A note on prognostic accuracy evaluation of regression models applied to longitudinal autocorrelated binary data

<p class="western" style="margin-bottom: 0cm; line-height: 200%;" lang="en-US" align="JUSTIFY"><span><span><span><strong>Background: </strong></span></span></span><span><span><span>Focus o...

Full description

Saved in:
Bibliographic Details
Main Authors: Giulia Barbati, Alessio Farcomeni, Patrizio Pasqualetti, Gianfranco Sinagra, Massimo Bovenzi
Format: Article
Language:English
Published: Milano University Press 2014-11-01
Series:Epidemiology, Biostatistics and Public Health
Online Access:http://ebph.it/article/view/10003
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:<p class="western" style="margin-bottom: 0cm; line-height: 200%;" lang="en-US" align="JUSTIFY"><span><span><span><strong>Background: </strong></span></span></span><span><span><span>Focus of this work was on evaluating the prognostic accuracy of two approaches for modelling binary longitudinal outcomes, a Generalized Estimating Equation (GEE) and a likelihood based method, Marginalized Transition Model (MTM), in which a transition model is combined with a marginal generalized linear model describing the average response as a function of measured predictors.</span></span></span></p><p class="western" style="margin-bottom: 0cm; line-height: 200%;" lang="en-US" align="JUSTIFY"><span><span><span><strong>Methods: </strong></span></span></span><span><span>A</span></span><span><span><span> retrospective study on cardiovascular patients and a prospective study on sciatic pain were used to evaluate discrimination by computing the Area Under the Receiver-Operating-Characteristics curve, (AUC), the </span></span></span><span><span>Integrated Discrimination Improvement (IDI) and the Net Reclassification Improvement (NRI) at different time occasions. </span></span><span><span><span>Calibration was also evaluated. A simulation study was run in order to compare model’s performance in a context of a perfect knowledge of the data generating mechanism. </span></span></span></p><p class="western" style="margin-bottom: 0cm; line-height: 200%;" lang="en-US" align="JUSTIFY"><span><span><span><strong>Results: </strong></span></span></span><span><span><span>Similar regression coefficients estimates and comparable calibration were obtained; </span></span></span><span><span>an higher discrimination level for MTM was observed. No significant differences in calibration and MSE (Mean Square Error) emerged in the simulation study, that instead confirmed the MTM higher discrimination level. </span></span></p><p class="western" style="margin-bottom: 0cm; line-height: 200%;" lang="en-US" align="JUSTIFY"><span><span><span><strong>Conclusions: </strong></span></span></span><span><span>The choice </span></span><span><span><span>of the regression approach should depend on the scientific question being addressed, i.e. if the overall population-average and calibration or the subject-specific patterns and discrimination are the objectives of interest, and some recently proposed discrimination indices are useful in evaluating predictive accuracy also in a context of longitudinal studies. </span></span></span></p>
ISSN:2282-0930