Annals of Internal Medicine, which attempts to represent internal medicine as a whole, has included the bright side [2] and the dark [3, 4] as the discipline struggles with these new, uncomfortable realities. Both sides are part of two articles in this issue of Annals, one by the Federated Council for Internal Medicine [5] and the other by Petersdorf and Goitein [6]. A common theme in these articles is the need for more generalists and the difficulties in getting them.

Discussions about the future of internal medicine usually have a relatively short time horizon, just a few years. The perspectives are political and economic (clinical privileges, payment reform, provider mix, autonomy, and other immediate questions); this approach emphasizes short-term effects far more than long-term ones [7]. We suggest that the political and economic changes are but expressions of more fundamental, long-term changes in demography, technology, and the social environment that will shape medicine and medical care in the decades to come.

The population is aging and it is the elderly who account for most medical care needs of the population. The elderly accumulate chronic, degenerative diseases that interact, increasing the complexity of care. Most physicians who manage these patients must consider their patients' medical problems in concert, not piecemeal [8], although the elderly will also need some specialized physicians and surgeons with the skill to repair body parts (for example, hearts, lungs, hips, and eyes) as they wear out.

Technologic advances—for example, in genetics and molecular biology—will make diagnosis and treatment easier and more powerful. Highly detailed and accurate diagnoses will become possible: Genetic mapping will disclose increased susceptibility to disease decades hence; computer-assisted imaging already can visualize internal anatomy without physically invading the body; and scans will display the chemical make-up of the body. With these advances, and more to come, it will be possible to describe patients' current health and predict their future with incredible clarity. Treatment of specific diseases will also be more effective. Some diseases that we now struggle to palliate will be cured—by gene transplantation, precisely targeted immunotherapy, or tailored drugs.

A consequence of advancing technology is that diagnosis and treatment are becoming less invasive. Diagnostic information and therapies that now require the skills of specialists—for example, endoscopy and angiography—will be gathered or done largely by technicians, and the facts (such as "The proximal left coronary artery shows 60% stenosis but the other coronaries are free of disease") will be made more directly accessible to less specialized physicians.

Computerized information is another technologic change that will reduce the need for the kinds of highly specialized physicians that now comprise a large proportion of the medical workforce, especially in internal medicine. An important role for subspecialists today is to carry and dispense complex information. For example, under the pressures of day-to-day practice, clinicians are grateful to find an infectious disease consultant to answer the question: "For my patient with Citrobacter septicemia, in this place and at this time, what is the antibiotic of choice and the best alternative?" In the future, such information will be readily available from computerized sources. So far, computers have not had much effect on day-to-day clinical decision making because most physicians are unused to them; present-day computers are usually stationary (on a desk top) whereas physicians move about, and the computers do not contain much clinically useful information in a readily accessible form. New generations of computers, soon to be available, will access vast amounts of information, sort it in a user-friendly fashion, and make it available at the bedside and in the clinic by means of palm-top computers and other technologies.

Despite the extraordinary power of recent discoveries in the biology of disease, medicine cannot limit its domain to biology, or it will define itself out of most of the action. Behaviors such as cigarette smoking, addiction, and violence account for most of the morbidity and mortality of people in modern societies. Much of the burden of suffering could be prevented or postponed by relatively simple changes in lifestyle, even without additional understanding of the biology of disease. Future physicians will have to be experts in communication and in helping patients to change their behaviors if they are to include the most important determinants of health within the boundaries of medicine.

Another major social movement is toward universal access to care and cost containment. With the growing conviction that everyone should have the care they need is the realization that society cannot pay more for medical care, let alone all effective care, for everyone. In the future, physicians will need to be far more sensitive to the social environment and the cost of care.

Thus, although highly specialized knowledge and skills will always be valued, they will be needed in fewer physicians than is now the case, as the main reasons for these kinds of physicians, technical skill and information transfer, are superseded.

What will be needed even more in the future are experts in the integration and weighing of complex clinical information. Someone will have to interpret a bewildering array of information, each part in relation to the whole, and make wise decisions. In doing so, it would be artificial to limit the scope of consideration to any one aspect of medicine, when all aspects of it—the biology of disease, the social and economic context, the individual patient's specific conditions and preferences—are important determinants of the end result. For example, what is done about genetic information predicting increased susceptibility to disease, which might be treated by gene transplant, must be decided in the context of other diseases that might otherwise shorten the patient's life or complicate the transplant, as well as the patient's preferences, and society's ability to pay for the procedure.

These new kinds of physicians will be needed not only in what are now called "primary care" settings but also in hospital wards and intensive care units. One could argue that as the complexity of care increases, whatever the setting, so does the need for this kind of general physician.

Make no mistake, this will not be a run-of-the-mill generalist, someone who simply did not have subspecialty training. The work of these physicians will be extraordinarily challenging. It will require expert clinicians with a strong grounding in the traditional subspecialties of medicine but also in newer disciplines such as clinical epidemiology, information management, behavioral medicine, clinical decision making, clinical economics, and ethics. The work will also require a high level of problem-solving skills. Although medical care should always be done with kindness and good sense, physicians will need far more than good intentions and a basic medical education to do this work well.

Many of the qualities of these new physicians are what internal medicine has been about all these years; they reflect its traditional values. If it aspires to a central role in the future, internal medicine begins with an advantage. It has emphasized scholarship and the care of patients with complex problems. Its members have been innovators in basing the practice of medicine on the results of clinical research, the social context, and cost containment.

But the new generalist will not be produced simply by carrying forward unchanged what internists have traditionally admired. These new physicians will be as concerned with management and treatment as with diagnosis, the classic interest of internists. They will know how to find and weigh information rather than to somehow commit it all to memory, as the "master clinicians" have done until now.

The role internal medicine plays in the future will depend on the direction the discipline itself takes. Internal medicine could choose to continue on its present course, toward even greater subspecialization, and as a consequence become a peripheral player in medical care, a federation of small disciplines made up of highly specialized physicians who are important to the care of a small number of patients and to teaching and research. Or internal medicine could take a leadership position for medicine as a whole, as it has in the past, by making systemic changes now to meet the future need for highly capable generalists.

If internal medicine decides it will not do this work, the work will still need to be done and it will be taken up by others. Ultimately, the needs of society will be met. The collective actions of the internal medicine community over the next few years will decide whether the discipline goes where it is most needed.