Activity Number:
|
538
|
Type:
|
Contributed
|
Date/Time:
|
Thursday, August 2, 2007 : 10:30 AM to 12:20 PM
|
Sponsor:
|
Biopharmaceutical Section
|
Abstract - #308601 |
Title:
|
Negative Binomial Analysis of Hypoglycemia Rates in Diabetic Patients
|
Author(s):
|
Cory Heilmann*+
|
Companies:
|
Eli Lilly and Company
|
Address:
|
Lilly Corporate Center, Indianapolis, IN, 46285,
|
Keywords:
|
Negative binomial ; Count data ; Zero-inflation ; Power ; Poisson
|
Abstract:
|
Hypoglycemia rates in insulin using diabetic patients are difficult to quantitatively model due to the highly skewed nature of their distribution, as many episodes are experienced by a few patients while many patients do not experience any episodes. A common method of analyzing hypoglycemia rates is to model the rank transformed rates. However, a rank transformation dilutes the available information and complicates inference on the actual difference in rates. This study uses data from four clinical trials as well as simulation to test different models that analyze rates of hypoglycemia. Test statistics such as BIC from real data sets show superior fit of the negative binomial model to Poisson based models. Simulations from a variety of models show that negative binomial models have higher power than non-parametric models to detect differences without inflation of type I error rate.
|
- 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 2007 program |