= Mixed panel and exact-death likelihood
{title2=$g_{rD}(t)=\sum_{s\text{ living}}p_{rs}(t)q_{sD}$}
A clinic-observed interval ending in a recorded state contributes a <transition probability>. If the next observation is exact entry into an <absorbing state> but the preceding state is unobserved, its <statistical probability density> is $\sum_{s\text{ living}}p_{rs}(t)q_{sD}$, not $p_{rD}(t)$. Multiplying these conditional contributions yields a <likelihood> for mixed panel and exact-event observations.
Back to article page
