|
Activity Number:
|
265
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, August 1, 2016 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Medical Devices and Diagnostics
|
|
Abstract #320896
|
|
|
Title:
|
Modeling Passively Collected Biomedical Data
|
|
Author(s):
|
Moumita Chakraborty* and Eric Laber and Ana-Maria Staicu
|
|
Companies:
|
and North Carolina State University and North Carolina State University
|
|
Keywords:
|
|
|
Abstract:
|
Passive data collection using wearable devices is an increasingly common mechanism for monitoring a patient's health status and evaluating intervention effectiveness in situ with minimal burden and cost. However, these data can be noisy and influenced by unobservable subject-specific characteristics which complicates estimation and inference. We propose a Random Effects Hidden Markov Model for a time series of passive measurements and describe conditions under which this model can be used to estimate the causal effect of an intervention applied at fixed time point during the observation period. We illustrate the proposed method using a veterinary study of arthritis medications in cats.
|
Authors who are presenting talks have a * after their name.
