Patients have the right to decide whether to authorize cardiopulmonary resuscitation (CPR). Physicians should provide adequate information and help clarify preferences.
The usefulness of decision analysis was investigated in two convenience samples: 20 healthy outpatient volunteers and 35 audience members at medical ethics grand rounds. Subjects quantified their relative preferences (utilities) for the outcomes of cardiac arrest. First, they rated them on a linear scale. Second, they participated in hypothetical gambles in which they indicated how much they would risk to avoid each outcome. The investigator then calculated the overall expected utilities of the CPR and no-CPR strategies.
Subjects were able to complete both the gambles and the rating scale. Utilities derived by the two methods differed greatly. Subjects had strong aversions to an outcome of severe long-term brain damage and widely varying ratings of an outcome of a short period of intensive care followed by death (intensive care unit death). Because intensive care unit death is far more likely than long-term brain damage, its utility was the prime determinant of whether CPR or no-CPR had the higher calculated expected utility.
The methods of decision analysis showed promise as a means not only of informing patients about CPR but of helping them make rational choices. They also revealed the inadequacy of current data on the key outcome of intensive care unit death.(Arch Intern Med. 1995;155:513-521)