The dominant approach to futility in medicine assumes that the probabi
lity and utility of medical interventions may be separated to provide
a quantitative (probabilistic) definition of futility. This assumption
is not only misleading but also responsible for much of the confusion
that futility has engendered in medical discussions. The divorce of u
tility from probability is the opposite of how clinicians reason: an i
mprobable intervention looks different if it is cheap, easy, and witho
ut morbidity than if it is technology intensive, expensive, and likely
to involve great pain and suffering. Futility is how physicians descr
ibe the sense of being compelled to proceed with resource intensive ca
re for marginal benefits. Outside the intensive care unit, physicians
weigh and sometimes reject patient requests without the need to invoke
futility. By examining the ways that physicians can legitimately eval
uate patient requests, we can show that appeals to futility are both u
nnecessary and counterproductive. In cases where such appeals are unav
oidable, the outpatient model suggests a process to adjudicate the com
peting claims of patient autonomy and physician responsibility.