The care of seriously ill patients, particularly when cure is unlikely and costs are high, poses a difficult and delicate moral problem. It seems hard to satisfy ordinary standards of efficient resource allocation by spending money on care for those for whom comparatively little permanent good can be done, and it is easy to think of other sick people who might benefit from having more resources expended on their care.

The care of persons with serious mental illness—those with a significant degree of functional impairment and a continuing high level of need for mental health services—unfortunately presents just this problem. Patients with chronic schizophrenia, severe bipolar disease, and severe depression ordinarily fit that description. These conditions are treatable, but treatment is expensive. Significant individual variation exists among patients with serious mental illness, and generalization can be hazardous. However, such conditions are likely to bring acute suffering, are difficult to treat, and often do not offer hope of a permanently good outcome.

This paper examines the difficulties of using conventional utilitarian, cost-benefit, moral, and political arguments to justify allocation of resources to the care of persons with serious mental illness and offers an alternative approach to this problem based on consideration of the goals of medicine.

Patients with serious mental illness pose particularly troubling problems of resource allocation that can be readily understood by a brief look at the usual standards for fair and efficient allocation of resources. The usual standards hold that resources should be allocated to treatment of medical conditions for which effective treatments exist. The resources should be allocated in ways that ensure a good cost-benefit ratio, that is, a good balance between the money spent and the treatment benefits realized. There should also be a good cost-effectiveness ratio, which is achieved when the money spent treating a particular condition produces an outcome as good as or better than that produced by spending the same amount of money to treat some other condition. In general, resources should be deployed in a way that will maximize utility, or satisfy the principle of the greatest good for the greatest number.

Many available treatments for persons with serious mental illness would have considerable trouble satisfying those standards. Money spent on those with less serious conditions—mild neuroses, for example—could help more people per dollar spent and achieve good therapeutic results. Helping persons with serious mental illness at the expense of other, more treatable patients would not satisfy the principle of utility or produce a good cost-benefit or cost-effectiveness ratio.

On the face of it, then, it is difficult to make a persuasive case for spending much money on the care of persons with serious mental illness. Those who are less seriously ill, for whom more can be done and a cure more often effected, would seem to have the strongest prima facie claim.

Moral theory does not offer much immediate help here. The field of ethics continues to wrestle with the long-standing puzzle of how to compare benefits for a large number of persons who have relatively minor problems with the same or fewer benefits, at a much higher individual cost, for a much smaller group of persons with more serious problems. And how are we to compare the benefits of high-cost treatment for those for whom cure is not likely with the benefits of lower-cost treatment for those who might be cured? Such puzzles do not make it easy to determine either what is fair or what might count as most humane.

We might try to sidestep these seemingly intractable issues by opting for a political solution and simply putting the allocation problem before the people and their elected representatives and letting them choose whatever solution they find most appealing. But that could be a dangerous solution. If people chose their own self-interest as the norm—which is what they usually do in political struggles—they will quickly perceive that they are far more likely to benefit from money that would be available for treatment of mild mental health conditions than they are from money available for the most serious conditions.

In sum, there are many daunting obstacles to making a good moral case and an equally persuasive political case for allocating resources to the care for persons with serious mental illness. Why should money be spent on the care of those who can, only with great difficulty, satisfy standards of economic efficiency, and who will—if they are helped and resources are tight—probably be taking resources away from those who might be more decisively helped?

This paper offers a different way of approaching this problem by considering allocation of resources in light of the goals of medicine. My suggestion by no means will solve all of the difficulties involved in achieving a fair allocation of resources, but it may at least provide a central place in our moral and political deliberations about care of persons with serious mental illness.

A recent Hastings Center project on setting new priorities for medicine proposed and explicated four goals of contemporary medicine (1):

• The prevention of disease and injury and the promotion and maintenance of health

• The relief of pain and suffering caused by maladies

• The care and cure of those with a malady and the care of those who cannot be cured

• The avoidance of premature death and the pursuit of a peaceful death.

