A Roundtable Discussion About the Future of Psychiatric Services

Editor's Note: In August seven psychiatrists met at the American Psychiatric Association to discuss how psychiatric services and their delivery might change over the next 50 years. Three participants sat on one side of the table and questioned the four others. The questioners were John A. Talbott, M.D., professor of psychiatry at the University of Maryland School of Medicine in Baltimore and editor of this journal; Jeffrey L. Geller, M.D., M.P.H., professor of psychiatry and director of public-sector psychiatry at the University of Massachusetts School of Medicine in Worcester; and Sally L. Satel, M.D., staff psychiatrist at Oasis Clinic in Washington, D.C., and lecturer at Yale University School of Medicine. The respondents were Howard H. Goldman, M.D., Ph.D., professor of psychiatry and codirector of the Center for Mental Health Services Research at the University of Maryland School of Medicine in Baltimore; Steven S. Sharfstein, M.D., president of the Sheppard Pratt Health System and clinical professor of psychiatry at the University of Maryland School of Medicine in Baltimore; Gary Tollefson, M.D., Ph.D., president of neuroscience products at Eli Lilly and Company in Indianapolis; and E. Fuller Torrey, M.D., executive director of the Stanley Foundation research programs and professor of psychiatry at the Uniformed Services University of Health Sciences in Bethesda, Maryland.

Tollefson: I would start with something that might not seem like a breakthrough, but I would see it as such: improving our use of the therapeutic tools we already have. I think there will be a greater appreciation for trying to achieve symptom reduction. Today we still see far too many patients who are burdened by residual symptoms of their psychiatric illness. We have paid less than optimal attention to achieving response and remission in all types of psychiatric disorders. One way to achieve higher response and remission rates is to raise clinicians' and patients' expectations about the outcomes they should obtain.

Torrey: In terms of breakthroughs, I'd say it is likely that in the next decade we'll be able to target the psychopharmacological agents much more closely and perhaps even tailor-make them to a person's central nervous system chemistry and genetic predisposition. Such an approach is likely to produce agents with fewer side effects.

Sharfstein: Two other areas where I think we can anticipate some new models are in assertive treatment and supervision in the community for patients who are better but not well and who need to be involved much more closely not only with the treatment system but with a variety of social services.

Tollefson: From the discovery perspective, the genome project is intriguing, but I see it as a project that will not deliver us useful products until quite a bit downstream. One of my areas of interest that I think might deliver sooner is the whole field of pharmacogenomics. We have the ability to begin to identify subpopulations that are more likely to respond to drug X or drug Y than others. With pharmacogenomics we will be able to predict who's more likely to have a superior outcome and who might be at risk of a serious adverse event, which is already occurring in the field of oncology.

Torrey: I don't think it's possible to predict the speed with which things will develop in relation to the human genome project. I think the project has generally has been oversold. However, in the area of pharmocogenetics, or pharmacogenomics if you like, it is likely to produce some results. Just before I left my office this morning, I had a call from a young woman who has breast cancer and who was given a choice of three drugs to take on the basis of the results of a gene study. So I don't think that it is unrealistic to expect we would have that kind of ability in psychiatry.

Sharfstein: In psychiatry perhaps more than in any other part of medicine, I think we have more of a comfort level of trying to force people into treatment. As a field we have dealt with a lot of questions related to civil liberties, and we have worked out some very interesting procedures for trying to protect an individual's rights against the need for the family and the community to be able to cope with that person's behavior. Also, now that we have effective treatments, noncompliance that leads to severe consequences for the family and community has a different dimension. If we can demonstrate that treatment leads to improved outcomes in behavior and symptoms, then we have a much stronger case to compel people to be in treatment and to compel providers to follow through.

Goldman: The trend away from hospital-based care is very likely to continue. It's a major way of reducing expenditures and societal allocation of resources to mental health services. I think the trend is in the direction of more primary-care-based delivery of the medically oriented mental health services. The technological advances that make treatments more specific also make it possible for them to be provided by primary care physicians.

