Abstract #301484

This is the preliminary program for the 2003 Joint Statistical Meetings in San Francisco, California. Currently included in this program is the "technical" program, schedule of invited, topic contributed, regular contributed and poster sessions; Continuing Education courses (August 2-5, 2003); and Committee and Business Meetings. This on-line program will be updated frequently to reflect the most current revisions.

To View the Program:
You may choose to view all activities of the program or just parts of it at any one time. All activities are arranged by date and time.

The views expressed here are those of the individual authors
and not necessarily those of the ASA or its board, officers, or staff.

Back to main JSM 2003 Program page
JSM 2003 Abstract #301484
Activity Number: 178
Type: Contributed
Date/Time: Monday, August 4, 2003 : 2:00 PM to 3:50 PM
Sponsor: Biometrics Section
Abstract - #301484
Title: Combined Multiple Imputation and Sensitivity Analysis to Test for Influence of Imbalance in Follow-up Time and Dependent Censoring in Observational Data
Author(s): Karl Mertens*+ and Els Goetghebeur and Stijn Vansteelandt and Carl Suetens and Herman Van Oyen
Companies: Scientific Institute Public Health and University of Ghent and Ghent University and Scientific Institute Public Health and Scientific Institute Public Health
Address: Rue J Wytsmanstraaat 14, Brussels, , B1050, Belgium
Keywords: survival analysis ; dependent censoring ; sensitivity analysis ; imputation
Abstract:

When trying to estimate the effect of nosocomial pneumonia (NP, a hospital infection) on mortality in intensive care units, specific data dependencies seem to invalidate relative risk estimates from logistic regression (LR) and Cox proportional hazards analysis. Main problems include the effect of NP onset on time of discharge from ICU which causes differential follow-up among NP and non-NP patients, and the effect of time of censoring (defined as when being discharged alive from ICU) on survival time which is called dependent censoring. The following strategy is used to check for the possible effect of these problems: (1) multiple imputation (MI) tries to correct for the imbalance in follow-up time, by augmenting the follow-up time of censored patients up to a fixed day D using cumulative hazards estimated in a survival analysis. (2) MI is combined with a sensitivity analysis, in which cumulative hazards used during MI are corrected so to deal with dependent censoring. In this study, relative risk estimates from a LR model are highly sensitive toward balancing follow-up time and specifying a dependent censoring effect.

  • 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 2003 program

JSM 2003 For information, contact meetings@amstat.org or phone (703) 684-1221. If you have questions about the Continuing Education program, please contact the Education Department.
Revised March 2003