|
Activity Number:
|
35
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Sunday, July 31, 2016 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Statistics in Epidemiology
|
|
Abstract #318992
|
|
|
Title:
|
A Statistical Framework for Using External Information in Updating Prediction Models with New Biomarker Measures
|
|
Author(s):
|
Wenting Cheng* and Jeremy M. G. Taylor and Bhramar Mukherjee
|
|
Companies:
|
University of Michigan and University of Michigan and University of Michigan
|
|
Keywords:
|
Constrained estimation ;
Bayesian methods
|
|
Abstract:
|
Prediction models are abundant in the clinical and epidemiologic literature. While these prediction models may be well established, new biomarkers are being proposed to add to these models for improved personalized prediction. The information from an existing prediction model could be available in the form of coefficient estimates (with or without measures of standard errors) or individual predicted probabilities. We pose a principled framework to incorporate such types of information while building a new prediction model that adds new candidate biomarkers to the existing model. The general premise is that this candidate biomarker is measured on a small group of subjects while the existing prediction model has been validated in large cohort studies. We formulate the problem in an inferential framework where the information is translated in terms of non-linear constraints on the parameter space of the new model. We develop both frequentist and Bayes solutions to this problem. Simulation results indicate that information from the established model can substantially improve the efficiency of model coefficients estimation and enhance the predictive power in the new model of interest.
|
Authors who are presenting talks have a * after their name.
