|
Activity Number:
|
515
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Thursday, August 2, 2007 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Statistics in Epidemiology
|
| Abstract - #308324 |
|
Title:
|
Bayesian Hierarchical Models to Evaluate Translational Research: Connecticut Collaboration for Fall Prevention
|
|
Author(s):
|
Terrence Murphy*+ and Heather G. Allore and Mary Tinetti
|
|
Companies:
|
Yale University and Yale University and Yale University
|
|
Address:
|
6 Hunting Ridge, School of Medicine, Hamden, CT, 06518,
|
|
Keywords:
|
spatio-temporal analysis ; Bayesian hierarchical models ; translational research ; study design ; falls ; intervention
|
|
Abstract:
|
Second stage translational studies are necessary and difficult to evaluate. A quasi-experimental design is used to compare the rate of fall-related health care utilization of two geographically disparate areas in Connecticut to evaluate an intervention designed to reduce fall-related injuries among older persons. This evaluation examines only the two baseline years prior to intervention. The experimental units are zip code tabulation areas for which data is gathered from public health sources. We estimate associations with sex, age, treatment, year of study and treatment-by-year interaction using spatio-temporal Bayesian hierarchical models. After adjusting for spatial variability, we find no association between rate of utilization and intervention and no difference in the temporal patterns of the rate of the intervention and usual care study arms in the two year baseline period.
|
