Death or survival, which you measure may affect conclusions: A methodological study.

To increase transparency around this paradox, researchers reporting one outcome should note if considering the opposite outcome would yield different conclusions. When possible, researchers should also report or estimate underlying risks alongside effect measures.