|
Activity Number:
|
278
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Tuesday, July 31, 2007 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Health Policy Statistics
|
| Abstract - #309983 |
|
Title:
|
Evaluation of Using Life Table Methods Combined with Microsimulation To Produce Estimates of Distributions
|
|
Author(s):
|
James Lubitz+ and Liming Cai*+ and Nathaniel Schenker and Paula Diehr
|
|
Companies:
|
National Center for Health Statistics and National Center for Health Statistics and National Center for Health Statistics and University of Washington
|
|
Address:
|
3311 Toledo Road, Hyattsville, MD, 20782, 3311 Toledo Road, Hyattsville, MD, 20782,
|
|
Keywords:
|
semi-Markov process ; multi-state life table ; healthy life expectancy ; Aging
|
|
Abstract:
|
Life table methods have been widely used to estimate healthy life expectancy (HLE). But in almost all of the studies the estimates have been averages, not distributions. In this study, we apply life table methods and micro-simulation to the Cardiovascular Health Study (CHS) data to produce estimates of the distribution of the 65 and over population on years with functional limitations and years in fair or poor self-reported health states. We compare the distributions from the simulated population with the distribution of the actual sample on the same measures to provide insights into whether, and to what extent, simulated populations may be used to estimate population distributions. We estimate two life table models: the multistate life table model and the semi-Markov process model estimated using the stochastic EM algorithm (SMP-EM).
|
