Related Experiment Videos
Methods for comparing cumulative hazard functions in a semi-proportional hazard model
D M Dabrowska1, K A Doksum, N J Feduska
1Department of Biostatistics, School of Public Health, University of California, Los Angeles 90024-1772.
Statistics in Medicine
|August 1, 1992
Summary
New graphical methods analyze differences in log cumulative hazard functions for semi-proportional hazard models. These methods test for treatment-covariate interactions and main effects, illustrated with kidney transplant data.
Area of Science:
- Biostatistics
- Survival Analysis
- Medical Statistics
Background:
- Semi-proportional hazard models are extensions of proportional hazards models.
- These models accommodate time-dependent effects of covariates.
- Analyzing treatment-covariate interactions is crucial in medical research.
Purpose of the Study:
- To develop graphical methods for analyzing a two-group semi-proportional hazard model.
- To create confidence procedures and test statistics for interaction and main effects.
- To illustrate the application of these methods using real-world data.
Main Methods:
- Analysis of differences between log cumulative hazard functions.
- Development of graphical diagnostic tools.
- Construction of confidence intervals and hypothesis tests for model parameters.
Main Results:
- The proposed graphical methods provide insights into model fit and parameter effects.
- The developed statistical procedures allow for formal testing of interactions and main effects.
- Application to kidney transplant data demonstrated the utility of the methods.
Conclusions:
- Graphical analysis offers a valuable approach for understanding complex survival data.
- The developed methods facilitate the assessment of treatment effects in the presence of interactions.
- These techniques enhance the interpretation of semi-proportional hazard models in clinical studies.