Broadening the scope of advance planning for medical care will not be simple. Physicians will need to engage their patients in periodic discussions about the goals of care. Establishing these goals, in turn, will require that patients gain an understanding of their overall physical and mental condition, and, given that condition, an appreciation of their ability to tolerate medical intervention. Finally, patients will need to work with their physicians to determine how their values shape the choice of goals for treatment. All this talking is time-consuming, it uses no procedures, and it necessitates raising issues that physicians are notoriously poor at discussing. But it is a crucial form of preventive medicine, designed to avoid excessive and undesired interventions in the final years of life. Just as the health care system supports other forms of prevention, physicians, hospitals, and health maintenance organizations should endorse comprehensive advance planning.

Why More Planning?

The prevailing model for medical decision making in the competent individual is an outgrowth of the theory of informed consent: At the time of an acute illness, patients are apprised of their diagnosis, their prognosis, and the options for treatment. Physicians should discuss the pros and cons of the different alternatives, and patients must make a selection, perhaps guided by a specific recommendation from their physician [2]. Although this approach goes a long way toward undercutting the tradition of medical paternalism [3], it can be expected to work far better if the groundwork is laid through planning.

In a medical crisis, patients may be technically decision-capable but sufficiently distraught that decision making is not optimal. Studies of patient recall of the provisions of consent forms for surgery, for instance, document an astonishingly poor grasp of the risks of the proposed procedure. This phenomenon is often attributed to the technical language of such forms or to hasty explanations by harried surgical residents [4]. At least as plausible is the possibility that patients may not be emotionally capable of objectively weighing the benefits and burdens of proffered treatments. As a result, patients often defer to their physicians or family members, implicitly recognizing that they are not equipped to handle choices when they are ill and anxious.

The second difficulty with crisis decision making, and one that has received less attention, is that the focus at such moments is often on the short-term rather than on the long-term prognosis. The physician's immediate goal in the acute situation is typically to determine whether there is potentially reversible disease. Often the physician does not step back to consider the natural history of the patient's underlying disease with and without the proposed intervention. In a patient with cardiomyopathy and pulmonary edema, the decision about whether to move the patient to an intensive care unit and insert a Swan-Ganz catheter usually rests on whether it is likely to improve his or her chances of surviving the current episode. The larger context—the fact that the patient's underlying disease process is progressive and irreversible—is often ignored, although the patient might choose to forego intensive treatment if she fully understood that she was destined for repeated bouts of pulmonary edema, rather than an isolated episode.

A third limitation of the prevailing model of medical decision making is that it focuses exclusively on a single, well-defined disease. The ideal physician, as presented in the informed-consent literature, explains to his or her patient the nature, prognosis, and options for treatment of a particular problem [5]. Together, patient and physician select the strategy that is most consistent with the patient's personal risk/benefit calculus [6]. The extent to which patient–physician interactions achieve this ideal is probably limited—only a minority of physicians tell patients the rationale for diagnostic and therapeutic procedures, and they rarely discuss risks or alternatives [7]. But the problem is not merely the gap between theory and practice; the problem is that the ideal itself is inadequate. For elderly patients with multiple medical problems, knowledge of the prognosis of each individual problem is insufficient. To be able to make reasonable choices, patients need to understand the effect of the interaction of their various medical problems. For instance, frail patients need to realize that even if no one of their problems is life-threatening, their organ reserves are so limited that they are at risk for a cascade of catastrophes if faced with an added acute illness.

Finally, limiting treatment—foregoing burdensome but potentially life-prolonging interventions—is a major decision, not to be taken lightly. Acceptance of treatment limitation is most likely to take root if the soil has been prepared. Older persons often do not grasp the full significance of their medical condition, unprepared as many are to accept the ultimate necessity of death and to perceive their own frailty [8], and unfamiliar as many are with the realities of being old and ill. Time to grasp and process information about their own condition is critical to informed decision making.

How to Plan

Planning while well for the eventuality of being ill should have three parts [9]. First, patients need to learn something about their overall medical condition. Second, they need to come to understand what treatment would be like for them personally, given that condition. And finally, armed with an understanding of their own circumstances and of their capacity to withstand medical interventions, patients need to evaluate this information using their personal calculus of benefits and burdens to formulate broad goals for medical care.

The process of learning about their medical status could take the form of an annual progress report, occurring on the occasion of a general physical examination. For patients who see their primary physician frequently, the evaluation can be done over an extended period. Patients and their families should emerge from their assessment with a knowledge of whether they are robust, frail, demented, or dying. They ought to come to understand that they are robust if their medical conditions are stable and have relatively little effect on their daily lives—even if they have a long list of medical problems and a correspondingly long list of medications. At the other extreme, they are dying if they have an illness that is uniformly fatal within months—typically, but not exclusively, metastatic cancer. Congestive heart failure, cirrhosis, or chronic obstructive pulmonary disease, in their final stages, can have as grim a prognosis as metastatic cancer. Between the robust and the dying are the frail: both the physically frail and the cognitively impaired. Individuals with multiple progressive chronic illnesses need to understand that they are physically frail. They have a limited capacity to withstand acute illness, and if they survive, are often left with a significant decline in functional status and a concomitant deterioration in the quality of life. Those elderly persons with early dementia need to understand while they still can that they are mentally frail: They have cognitive impairment that transcends normal aging, and their impairment will ultimately result in their being bedbound and totally dependent.

