a fair statement of the odds that a rational, well-informed person would give or take for the outcome of an experiment. The experiment may be unique and not rationally understood (precluding both theoretically sound predication and empiric experience). The formulation is applicable to experiments that have been carried out but the outcome unknown. (For instance, a certain statement about the gender of the fetus early in pregnancy is established but perhaps not accessible until amniocentesis can be done.) Unlike personal probably, the subjective probability should be the same from all competent counselors in possession of the same evidence.
Search Stedman's Medical Dictionary
Examples: glitazone, GI cocktail, etc.