Fox, Ellen MD
Editorials represent the opinions of the authors and THE JOURNAL and not those of the American Medical Association.
Despite a multitude of efforts in recent years to reform medical education, and despite the significant gains that have been achieved, some refractory problems remain. Two articles in this issue of THE JOURNAL describe persistent shortcomings in the present medical education system. Billings and Block  summarize the results of an extensive literature review of end-of-life education in medical schools, concluding that the training currently provided inadequately prepares students to care for dying patients. Christakis and Feudtner  describe anecdotal experiences of medical students and residents on inpatient services, illustrating how the transitory and time-pressured nature of that environment can have disturbing effects on trainees' ethical development. On the face of it, these two articles may seem to have little in common. On a deeper level, however, these articles share a common theme: Both can be construed as serious critiques of medical education's prevailing overemphasis on a curative model of medical care.
The term "curative model"-as well as related terms such as "cure-oriented medicine," "cure-over-care mentality," and "curative bias"-is commonly used in professional and public discourse as if its meaning were self-evident and universally understood. In fact, however, no definition of curative model has been widely accepted. For the sake of this discussion, let us define it as an approach to clinical medicine that narrowly focuses on the goal of cure. Let us further agree that in this context the word "cure" is used in what Pellegrino and Thomasma  call "a radical sense-to refer to the eradication of the cause of an illness or disease, to the radical interruption and reversal of the natural history of the disorder."
Although cure is unquestionably an appropriate goal of medicine, other goals are important as well: promoting health, preventing illness and injury, restoring functional capacity, avoiding premature death, relieving suffering, and caring for those who cannot be cured.  In its purest form, the curative model concentrates solely on the goal of cure and in the process neglects medicine's other goals.
Elaborating on this theoretical construct, the curative model is characterized by an inherent set of assumptions, attitudes, and values. For example, its way of thinking is distinctly analytic and rationalistic. Clinical concerns are approached as puzzles to be solved; clinical encounters are treated as occasions for scientific inquiry. Because the object of analysis is the disease process and not the patient, symptoms are treated as clues to diagnosis, instead of as phenomena that are themselves worthy of treatment. As Christakis and Feudtner put it, patients' revelations are valued "chiefly if they lead to a diagnosis or an effective treatment plan, as when important intimate details... 'crack' the case."
According to this model, effective cure is contingent on effective diagnosis and treatment. Diagnosis, in turn, is derived from pathophysiology, while treatment is derived from empirical research on clinical outcomes. As a result, the cure-oriented approach is highly invested in a scientific, biomedical-or what Christakis and Feudtner call a "medi-centric"-approach to clinical practice. With regard to diagnosis, the curative model values scientifically based data over other types of information. Laboratory test results and radiographic findings, for example, are generally trusted more than patient self-reports. Devalued is anything subjective, immeasurable, or unverifiable. With respect to treatment, the curative model deems a therapy to be "medically indicated" if it has been demonstrated through rigorous empirical studies to eradicate or at least to slow or halt the progression of disease. Thus, treatment is directed toward the underlying cause of illness rather than its outward manifestations. Treatments considered most effective are those that improve objective, disease-related outcomes such as tumor size or disease-free survival. Treatments that improve subjective and nonspecific outcomes, such as the patient's perceived quality of life, may be seen as less important.
The curative model also belittles phenomena that are not entirely explainable by science. And because brain physiology has not been elucidated in much detail, psychological factors may be regarded as trivial or even spurious. For example, symptoms that cannot be understood in pathophysiological terms are sometimes disparagingly described as "all in the patient's head." When the only goal is cure, facts become differentiated from feelings, and the body becomes dissociated from the mind.
Another manifestation of the curative model's scientific worldview is its tendency to perceive patients in terms of their component parts. Cure-oriented medicine reveres the "hard" medical sciences, which fractionate human beings into molecules, cells, and organ systems. It also reveres the clinical sciences, which divide the care of patients among various subspecialties. Conversely, the "soft" social sciences, which examine intact human beings and their complex relationships, are relegated to a lower, "nonmedical" status. The curative model's reverence for hard science can lead to patients being treated less as whole persons than as repositories for disease.
