= Solution
A <frailty random variable> is an unobserved positive multiplicative risk factor. A proportional frailty model has
$$
h(t\mid U)=Uh_0(t),
\qquad
S(t\mid U)=e^{-UH_0(t)}.
$$
The scale of $U$ is not separately identifiable from $h_0$: multiplying $U$ by a constant and dividing $h_0$ by it leaves the model unchanged. We may therefore normalize $\mathbb EU=1$, which lets $h_0$ represent the mean initial hazard multiplier and makes relative frailty interpretable.
If $U\sim\operatorname{Exponential}(1)$ and $h_0(t)=\lambda$, then the <Laplace transform> of $U$ gives
$$
S(t)=\mathbb E[e^{-U\lambda t}]
=\frac1{1+\lambda t},
\qquad
h(t)=-\frac d{dt}\log S(t)
=\boxed{\frac\lambda{1+\lambda t}}.
$$
Thus $h(0)=\lambda$, while $h(t)\to0$ as $t\to\infty$. Survivors become increasingly enriched for low-frailty individuals.
Back to article page
