When data interpretation should not rely on the magnitude of P values: the example of ANDROMEDA SHOCK trial

Francesco Corradi, Guido Tavazzi, Gregorio Santori, Francesco Forfori