If medicine is limited to the goal of cure, then medical indications alone dictate which diagnostic tests and therapeutic treatments are appropriate. All medical decisions are therefore reduced to purely empirical questions. It follows, then, that two patients in identical clinical circumstances are treated exactly the same, so individual patient differences are insignificant. A cure-oriented physician has no real need to get to know patients, no need to understand their individual values, and no need to explore how illness affects their lives. In the curative model, medical ailments are all but separated from the persons who serve as their hosts.
The curative model is also typified by a hierarchical structuring of relationships among health care professionals, in which those who command the most biomedical knowledge (typically, physicians) also command the most authority. In the words of Christakis and Feudtner, "The hierarchy of rank also determines to what degree other staff members can question a certain decision and course of therapeutic action."
To the extent that cure is the overriding goal of medical care, clinical investigation can never be complete until the pathophysiology is fully understood. And once a definitive diagnosis is established, the curative model demands that diseases be treated. Thus, the model assumes not only that diagnosis and treatment are possible, but that they are also necessary and desired.
On a basic level, the curative model conflicts with the notion of a good death. Where cure is the ultimate goal, death is the ultimate failure, a symbol of medicine's inadequacy-hence, the "aggressive" treatment approach alluded to by Billings and Block. In the jargon of the curative model, patients whose diseases cannot be stopped or slowed are termed "untreatable" or "beyond help." From a cure-oriented perspective, the care of such patients is considered outside the purview of medicine.
Ultimately, it is overreliance on the curative model that underlies the neglect of palliative care in medical education. As Billings and Block explain:
"The culture of the academic health center remains highly invested in "aggressive" treatment of terminal disease and the notion that death represents medical failure, and tends to neglect psychosocial issues in illness. These messages constitute a "hidden curriculum" that undermines many of the attitudes that form the basis for effective and compassionate care of the dying."
It is also the curative model that drives the "efficient" residents described by Christakis and Feudtner to "turf" psychosocial tasks, avoid intimacy, embrace hierarchical structures, and "float" above commitment.
The article by Billings and Block reflects an alternative model to current medical care that has been labeled the "palliative care" approach. In many ways the curative model and the palliative care model are antithetical. Whereas the former assumes that cure is the primary goal, the latter assumes that cure is not a goal. The World Health Organization defines palliative care as "the active total care... of patients whose disease is not responsive to curative treatment."  Thus, the palliative care model supports a variety of goals, including relief of suffering, control of symptoms, and restoration of functional capacity. However, it supports neither the goal of cure nor the hastening or prolonging of death. 
Another difference is that the palliative care model is far more sympathetic than the curative model to subjective phenomena. For example, palliative care medicine is often intensely concerned with the treatment of pain, despite the fact that pain cannot be definitively verified and at times cannot even be explained. But palliative care considers the patient's subjective experience of illness to be as important as objective clinical data. Accordingly, it treats the amelioration of pain and other symptoms as a legitimate goal of medicine.
Furthermore, the palliative care approach is more tolerant of incomplete medical knowledge and accepting of medicine's limitations. Diagnosis is not a predetermined goal; rather, it is pursued only if so doing conforms to patient-determined goals of medical care. Similarly, a specific treatment is regarded as appropriate only if it is worthwhile from the patient's perspective. Moreover, the palliative model does not equate death with defeat. To the contrary, it accepts death as the natural conclusion of life. When death approaches, the response of palliative care medicine is to comfort, rather than to withdraw.
Perhaps the most important difference between the 2 models is that in palliative care, the management plan is tailored specifically to each patient, according to the patient's personal values and preferences. This approach contrasts with the curative model, in which the appropriate treatment is considered a scientific question that is best decided by the physician. Tailoring the care plan to the individual patient requires getting to know every patient as a person and understanding the unique concerns of each. Palliative care's emphasis is on comprehensive care of the whole patient, inclusive of psychological, social, cultural, ethical, and spiritual concerns. Attending to all of these needs demands diverse areas of expertise; interdisciplinary teamwork-not an authoritative hierarchy-is central to the model of palliative care.
