Keskeiset käsitteet
This paper proposes a new framework for selecting causal estimands in observational studies, particularly when facing limited overlap between treatment and control groups, aiming to balance statistical performance with preserving the intended target population.
Tilastot
The study was conducted between 1989-1994.
The data consists of 5735 individuals.
2184 individuals were treated (RHC applied within 24 hours of hospital admission).
3551 individuals were in the control group.