The consequences of uncertainty. An empirical approach to medical decision making in neonatal intensive care.

Abstract

The policy of the Stanford University Medical Center on care for extremely premature neonates has been generally to initiate intensive care "based on a nonprobabilistic paradigm with a goal of saving every infant's life." Fischer and Stevenson analyzed the mortality rates, weights, gestational ages, and total costs of care for a sample of 68 low birth weight neonates. They also examined the outcomes of care in terms of degree of disability for the 24 surviving infants. They concluded that the existing policy resulted in undue suffering and in a substantial economic cost for the suffering and loss of life. They advocate use of an "individualized prognostic strategy" that takes into account the interplay of statistical prediction, the physician's assessment of clinical course, and parental responses when physicians suggest discontinuing aggressive care.