Activity Number:
|
269
|
Type:
|
Contributed
|
Date/Time:
|
Tuesday, August 4, 2009 : 8:30 AM to 10:20 AM
|
Sponsor:
|
Section on Risk Analysis
|
Abstract - #304941 |
Title:
|
Interpreting Differential Effects in Light of Fundamental Statistical Tendencies
|
Author(s):
|
James P. Scanlan*+
|
Companies:
|
James P. Scanlan, Attorney at Law
|
Address:
|
1527 30th Street NW, Washington, DC, 20007,
|
Keywords:
|
subgroup analyses ; differential effects ; interaction ; baseline rates
|
Abstract:
|
Subgroup analyses are subject of increasing discussion in epidemiology. Overlooked in the discussions is the statistical basis for expecting interventions to yield different proportionate changes in rates of experiencing an outcome for groups with different baseline rates. This presentation would explain the statistical basis for expecting an intervention that reduces an outcome rate to cause the group with the lower baseline rate to experience a larger proportionate reduction in the outcome while causing the comparison group to experience a larger proportionate increase in experiencing the opposite outcome. It will also address the way that expectation comports with observed patterns of differential effects in individual studies and meta-analyses of subgroup effects and the implications of the expectation for the interpretation of seeming differential effects.
|