Event-driven proportional-hazards planning
Estimate required events for a two-sided log-rank design using the Schoenfeld approximation.
Your entries stay in this page’s memory and are not submitted to us. Downloads are generated locally. Reloading or leaving the page clears your entries. Do not enter patient information.
HR must be positive and different from one; allocation must be inside 0–1, alpha 0.001–0.2 and power above 0.5 and below 0.999.
The starting values are hypothetical. Results update as you edit. Use aggregate or synthetic inputs only.
Calculation results
- Unrounded required events
- 246.7871
- Required events rounded up
- 247
- Allocation fraction
- 0.5
How to interpret the output
Required events are (z(1−alpha/2)+z(power))²/[p(1−p)log(HR)²], rounded upward. This large-sample approximation assumes proportional hazards and a fixed design. It does not determine participants without accrual, censoring and event-rate assumptions; unequal censoring and nonproportional hazards need fuller simulation.
Method referenceSource checked 2026-11-06. Read the reproducible calculator validation report.