Articles
NPH technical background
Some technical background introduction to non-proportional hazards, such as Z-score, B-values, and canonical form.
- Non-proportional effect size in group sequential design
- Canonical joint distribution of Z-score and B-values under null and alternative hypothesis
- Computing bounds under non-constant treatment effect
- Numerical integration non-proportional effect size in group sequential design
- Approximating an arbitrary survival distribution
- Power evaluation with spending bounds
- Computing expected events by interval at risk
- Statistical information under null and alternative hypothesis
Comparision of multiple NPH tests
Comparison of power among different tests, such as regular logrank test, weighted logrank test with different weights (i.e., Fleming-Harrington, Magirr-Burman, etc.), and MaxCombo.
Designs with futility bounds
Designs with both efficacy and binding/unbinding futility bounds.
Designs by AHR
Develop designs for time-to-event endpoints by average hazard ratio.
Designs with binary endpoints
Develop designs for binary endpoints.