Medical Decisionmaking Under Uncertainty

Abstract

Should aggressive treatment of premature or low-birth-weight infants be undertaken when prognosis is uncertain? Doctors know how to treat many premature or low-birth-weight infants so that they grow up to lead normal lives. But there are some premature or low-birth-weight infants that doctors do not know how to treat. Given the current state of medical knowledge, these infants cannot be saved, and their premature death is inevitable. The question posed does not ask about either of these two classes of infants. In their case, prognosis is not uncertain. Within the uncertainty that the ineluctably attends all empirical knowl-edge, we know what will happen to them, both if they are, or are not, treated. Nor do they present any peculiar conceptual problem. For ex-ample, the only difference between infants in the first class and normal infants, is that caring for the former costs more. Similarly, infants in the second class also present no peculiar problem. If an infant is dying and doctors know only how to prolong its life by a few hours or days, treatment will be either experimental or given to reduce the pain. These are exactly the alternatives we face when it is an adult who is dying.