Activity Number:
|
311
|
Type:
|
Contributed
|
Date/Time:
|
Wednesday, August 14, 2002 : 10:30 AM to 12:20 PM
|
Sponsor:
|
Section on Statistics in Epidemiology*
|
Abstract - #301106 |
Title:
|
The Optimal Strategy in Designing a Case-Control Study
|
Author(s):
|
Peng Liu*+
|
Affiliation(s):
|
Food and Drug Administration
|
Address:
|
5100 Paint Branch Parkway, College Park, Maryland, 20740, USA
|
Keywords:
|
Retrospective study ; Sample size ; Equivalent sample size ; Maximize power ; Minimize cost ; Unequal-size design
|
Abstract:
|
The nonlinear programming technique is adopted to determine the best design strategy for treating a non-pairwise matching case-control study with constraints. The graphical solution technique is presented for insight into the mechanisms of the optimization process. The optimization procedure can be summarized into four steps: 1.) graph the region of each constraint expressed by an inequality equation; 2.) determine the feasible solution region satisfied by all the constraints; 3.) graph a family of objective functions; and 4.) determine the optimum solution by the intersection between the objective function and the boundary of the feasible solution region or the critical point.
|