|
Activity Number:
|
328
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Tuesday, August 8, 2006 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Statistics in Epidemiology
|
| Abstract - #305986 |
|
Title:
|
Using Multiple Imputation To Improve Race-Specific Disease Rate Reporting in a National Active Surveillance System
|
|
Author(s):
|
Elizabeth R. Zell*+
|
|
Companies:
|
Centers for Disease Control and Prevention
|
|
Address:
|
1600 Clifton Road, Atlanta, GA, 30333,
|
|
Keywords:
|
incidence rate ; racial disparities ; multiple imputation ; disease surveillance
|
|
Abstract:
|
Reducing health disparities is a Healthy People 2010 objective. CDC's Active Bacterial Core surveillance (ABCs) monitors racial disparities among invasive meningococcal and pneumococcal infections in the U.S. Race data are missing in 15% of reported ABCs cases. We have explored multiple imputation models which showed improvement over simple methods to account for unknown race. ABCs is an evolving system that adds variables over time to improve disease monitoring and for which specific variables included in case reports often varies among participating sites. A more complex approach to assigning missing race values is needed. To improve usefulness for larger epidemiologic evaluations we explore several imputation models. We evaluate multiple imputation using the multivariate normal and sequential regression multivariate models. A comparison of incidence rates by race will be presented.
|