The group that fashioned these goals, drawing both on historical and on contemporary debates, adopted a perhaps surprising overarching principle—that no one of these goals takes logical priority over or is more important than any other goal. On the contrary, the group argued that the goals of medicine should be dependent on context, both in guiding individual patients' care and in determining the aims of health care systems. For example, disease prevention should take a logical priority in the care of those who are well, but give way to other goals when people become sick. Similarly, the care of those who are sick calls for a different set of goals than does the care of the same persons when they are dying.

One purpose of this strategy is to emphasize that the variety of medical circumstances can call for different and shifting goals over time. Another important purpose is to displace the cure of illness and the saving of life from their de facto priority in most health care systems. They are important goals but are not, in principle, more central than the goals of relieving the suffering of those who cannot be cured or helping patients achieve a peaceful death. The care of the chronically ill and of those who are dying is given a status in principle perfectly on a par with the cure of illness and the saving of life.

This way of understanding the goals of medicine has an immediate and critical relevance for the care of the seriously mentally ill. In light of these goals, both the treatment of individual patients with serious mental illness and the allocation of resources for this patient group have the same high status as the care of those who can be cured. Thus persons with serious mental illness are no longer treated as second-class citizens, and their need for treatment resources is no longer displaced by the supposedly stronger claims of those who are potentially curable. In this way of thinking, persons with serious mental illness are to be treated as equals in the allocation of resources.

Some additional considerations are worth mentioning. All of us are at risk of contracting an incurable condition, physical or mental. Indeed, at the end of our life we will, by definition, have an incurable condition. Given that universal eventuality, allotting a higher medical status—in principle or often in practice—to cure makes no sense. When a cure can be achieved, it should be sought; then it is a reasonable primary goal. But when it cannot be achieved, it should be put out of mind, and attention should be focused on what is possible.

Sooner or later, for all of us, the potential for a cure will run out. We will then find ourselves in a situation analogous to that of persons with serious mental illness—incurable, expensive to care for, and seemingly a dubious economic investment. The goals of medicine must give equal status to responding to this inevitable circumstance in life.

I return now to an earlier question: what is a fair way to allocate resources to the care of persons with serious mental illness? So far I have tried to establish one point only: as a group, patients with serious mental illness have as significant a claim on health care resources as any other patient group. Given the context of their lives and medical conditions, their needs count as much as those of persons who can be cured or who can experience substantial improvement. There should never be any question about whether their condition puts them automatically in a category in which their claim to good care is less significant than that of other patients.

But does that conclusion help us to decide how we should best and most fairly distribute scarce resources to different groups of mental health patients? In part yes, for it means that those with serious mental illness have equal standing with all other patients, contrary to the pervasive belief that curable conditions have a greater, more compelling claim on resources. But in part no, because the limits of available resources may make it impossible to meet all the legitimate needs of different groups of patients, some seriously ill and others not, some who can be decisively helped and others who cannot. What then?

My suggestion—which is not without problems of its own—is to allocate resources by considering the extent of suffering experienced by different groups of patients. Those who can be helped and those who cannot be helped have one thing in common: both groups are suffering, and their most immediate need is to have that suffering relieved. The medical goal is the same in each case. But one group—those with serious mental illness—will, on the whole, be suffering more intensely and with less hope. Moreover, their illness is more likely to be a significant problem for their families, who share in their suffering. Those whose suffering is less intense and who can function with less assistance from their families could reasonably be given fewer resources. I specify fewer resources, not a deprivation of resources altogether. But if both groups deserve to have their suffering relieved, it would seem most reasonable to begin with those comparatively worse off, helping them first, but in the process making certain that an appropriate amount of resources is retained for those who are suffering comparatively less.

A major problem with this approach is the difficulty of calculating the degree of suffering that different groups undergo and of taking into account the variation in individual suffering within each group. Nonetheless, as in many other matters of policy, rough calculations are better than none. It should not be impossibly difficult to make the needed distinctions.

Policy makers and others should not be intimidated by the fact that care for persons with serious mental illness does not meet the usual standards of efficient health care spending. This group meets the more important test of having their treatment justified by central and long-standing goals of medicine. The idea of efficiency in health care spending needs to be adapted to those goals, not the other way around.

Dr. Callahan is director of international programs for the Hastings Center, Route 90, Garrison, New York 10524-5555 (e-mail, callahan@thehastingscenter.org). This paper is one of a series of three papers in this issue on care of the least well off in mental health services.