|
Activity Number:
|
430
- Contributed Poster Presentations: Section on Statistical Consulting
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Tuesday, July 31, 2018 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Statistical Consulting
|
|
Abstract #330838
|
|
|
Title:
|
Comparison of Latent Variable Models with Black Box Methods to Classify Disease Flare
|
|
Author(s):
|
Jonathan Grotts*
|
|
Companies:
|
|
|
Keywords:
|
Patient reported outcomes;
longitudinal analysis;
Quality of life measures
|
|
Abstract:
|
Methods for identifying and predicting disease flare have utility for clinicians and their patients. Flare is often described ambiguously as a worsening of health from a baseline state and quantitative methods for examining patient flare are limited. We examined the ability of a longitudinal multivariate latent class model to identify patient flare when patients were asked to routinely respond to multiple pain scales. This model attempted to capture the correlation within patients and between patient reported pain scales. The latent class model was compared to black box methods for predicting flare. Comparison between flare prediction methods was done to understand the trade-off between incorporating flare etiology into a model and prediction performance. Patient reported flare status was used to assess model performance. Multivariate longitudinal data was from a study on pelvic pain management where data was collected every two weeks for a year.
|
Authors who are presenting talks have a * after their name.
