232 Discrete Event Simulation for Health Technology Assessment
population so that it matches the clinical trial’s reference. This is done by
deriving the likelihood that a simulated patient would have been included
in the reference arm of the incompatible clinical trial and then using these
probabilities as weights to reconstitute the simulated population, which will
now match, on average, the clinical trial’s reference population. The simu-
lated results with treatment S are obtained for this reweighted population
and compared to what was observed for treatment E in the incompatible
clinical trial. This estimate can then be used to represent (Δ
E–S
) and derive
the course with treatment E in the simulation.
6.6.2 Joining Time-to-Event Distribution ...