Torrey: I think the settings in which future care will be delivered will be driven largely by how the services are reimbursed. The two reimbursement systems we've had in my practicing lifetime, first in states and counties and now managed care, have both demonstrated virtually complete bankruptcy at this point, in my opinion. I think that the reimbursement system is likely to change in a dramatic way over the next decade toward outcomes reimbursement. When you start to reimburse by outcomes, you give contracts to the teams and agencies that demonstrate the best outcomes.

Sharfstein: In terms of financing, I agree with Fuller. There have been two major sources of funds. One—the states and counties—has funded long-term care and has led to various custodial strategies. It has underfunded treatment in the community. The other—insurance managed care—has been overwhelmingly an acute model that provides care for the most short-term, most disruptive kinds of symptoms and does not deal with the long-term issues. I would hope that in the future we would be able to devise strategies that link together acute and long-term financial needs. I don't think that hospitals will necessarily disappear, but I think we're seeing a redefinition of a concept of hospitals and a way of trying to develop continuity of care that especially includes psychosocial rehabilitation, housing, and the clubhouse model along with various day programs and some inpatient care.

Geller: The assertive community treatment model is still largely being driven by current reimbursement schemes. Some publications are now using data from a "scaled-back" assertive community treatment model to argue that such a model is more cost-effective and has comparable outcomes. In such a model, the team does not go into the community. Instead the patients come to the office. The case management is subcontracted out.

Another change we might see is toward residential programs with congregate living. An argument could be made for reimbursing such a program on the assumption that it is substantially less expensive to take care of 60 people in the community if they all live in one building than if they live scattered all over the city in ten six-bed residential programs. If it is more cost-effective, why won't we simply see a proliferation of this type of community institution?

Torrey: I'm not sure what studies you're referring to about scaled-back assertive community treatment. However, if you actually measure outcomes and look at quality of life, I think that maintaining continuity of the caregiver is much more cost-effective. When you subcontract out the case management and dissolve the team concept, you lose the continuity of the caregiver, which I believe is the absolute essence of the assertive community treatment model. So I doubt that such a scaled-down program would be cost-effective.

Sharfstein: I think the problem with the model of a 70- or 80-bed residence is that it goes back to small state hospitals. Even though there may be economies of scale, I think that most payers, including the public payers, are reluctant to undertake that long-term kind of financial responsibility, especially when many of these patients can be managed with a model that focuses on the continuity of the caregiver.

Geller: Let me just clarify: What I'm concerned about is not the kind of institution you're talking about. I'm concerned about the funding system viewing this aggregate living situation as no different from an arrangement in which people are living more individually. When you have 100 people living under one roof, it will be much more cost-effective to deliver services than it would be if people are living all over the city. In fact, some people can make the argument that they will live in a better neighborhood because the residence can be built in a better neighborhood.

Goldman: I see the trend going in the direction of decoupling, or uncoupling, what are normally transfer payments—that is money for people's housing, food, and so on—that were built into the costs of institutional care. I see continued decoupling of these payments from what we conventionally think of as payment for clinical services.

Tollefson: And it also will become more prescribed, which I think will be an interesting controversy. The approach to the administration of health care will be much more pragmatic and businesslike. It may threaten some of the autonomy of decision making, for both the patient and the provider.

Torrey: Let me try to answer John's question more directly of where the change is going to come from. First, let me say it is not going to come from Washington. Virtually no important change in human services or medical services has originated in this city in this century. Change almost always comes from the states, and in some cases from the counties. I think we are already seeing the earliest stages of change as more states apply for waivers from Medicaid and start to experiment. I think that we are going to get at least one bright, forward-looking director of a state department of mental health who has a governor who will back him or her in conducting experiments on the county or state level. Experiments will show that reimbursing by outcomes is very cost-effective and much better for patients and families. The successful policies and programs will spread to other states.

