|
Activity Number:
|
424
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 9, 2006 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Biopharmaceutical Section
|
| Abstract - #306418 |
|
Title:
|
Using Stochastic Differential Equations for Imputation of Missing Values in Longitudinal Clinical Data
|
|
Author(s):
|
Naum Khutoryansky*+
|
|
Companies:
|
Novo Nordisk
|
|
Address:
|
100 College Road W., Princeton, NJ, 08540,
|
|
Keywords:
|
missing data ; imputation ; stochastic differential equations
|
|
Abstract:
|
Missing data are common in clinical trials. In longitudinal studies missing data are mostly related to dropouts. Some dropouts appear completely at random. The source for other dropouts is withdrawal from trials due to lack of efficacy. For the latter case standard analyses of the actual observed data can produce bias. This paper considers application of stochastic differential equations (SDE) for imputation of missing values for primary and secondary endpoints. Coefficients of the SDE (their expected values and variances) are estimated using available data at discrete time points. The next step is imputation of missing values (estimation of their expected values and variances) for each subject based on the SDE and available data for this subject. It is shown that there is a relationship between the SDE approach and the incremental mean imputation method introduced previously.
|
