July 1, 1989Statistics in Medicine
Bootstrap investigation of the stability of a cox regression model
Authors
Discussion
Loading...
Member takes
Implication
Randomized trial investigates model stability and predictive ability in patients with primary biliary cirrhosis, highlighting confidence interval differences.
Key Points
- The aim is to investigate the stability of a Cox proportional hazards regression model in a clinical trial setting.
- Conducted a bootstrap investigation using 100 samples from a trial of azathioprine versus placebo.
- Analyzed 17 candidate variables in stepwise Cox regression.
- Constructed bootstrap confidence intervals for two-year survival probabilities.
- The original variable selection was confirmed, with no additional important variables identified.
- Bootstrap confidence intervals for survival estimates were wider than those from the initial model.
Cite This Study
Altman et al. (1989) studied this question.