Benefit–risk workbench

Simplified MCDA for teaching

Endpoints in the analysis

Favourable effects

Unfavourable effects

Deterministic result · point estimates

Calculating…

Stochastic result · SMAA

Probability of being preferred

Weights

Deterministic results

Point estimates from the effects table, the value functions below and the current weights. No uncertainty is taken into account here.

Value profile

Partial value: 0 = worst end of the scale, 1 = best end. Contribution = weight × partial value; the contributions add up to the total value.

Sensitivity to one weight

How to read this result

A plain-language explanation for non-specialists. The numbers update with your choices.

Effects table and value functions

Set which direction is better and the scale range: the worst end scores 0, the best end 1, linear in between. Choose endpoints at the top of the page.

Teaching tool inspired by the drugis.org MCDA software: linear partial value functions, swing weights, SMAA with measurement uncertainty (beta or normal distributions from the effects table), 4,000 simulations with a fixed seed. Cases are taken from published EPAR effects tables; the example weights are illustrative only. Files opened here stay in your browser.