Online Program Home
  My Program

Abstract Details

Activity Number: 494 - Statistical Methodologies for Identifying, Modeling, and Managing Subpopulations at Risk
Type: Topic Contributed
Date/Time: Wednesday, August 2, 2017 : 10:30 AM to 12:20 PM
Sponsor: Section on Statistics in Epidemiology
Abstract #323375 View Presentation
Title: Data Analysis Techniques for Large and Changing Populations
Author(s): Jason Lee*
Companies: Johns Hopkins University Applied Physics Laboratory
Keywords: repeated measures ; multiple comparisons ; regression ; mixed models ; value-based care
Abstract:

Value-based care proposes to maximize value for patients by achieving the best quality care while lowering cost. Determining the effect of value-based care and its components on health outcomes has not been studied from a systematic and longitudinal perspective. However, many of the statistical and economics methods in use today (specifically, those examining changes over time with repeated measures) may be biased, inaccurate, or limited by data quality or missingness. These problems could be exacerbated if value-based care is implemented in a national health system, which offers geographic, volume, and condition variation. We are therefore exploring new and innovative study designs and statistical analysis techniques for use in a large, diverse, and changing population. We believe proper use of design and analysis techniques will minimize bias and produce more accurate and precise measurements, especially over many outcomes and times. This discussion will cover the basis of value-based care, its potential application within data-rich populations and health systems, and the statistical methodology and considerations used to develop statistical models and test program effects.


Authors who are presenting talks have a * after their name.

Back to the full JSM 2017 program

 
 
Copyright © American Statistical Association