Learning about a patient also depends on effective communication and relationship building. As a result, the palliative care model stresses the importance of humanistic qualities and interpersonal skills. Because getting to know a patient also takes time, the palliative care model is best implemented under conditions that allow for lengthy encounters over a long period. This explains why the efficiency-driven and time-limited setting of acute care hospitals, as described by Christakis and Feudtner, is virtually incompatible with a palliative care approach and instead reinforces the curative model.
No one model of medical care is ideal for all patients in all clinical circumstances. The most suitable model is that which best reflects an individual patient's needs and goals. Thus, the curative model might be properly applied to a previously healthy patient presenting with an acute, well-defined, and easily treatable medical condition. Usually there would be no pressing reason for a physician to learn intimate details about the life of a patient with a broken bone, acute appendicitis, or community-acquired pneumonia. For situations in which medicine is capable of restoring health completely, patients and physicians tend to agree that the overriding goal is cure: in such cases, the curative model is wholly appropriate. At the other extreme, the palliative care model is often a good fit for patients who are entering the final phase of a predictably progressive terminal illness. Over time, such patients typically accept their plight, and many elect to forgo all efforts at cure and to accept palliative care alone.
If all medical problems were imminently curable, a strong emphasis on the curative model might be desirable. There would be no overwhelming need for physicians to establish meaningful, long-term relationships with patients. Transient encounters would be perfectly satisfactory. Likewise, were the biomedical sciences powerless to change the natural history of disease, the curative model would not exist at all; medicine would consist entirely of palliative care.
In reality, relatively few patients present with curable conditions alone; most have chronic, disabling conditions as well. By the same token, patients with no reversible component to their disease constitute only a small subset of the population. For most patients, neither a purely curative model nor a purely palliative care model is altogether suitable.
It is a mistake to view the curative model and the palliative care model as the only 2 options available. Instead, the two models represent opposite ends of a spectrum in which limitless variations are possible, and leaping directly from 1 extreme to the other is seldom appropriate. Between the curative model and the palliative model lies an unnamed approach that supports all legitimate goals of medicine-health promotion, prevention, rehabilitation, life preservation, comfort, and care-and is willing to combine them in whatever manner best reflects the values of an individual patient. The flexibility of this approach makes it appropriate for all cases in which neither the curative model nor the palliative care model alone will do.
In theory, there is nothing wrong with focusing narrowly on a single goal of medicine, whether it be cure or palliation. In practice, however, overreliance on a single model can have unfortunate consequences. When the curative model dominates medical education, as it does today, its inherent assumptions, attitudes, and values tend to define the culture of academic medical centers. Trainees who are thoroughly immersed in such a cure-oriented milieu can hardly avoid being influenced by it.
The 2 articles contained in this issue support the claim that in medical education today, the curative model of medical care is overemphasized. Billings and Block support this claim indirectly by arguing that the current training in palliative care is inadequate. Moreover, they assert:
"Studies of undergraduate medical education indicate that gains in humanistic attitudes and skills... in the first 2 years of medical school are regularly reversed during the last 2 years of school and in residency training. Even when teaching emphasizes these humanistic competencies in the first 2 years, the clinical environment, especially on inpatient services, fails to support, and often undermines, these attitudes and practices."
In their article, Christakis and Feudtner describe multiple anecdotal examples of how the cure-directed approach dominates inpatient services and deleteriously affects trainees. They write:
"Yet this dominant mind-set pressures residents to attend only to those medical issues that necessitated a patient's admission in the first place and that must be addressed before the patient's discharge. Personal capacities to deal with psychosocial problems... are left underdeveloped."
Ideally, medical education's emphasis on a particular model of medical care should reflect the importance of that approach in clinical practice. Yet the current medical education system seems unbalanced, neglectful of palliative care and inordinately focused on cure. Admittedly, this observation is difficult to prove by rigorous quantitative methods, as "degree of cure orientation" among medical schools would be a difficult variable to measure. A few concrete figures might shed some light on the magnitude of the problem.
