The views expressed here are those of the individual authors and not necessarily those of the JSM sponsors, their officers, or their staff.
Abstract Details
Activity Number:
|
139
|
Type:
|
Contributed
|
Date/Time:
|
Monday, August 1, 2011 : 8:30 AM to 10:20 AM
|
Sponsor:
|
Section on Statistics in Epidemiology
|
Abstract - #302405 |
Title:
|
Risk Estimation for Non-Rare Events
|
Author(s):
|
Shailendra N. Banerjee*+
|
Companies:
|
Centers for Disease Control and Prevention
|
Address:
|
4770 Buford Highway, Atlanta, GA, 30341,
|
Keywords:
|
log-binomial ;
iterative estimation ;
converge ;
non-rare event
|
Abstract:
|
A fitted logistic regression model can estimate risks adjusted for confounders, but it is known to over-estimate the risks when the outcome event is common or non-rare with an incidence or prevalence rate of roughly 10% or more. One of the alternatives is the log-binomial model, which is a generalized linear model with log-link and binomial errors. Since, this model estimates relative risk and also because odds ratio does not approximate to relative risk for non-rare events, this model is more appropriate compared to logistic regression in this situation. However, one of the known problems with the use of this model is that the iterative estimation algorithm may fail to converge. We used logistic regression, log-binomial and Poisson regression models on a prostate dataset listed by Collett (1991), a non-rare event of 38% in this dataset. Both log-binomial and Poisson model gave smaller and narrower range of risks compared to that of logistic regression model. The log-binomial regression model also resulted in some non-convergent, out-of-bounds predicted probabilities. This will be further investigated with simulated and more observed data.
|
The address information is for the authors that have a + after their name.
Authors who are presenting talks have a * after their name.
Back to the full JSM 2011 program
|
2011 JSM Online Program Home
For information, contact jsm@amstat.org or phone (888) 231-3473.
If you have questions about the Continuing Education program, please contact the Education Department.