|
Activity Number:
|
104
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, July 30, 2007 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Biometrics Section
|
| Abstract - #309103 |
|
Title:
|
Analysis of Longitudinally Measured, Left-Censored Biomarkers of Severe Sepsis with Dropout, and Death
|
|
Author(s):
|
M. Abdus Sattar*+ and Lisa Weissfeld
|
|
Companies:
|
University of Pittsburgh and University of Pittsburgh
|
|
Address:
|
120 Ruskin Ave, Pittsburgh, PA, 15213,
|
|
Keywords:
|
Longitudinal censored data ; tobit regression ; sensitivity analysis
|
|
Abstract:
|
Our goal was to address the problem of analyzing longitudinally measured left censored biomarker data when subjects are lost to data collection for multiple reasons. We used an un-weighted and weighted tobit regression method for interleukin-6 (IL-6) and interleukin-10 (IL-10) left censored biomarkers data; where the weight for the weighted method accounted for the various reasons of dropout. For IL-6, analysis of all of the predictors: race, steroid use, score of organ failure assessment (SOFA) and Pneumonia Severity Index (PSI), are significant though the standard errors of the estimated coefficients for the weighted tobit regression are smaller. For the IL-10 analysis, all of the above mentioned predictors are significant except PSI in the un-weighted tobit regression. The standard errors for the estimated regression coefficients are higher in the weighted tobit regression models.
|
- 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 |
