|
Activity Number:
|
25
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Sunday, July 29, 2007 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Biopharmaceutical Section
|
| Abstract - #309492 |
|
Title:
|
Developing Statistical Methods for Setting Specification Limits
|
|
Author(s):
|
Michelle Quinlan*+ and Erin Blankenship and Walt Stroup and James Schwenke
|
|
Companies:
|
University of Nebraska-Lincoln and University of Nebraska-Lincoln and University of Nebraska-Lincoln and Boehringer Ingelheim Pharmaceuticals, Inc.
|
|
Address:
|
1601 N 35th St, Lincoln, NE, 68503,
|
|
Keywords:
|
shelf life ; specification limit ; confidence/prediction interval ; random effect
|
|
Abstract:
|
A product's shelf life can be estimated from the time of first intersection between the specification limit for a stability limiting characteristic of a pharmaceutical product and an agreed upon interval estimate. In addition to defining a shelf life, a specification limit also defines bounds for future batches to confirm a product's shelf life. Any statistical methodology developed to facilitate setting specification limits must provide a reliable estimate of future batches. In addition, when multiple batches are used to support setting specification limits, controversy exists as to whether batches should be considered fixed or random effects. After assessing alternative interval estimate approaches, the goal is to develop multi-batch (fixed/random) extensions and compare the alternatives via simulation to determine which yield interval estimates suitable to shelf life estimation.
|