Goldman: Is payment by outcome a little bit like in the Middle Ages, where if the doctor fails the patient, the doctor is killed or not paid? How do you adjust it if a program decides to take on people who are bad risks? Or if a program avoids people who are high risk—cherrypicking?

Torrey: First of all, you don't get a choice. You're assigned 100 patients or 500 patients, and then you are reimbursed on the basis of the outcomes of those patients. You measure quality of life, both as patients perceive it and as others perceive it. You measure things like how many are working at part-time jobs, how much time they spend in jail, how happy their families are with services. You pay everyone a minimum amount, and then you pay certain teams or companies more money if they do a better job.

Sharfstein: I don't entirely agree with Fuller that nothing can happen out of Washington. I think some modifications can occur in the available health monies that could provide some help for continuity of care and for more accountability. I am not as optimistic as Fuller that people are really going to be interested in being held accountable, but I do think there can be changes in Medicare and Medicaid policy. The waiver is one example that encourages experimentation, but I think that at the federal level there are some demonstration projects and other kinds of efforts that can have more of an enduring impact over time.

Geller: I think that one future development will be much better defined outcome measures and better research into outcomes—because I think what Fuller is suggesting is that we don't have the ability to measure outcomes.

Sharfstein: I would like to see patients and families have more control over the resources that are available to them. I don't know exactly how we would do that. I would like to see some experiments with vouchers. I would like to see people be able to set aside some resources and to get knowledge through the Internet and get more consumer education and be able to pick and choose among options.

Tollefson: Some preliminary uses have been developed. For instance, patients can voluntarily sign up with their own personal account and get reminders to renew their medication. Providers can also follow up on patients' symptoms. These things are quite intriguing, but where they will go is double-edged.

Torrey: The Stanley Foundation is funding a project to develop a home monitoring kit for lithium that is similar to home monitoring for diabetics. You can send the data over the telephone system to your internist.

Geller: I think the computer will be used to a much greater degree as a rehabilitation tool itself. We are getting some surprising results using computer programs with a state hospital population of chronic patients with profound neurological damage. Such programs are also useful for patients who do not do well with interpersonal conversations and for those who need to go slowly on measured tasks.

Tollefson: We have been working with a computer program that allows us to monitor participation in treatment for mood disorders. The patient completes a standard subjective assessment tool on a regular basis, which provides the patient with direct feedback on clinical signs and symptoms. The assessment data can be transferred to the physician's office, and thus we can also be alerted when a patient in remission suddenly reports a significant increase in symptoms. It is another way of looking at access and early identification. It is quite intriguing and may also rewrite the settings of care.

Geller: I think it is entirely possible that we will see the disappearance of the concept that if there is no treatment we can offer a patient, the best we can do is care. Instead, we will have some treatment we can offer to everybody in the direction of symptomatic relief.

Goldman: I hope we will live to see the day when saying that people will be better but not well is an anachronism. People will get well.

Goldman: I think that as a specialty psychiatry will diminish. Many of the tasks that are not medically oriented will be picked up—already are being picked up—by nonmedical professionals.

Tollefson: I think there will be fewer psychiatric units of service available for the demand. It is going to put a premium, I suspect, on the field, with the physician-trained field of psychiatry increasingly assuming more of a concentrated role. I can't see any deviation away from increased use of physician extenders, people who are part of a team and who allow the psychiatrist to accomplish much more than he or she could in direct care. So I see psychiatrists taking more of a management and consultative role.

Sharfstein: The issue comes down to economics again. The trend has been to go to ever less trained or less experienced people to take care of sometimes very difficult patients. I think it is going to be difficult to counter that trend. There is a real push to try to get care in the hands of lower-cost persons.

Geller: Steve, have you noted at Sheppard Pratt a difference in the rate at which this change is happening across disciplines?

Sharfstein: From what I've seen, the rate in the other disciplines is actually faster than it is for psychiatrists. The pressure is on psychiatrists to be responsible for more and more patients, but the actual pay for psychiatrists who are doing that is going up. Also, I think there is always going to be in our society a different kind of marketplace for private pay. A large number of Americans with discretionary income are going to be willing to pay out of pocket, given concerns for confidentiality and other reasons. Some psychiatrists will have a 40-hour clinical practice and get paid out of pocket.

