Drs. Patrick, Lindquist, Kafka, and Rifas lament the assault against the longitudinal, personal nature of the physician-patient relationship and the associated obligation of nonabandonment for managed care and the progressive corporatization of medicine. Physicians must become actively involved in the formation of these emerging systems and advocate for quality and continuity of care, lest the essence of our profession become lost in this economic transformation [1].

Dr. Cohen suggests substituting a "commitment to caring" for nonabandonment because it is a more positive, intuitively appealing concept. Although caring about and caring for patients are essential elements [2], they do not capture the responsibility for continuity and joint problem solving in the face of uncertainty. "Continued clinical presence" comes close but is too detached. The anathema of abandoning the sick resonates deeply with the public and within the profession, for there is nothing more terrifying and clearly wrong than forcing patients to face severe illness and death alone.

We agree with Dr. Engel that the experiential, behavioral, and phenomenologic aspects to the physician-patient relationship need deeper exploration and that moral decision making should not simply be guided by what "feels right." However, the effect of medical decisions guided exclusively by abstract ethical principles on the lives of real patients also needs much more exploration. For example, Dr. Levy made assumptions about the significance of a specific medical intervention (that is, that artificial feeding has no utility when someone is dying of cancer) based on his beliefs and reading of the literature. However, this intervention enables our patient to spend an extraordinary month at home on hospice when she had no other ability to eat or drink. Later, when her symptoms became overwhelming, and she was more fully prepared to die, she could stop this treatment, allow high-dose opioids to relieve her pain, and die relatively peacefully. Simplistic assumptions about rules, utilities, or principles often do not allow such personally tailored decision making.

Dr. DeHart wrongly suggests that we are taking responsibility for disease and death itself. We have no intention of oversimplifying the process of dying or devaluing our many colleagues who actively struggle to comfort their patients. Yet, death need not be dominated by "disintegration and humiliation," nor must it be approached passively by physicians and patients. Using modern palliative methods, we can relieve much of the pain and other symptoms that accompany dying, thereby freeing patients to work more on their emotional, spiritual, and interpersonal issues. Although "holding hands, wiping brows, and sharing tears" are essential caring elements, there is also a medical obligation to relieve symptoms and to jointly face the many unpredictable, uncertain challenges posed by the patient's clinical situation and individual values.

Dr. Prechter sheds light on the inherent tension in our suggestion that physicians should carefully examine, but not violate, their fundamental values to be responsive to their suffering patients. Modern medicine is searching for a balance between our dual obligations to protect the sanctity of life and to relieve human suffering. We must design an ethical system that is responsive to the values and clinical circumstances of suffering persons, while at the same time respect our professional and societal traditions. We believe the obligation not to abandon those whose need is greatest should be a central element in these deliberations.