Risk difference
= Risk difference
{title2=$RD=p_1-p_0$}
The difference between two outcome <probabilities> over specified populations and follow-up periods. With intervention risk $p_1$ and comparison risk $p_0$, a beneficial reduction in an adverse outcome has $RD<0$ and absolute risk reduction $p_0-p_1$. A causal interpretation requires a suitable design or identification assumptions; a simple difference between selected observational groups is not automatically a treatment effect.