Torrey: I predict that in 20 years there will be no psychiatrists—no psychiatric profession as such. There will be physicians who will be trained in a neurological specialty in some form of what you may want to call neuropsychiatry. Neurology and psychiatry will be merged. The National Institute of Neurological Disorders and Stroke and the National Institute of Mental Health will be merged long before 20 years. A national brain research institute will encompass the whole field.

Sharfstein: The rest of medicine will have major breakthroughs over the next few decades. As a result of that—we've certainly seen it in the latter part of the 20th century—more attention will be paid to the brain and brain failure. Many general physicians are going to get much more psychiatric or neuropsychiatric training. I've made the prediction, in direct contrast to Fuller's prediction, that in 30 or 40 years we will have many more psychiatrists than today, although they are going to be different from the current group of psychiatrists. Whether they will be called psychiatrists is an interesting question. But I think we are going to have many more people who will be spending most of their time on the brain, brain failure, and the psyche.

Satel: I think the future of substance abuse treatment is through the criminal justice system. We are seeing that now with drug courts and diversion programs. We are trying to put more and more people into the treatment system through the criminal justice system in ways that I think are extremely promising, and the data suggest this, too. Because of the leverage and the coercive aspect, people stay longer and do just as well if not better than those who go to treatment voluntarily. Hard-core addicts who somehow manage to escape the criminal justice system, and not many do, will obviously have to be treated in a public-sector setting. I have always thought that the treatment system has the least clout with these patients. For example, in methadone clinics, there are patients who continue to use cocaine. We can threaten to throw them out. That is our leverage, and sometimes it works. Other social systems, such as public housing or the welfare system, could apply this kind of leverage with urine testing. That's where I think the future of treatment and behavioral modification is for hard-core addicts—using the other social systems to establish incentives and sanctions.

Torrey: I think one set of questions will be focused on the brain and behavior. Even when you understand the circuits that are activated on any given behavior, what does this have to do with motivation? What is personal responsibility when translated into neuroanatomical and neurophysiological terms? I think that is something we are going to continue to struggle with for a long time, and probably forever. Improvements in the technology to measure brain structure and function are going to bring such questions more to the fore.

Sharfstein: I think one of the questions that will always be asked is, What is the right balance between individual liberty and forced treatment? Related questions are, Can people choose to be homeless? To commit suicide? These are the kinds of questions we will always be dealing with, both from a scientific perspective and from the perspective of social policy.

Goldman: I would add questions about social welfare. What is the societal responsibility for people who have different capacities to participate in the person-defining aspects of life, such as work and interpersonal relationships? What will science tell us about that?

Tollefson: Two questions come to mind for me. Is it necessary to age? And, what is happiness? I think as technology continues to increase demands on individuals, and as individuals take on more and more in their lives, we will deal with the fundamental issue of what happiness is. Are we becoming a more happy or a less happy society?

In the past, I think there was a lot more homogeneity when it came to defining happiness. The vast majority of people would have agreed on a small set of things that would define happiness. But over the past 20 or 30 years, we really have diversified as a society in terms of the number of things that people do that they associate with being happy. Can any one individual define for any other what happiness is? If someone comes into your office and complains that he is unhappy, where do you start? I think there are going to be many, many potential definitions and a tremendous diversity across people's lives.

Torrey: Diversity is inevitable. By about 2020 Anglo-Saxons will be in a minority in the United States. We will have an increasingly multicultural society. So the kinds of questions you're asking will have multicultural answers. We will increasingly have to question the concept of America as a single country because economically and culturally we will be so interdigitated with the rest of the world in ways that we can't even imagine now. When you talk about resource allocation, we are used to thinking of 270 million people. Instead, we'll think in terms of billions of people, and that will change the way we even conceptualize the questions you are asking.