To plan for future illness, patients need to become familiar not merely with their overall state of health in the absence of any new, acute problems, and not merely with their likely trajectory over the ensuing months or years, but also with what the experience of illness and treatment would be like for them. For although most younger patients are willing to put up with almost any degree of transient misery in exchange for a good chance of prolonging their lives, and some older patients are similarly inclined [10], many older persons are extremely reluctant to undertake burdensome treatments, understanding that the gains vouchsafed by these treatments—if any—will be of short duration [11]. Clearly, physicians cannot strive to describe what it would be like to experience the full gamut of possible illnesses, the effect of the modalities designed to diagnose them, and the interventions used to treat them—from strokes, magnetic resonance imaging, and endarterectomy to leukemia, bone marrow aspiration, and chemotherapy. But physicians can explain to patients that if they are robust, they are likely to tolerate the rigors of medical intervention as well as they did at a younger age. Age alone does not affect outcomes of treatment of most invasive treatments studied, from thrombolysis [12] to surgery [13], nor is treatment more painful or uncomfortable due to age alone. On the other hand, if the patients are dying, attempts at curative treatment are likely to produce added suffering and have little chance of affecting outcome. Although physicians must avoid depriving their patients of hope, an unrealistic appraisal—or no appraisal at all—may stimulate the desire for futile treatment. What dying patients need above all is the assurance that that they will not be abandoned and that every effort will be made to maximize quality of life in their remaining days. If they are frail, aggressive treatment is less likely to be successful and more likely to entail a protracted period of disability—perhaps weeks in the hospital and a discharge to a nursing home—than was the case before their frailty developed. If the patients are developing dementia, both they and their surrogates need to understand that treatment is apt to sharply increase their level of confusion [14], that aggressive intervention will probably require that they be physically or chemically restrained, and that this in turn will place them at risk for additional complications [15].

Equipped with this kind of understanding, elderly patients and their physicians can articulate general goals for treatment. These goals should serve as guidelines for treatment when illness develops, not as specific directives. Using the guidelines as a starting point, patients and their physicians will need to delineate a specific plan of treatment when faced with a concrete illness. Typical goals might include cure for the robust, comfort for the dying, and preservation of function for the frail.

Barriers to Planning

Broadening the scope of discussions about future treatment beyond situations of incompetence will be a challenge. Currently, only a minority of patients draw up a written directive to direct care in the event of incompetence, and even fewer discuss their preferences with their physicians [16]. Although the percentage of patients who complete advance directives is small, most of those who are asked state that they want to think about and plan for the future [17, 18]. Failure to complete a written advance directive may reflect the inadequacies of such documents—and a preference for relying on surrogates in the event of incompetence—rather than a lack of interest in engaging in a conversation about the direction of future care [19]. Patients who are interested in contemplating future incompetence should be all the more concerned with thinking about what lies in store if they remain competent, a far more likely scenario. If individuals feel strongly about influencing their treatment course when they are comatose or so demented that they cannot recognize their families—the conditions typically addressed by conventional advance directives—surely they will want to participate fully in decisions that affect them when they are alert and lucid.

Another difficulty with expanding the role of advance planning is that there are so many contingencies to consider. At least with planning for incompetence, only a handful of conditions need to be considered, principally sudden death, coma, dementia, and delirium. The varieties of life-threatening illness to which the elderly are susceptible fill entire textbooks of medicine, even excluding those sicknesses that deprive patients of their reasoning capacity. The number of possible diagnostic tests and treatment modalities is likewise enormous. But the point of the proposed planning is not to acquaint people with all the disasters that might befall them; rather, it is to prepare them for the inevitability of sickness by helping them establish where they are in the overall schema of their lives. The goal is to remind them, however painful this may be, that their lives are drawing to a close, and to help them begin to plot their course for the final lap of the journey.

A further barrier to extensive planning is physicians themselves. Physicians now only rarely bring up the subject of advance directives; how much more reluctant will they be to rehearse matters that their patients will ostensibly be able to discuss later? And if physicians currently fall short in explaining the natural history of an individual disease with their patients, how can they be expected to amplify their discussions to include an overall assessment of their patients' health status? First, patients and their advocates have a role to play. Patients can take the initiative by asking their physicians for such an assessment [20]. Educational outreach programs such as those promoting advance directives in community centers, congregate meal sites, and synagogues and churches, can stimulate interest in raising important questions at the time of a physical examination. Advocacy groups such as the Alzheimer's Association and Choices in Dying can disseminate information promoting additional planning. Such simple interventions as providing a brochure describing the durable power of attorney for health care resulted in an increase in designation of a health care proxy from 0.4% to 18.5% in one group of elderly patients recently discharged from a hospital [21]. Secondly, the health maintenance organizations, group practices, and hospitals in which physicians practice should develop structures to promote planning. This could include allotting time to primary care physicians to engage in assessment and planning, as well as a separate billing code to correspond to the time spent. Health maintenance organizations and group practices could offer incentives to physicians if they endorse and implement a program of broad planning with a majority of their elderly patients. Thirdly, educational efforts are necessary to stimulate interest in planning. Hospitals should devote a grand rounds to spelling out in detail what planning actually involves. At the residency and medical school levels, as well as in continuing education programs, trainees and practicing physicians can learn interactively about comprehensive planning through role playing exercises and videotaped interviews.

Significant change in the way physicians think about their elderly patients and in the way they practice will be necessary before patients can expect to receive a comprehensive analysis of their condition on a regular basis. But if the idea of broadening the scope of advance planning has merit, then physicians will gradually come around to implementing it in practice, just as physicians broach the subject of planning for incompetence far more often today than they did 10 years ago. Only if patients know where they stand and have some sense of what lies in store can they realistically plan for their future medical care and avoid burdensome, unwanted interventions in the twilight of life.