Abstract
Decisions about cardiopulmonary resuscitation may be based on medical prognosis, quality of life and patients' choices. Low survival rates indicate its overuse. Although the concept of medical futility has limitations, several strong predictors of non-survival have been identified and prognostic indices developed. Early results indicate that consideration of resuscitation in the elderly should be very selective, and support "opt-in" policies. In this minority of patients, quality of life is the principal issue. This is subjective and best assessed by the individual in question. Patients' attitudes cannot be predicted reliably and surrogate decision-making is inadequate. Lay knowledge is poor. However, patients can use prognostic information to make rational choices. The majority welcome discussion of resuscitation and prefer this to be initiated by their doctors; many wish to decide for themselves. There is little evidence that this causes distress. The views of such patients, if competent, should be sought actively.