraser University
Keywords: cancer survivorship research ; extended GEE ; mean-variance model ; medical cost ; physician visit frequency
Abstract:

A specific aim of a cancer survivorship program is to classify the survivors according to whether they are at the risk to the consequences of the initial cancer diagnosis. To achieve the goal, we adapt the GEE approach to analyze the longitudinal physician visit claims collected for a young cancer survivor cohort with latent class mean-variance models. The physician visit claims from a subset of the cohort, in which subjects are identified at the baseline and generally viewed as suffering the consequences, are used to quantify the characteristics for "at-risk" in physician visit frequency or cost. On the other hand, the claims from a selected group of the general population, matching the cohort in sex and birth-year, set the standard of "not-at-risk". These enable our approach easy to implement and efficient in inference, while enjoying the inherent robustness of the GEE approach. We validate the classification via a fitted latent class mean-variance model using another subset of the cohort, presenting late effects of the initial diagnosis during the follow-up period.


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

Back to the full JSM 2015 program





For program information, contact the JSM Registration Department or phone (888) 231-3473.

For Professional Development information, contact the Education Department.

The views expressed here are those of the individual authors and not necessarily those of the JSM sponsors, their officers, or their staff.

2015 JSM Online Program Home