In their recent article, Payne and colleagues [1] began the critically important process of seeking to define what constitutes appropriate care for persons in the persistent vegetative state (PVS). Although the opinions of physicians alone are insufficient to determine a societal consensus, finding agreement among a group of professionals exposed to the clinical realities of severely impaired patients is an important first step. Although the authors carefully defined PVS in their questionnaire, I remain concerned that some of the respondents, particularly nursing home directors, may have answered questions about the suitability of various medical interventions with the severely demented nursing home resident in mind. Working in a long-term care facility, I frequently see physicians use the everyday rather than the technical meaning of "vegetative" in describing their patients. By these far more casual criteria, a person whose vocabulary is limited to a handful of words, who responds to stimuli by groaning or pushing away the aide who tries to feed her, and who spends most of her days sitting in the hallway staring in front of her is considered "vegetative." The investigators were careful to define PVS as a state of unconsciousness, but the nursing home directors' surprising enthusiasm for recommending dialysis, chemotherapy, and other interventions suggests that they may have been applying their own definition of the term. The similarity of views between the neurologists, who are likely to accept the narrow technical meaning of the PVS, and the nursing home directors may be real or may reflect differing conceptions of vegetative patients.

To be sure that all the respondents really were referring to the PVS rather than to advanced dementia, the authors need to verify their respondents' understanding of the terms used. If the two groups are truly concordant, then the study raises another question: How much uniformity is required to constitute consensus? Unlike the authors, who state that strong agreement exists among the physicians sampled, I suggest that the fact that 60% to 80% favored withholding treatment is insufficient to establish consensus. Although 99% agreement is unlikely, a simple majority is probably not adequate. To persuade the public that limitation of care is appropriate in PVS (and to later extend the discussion to dementia), we will need far more than a simple majority of doctors agreeing on a reasonable standard.