Although there is not a direct correlation between different models of medical care and different locations for health care delivery, the curative model is, as both the Billings and Block article and the Christakis and Feudtner article illustrate, especially conspicuous on inpatient services. In the United States today, a small proportion of medical care is delivered in such settings: only 13% of all physician-patient contacts occur in acute care hospitals.  The remaining 87% of contacts take place in other clinical settings, such as physicians' offices, ambulatory care clinics, intermediate-term and long-term care facilities, and hospice and home care environments.
In stark contrast to the proportion of medical care that is delivered in acute care hospitals is the proportion of training that is conducted there. Medical students spend between 60% and 82% of their required clerkship time on inpatient services (except during family practice rotations, in which only 10% of training occurs in inpatient settings).  Although aggregate data are not available for all graduate medical education programs, the proportion of time internal medicine residents spent in inpatient settings in 1991 ranged from 69% to 81%. 
Meanwhile, a significant proportion of patient care is delivered in settings that are more amenable than hospitals to a palliative care approach, such as long-term care facilities and hospice programs. Nevertheless, as Billings and Block report, "Nonacute and outpatient care, particularly home care and nursing home care, are underutilized in formal medical training."
The suggestion that students and residents should spend less time on inpatient rotations and more time in other settings is certainly not new. Neither is Christakis and Feudtner's suggestion that clinical training provide social resources that promote the ethical development of medical students and residents. Multiple proposals to reform education have included these suggestions as well as many others that would counteract the present overemphasis on curative care. 
Although progress has unquestionably been made, [7,8] certain aspects of medical education have been highly resistant to reform.  In any case, it is abundantly clear that previous reform efforts have not gone nearly far enough. The articles in this issue and the statistics above demonstrate a significant mismatch between the type of care taught in medical schools and the type of care actually delivered in clinical practice. The curative model of medical care predominates medical education in this country, and this is a problem. Additional reform of medical education is needed if other models of medical care are to receive the emphasis they deserve. Reframing the problem with medical education in terms of its relative emphasis on the various models of medical care may help to marshal resources to effect essential changes in medical school curricula and graduate medical education programs.
Ellen Fox, MD
Reprints: Ellen Fox, MD, Ethics Standards Division, American Medical Association, 515 N State St, Chicago, IL 60610 (e-mail: Ellen_Fox@ama-assn.org).
1. Billings JA, Block S. Palliative care in undergraduate medical education: status report of future directions. JAMA. 1997;278:733-738. Ovid Full Text Library Holdings Bibliographic Links [Context Link]
2. Christakis DA, Feudtner C. Temporary matters: the ethical consequences of transient social relationships in medical training. JAMA. 1997;278:739-743. Ovid Full Text Library Holdings Bibliographic Links [Context Link]
3. Pellegrino ED, Thomasma DC. Helping and Healing. Washington, DC: Georgetown University Press; 1997:27. [Context Link]
4. Callahan D, Parens E. The ends of medicine: shaping new goals. Bull N Y Acad Med. 1995;72:95-117. Library Holdings Bibliographic Links [Context Link]
5. World Health Organization. Cancer Pain Relief and Palliative Care: Report of a WHO Expert Committee. Geneva, Switzerland: World Health Organization; 1990. [Context Link]
6. National Center for Health Statistics. Vital Health Stat 10. 1995;10:113-114. [Context Link]
7. Barzansky B, Jonas HS, Etzel SI. Educational programs in US medical schools, 1995-1996. JAMA. 1996;276:714-719. Ovid Full Text Library Holdings Bibliographic Links [Context Link]
8. Andersen RM, Lyttle CS, Kohrman CH, Levey GS, Clements MM. National Study of Internal Medicine Manpower, XIX: trends in internal medicine residency training programs. Ann Intern Med. 1992;117:243-250. Library Holdings Bibliographic Links [Context Link]
9. Christakis NA. The similarity and frequency of proposals to reform US medical education: constant concerns. JAMA. 1995;274:706-711. Ovid Full Text Library Holdings Bibliographic Links [Context Link]
10. Anderson MB. Medical education in the United States and Canada revisited. Acad Med. 1993;68(suppl 6):S55-S63. Library Holdings Bibliographic Links [Context Link]
Education, Medical; Interpersonal Relations; Interprofessional Relations; Palliative Care; Terminal Care