|
Activity Number:
|
425
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 5, 2009 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Health Policy Statistics
|
| Abstract - #305598 |
|
Title:
|
Individual Heterogeneity and Individual Inefficiency
|
|
Author(s):
|
Zhehui Luo*+ and Jeremy W. Bray and Joseph C. Gardiner
|
|
Companies:
|
RTI International and RTI International and Michigan State University
|
|
Address:
|
3040 Cornwallis Rd, Research Triangle Park, NC, 27709,
|
|
Keywords:
|
stochastic frontier model ; random effects ; health production ; parametric ; nonlinear model ; inefficiency
|
|
Abstract:
|
We use the stochastic frontier model framework to demonstrate the importance of differentiating individual heterogeneity and individual inefficiency in producing health outcomes that are bounded between zero and a fixed upper limit. Our model is motivated and illustrated by the COMBINE trial for alcohol dependence where nine treatment arms are evaluated. Preliminary results indicate that naïve models may underestimate average treatment effects, and additive random effects models do not readily identify factors that distinguish subgroups. However, stochastic frontier models with individual inefficiency can produce meaningful and interpretable results.
|
