Change the assumptions to estimate the number of trial participants needed.
The starting treatment effects are hypothetical. They are not measured effects of a drug.
Model, assumptions and sources
This reproduces the source workbook’s meta-analysis approach: a weighted mean of log hazard ratios, an independent-condition composite incidence, and an event-count formula divided by trial length and incidence. It preserves the workbook’s narrow, unequal hazard-ratio bounds, which affect the weights. These bounds are assumed, not measured confidence intervals.
Diseases are treated as independent and incidence as constant. The calculation does not account for dropout, competing risks or changing age. It is an illustration for exploring study design, not a validated trial plan. The source essay discusses other models with different results.
