Dr. Gillick's concern that survey respondents may have included severely demented patients in the category of the PVS is valid, especially given the fact that almost one third of respondents believe these patients experience pain and suffering and 13% believe they experience hunger and thirst. These similarities between the beliefs of neurologists and medical directors about patients in the PVS suggest that both groups were referring to the same category of patients. Our survey determined neither whether physicians are unfamiliar with the diagnostic criteria for the PVS nor whether they are familiar with the criteria but do not accept them.

The consensus to which we refer is our finding that fewer than 8% of respondents believe that cardiopulmonary resuscitation, mechanical ventilation, and dialysis should generally be provided. A separate question determined that more than 96% of respondents believe it is ethical to withhold or withdraw these therapies. Even with this remarkable degree of agreement, we are left with the question, To what extent should the beliefs shared by the majority of physicians be considered when practice guidelines and health care policies are established? As we stated in our article, our conclusion is that these questions must be addressed not only by physicians, but, more important, by society.

Dr. Stead's comments on the care of his sisters poignantly illustrate physician behavior that was recently empirically documented in the findings of the Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatment (SUPPORT) [1]. More tragically, perhaps a financial disincentive (that is, a switch to Medicaid payments) influenced the physician's decision to continue treatment despite the patient's living will. Dr. Cassell's editorial [2] on our article offers an insightful discussion about the confusion and incoherence of physicians' thoughts, judgments, and clinical decisions that may also contribute to such behavior.