2020 Vision: Health in the 21st Century (1996)

Jeff Goldsmith, Ph.D.

President, Health Futures, Inc.

As you might surmise from the title "Risk and Responsibility," this discussion reflects a deep personal interest in the issue of risk and how we as individuals in society manage it. That interest is more than an intellectual interest; it is a visceral one. In my professional work, I think about how changes in science and technology and in the organization and financing of care are going to change this trillion dollar health care activity that we are all a part of.

In my personal life, I train in Tae Kwon Do and ski off cliffs and out of helicopters. I believe that my professional activities are more dangerous than my recreational activities. This topic, the issue of how health care financing is going to change, is fraught with particular danger because so much of the beta, the variability in our financing system, is accounted for by political and normative forces that are notoriously difficult to predict.

As a person who has been doing this for a while, I have discovered that there is a great deal of difference between forecasting 25 years ahead and skiing off a cliff. In skiing, the more air you put under your skis, the more danger there is. In forecasting, the more air you put under your forecast, the less danger.

I discovered this about 10 years ago when I was asked by Hospitals magazine to write an article on the U.S. health care system in the year 2036. My first reaction to that request was, I suspect, not all that different from the

way anyone would react, "Who knows?" Then, I realized that not only would I be dead in the year 2036, but so would every single person who had read the article. It was a liberating moment, because I realized that I could say whatever I wanted and no one would know how wrong I was.

Now, with the year 2020, I think we are in a slightly more dangerous situation, because there is the prospect that a few of us will be alive in the year 2020. Christine Cassel, a gerontologist and a friend from the University of Chicago (who is now at Mount Sinai in New York), gave me these prevalence forecasts for Alzheimer's disease suggesting that the survivors among us probably will not remember what was said. So it seems worthwhile taking some predictive risks.

It seems to me that the central direction of change in our financing system in health care in the last 15 years, which I see persisting for the next generation, is the shifting of economic risk and responsibility for health care costs from the government and employers to health plans and the health care provider community, and subsequently to individuals. I see this trend accelerating in the next 15 years as high-risk populations enter managed care pools. Of course, the growing number of uninsured people in our country fits this trend because what we are seeing there is a shift in economic risk from the government and employers, who are trimming their responsibility for certain segments of the employed population, to individuals and to a lesser extent, to the health care institutions that have responsibility for them.

The idea of shifting economic risk is the central theme of this discussion—how we manage that shifting of risk and how an important societal institution, namely managed care, is going to change as it assumes a steadily increasing amount of risk—the risk involved in an aging population.

Even though managed care has been with us as an institution for the better part of 60 years, it is my contention that it is still in its adolescence as a societal form. Three powerful forces are operating now in the health care financing landscape that are going fundamentally to transform that business as it moves forward into the first part of the next century.

Fifteen years ago there were only about 10 million people enrolled in HMOs (health maintenance organizations) in the United States, and that population probably represented the healthiest 10 million people in the country. The typical HMO subscriber was probably white, a blue- or white-collar worker with a family comprised of people who did not get sick very often. The largest costs in a typical health plan were obstetrics and pediatrics.

Clearly, a major change that is taking place in the composition of the managed care population at risk is the flood of high-risk individuals entering that pool. These individuals come from the two highest-risk subgroups in our population: the multifunctionally impaired, chronically ill elderly population, for whom Medicare is currently responsible, and the poorest of the poor—people enrolled in Medicaid programs—who bring all of the societally driven risks associated with their socioeconomic status into the risk pool.

So if managed care 10 years ago was about managing the care of people who did not get sick very often, managed care in 10 or 15 years is going to be about trying to manage the risk of the sickest people in our society.

The second major change is the change in the competitive structure of health insurance itself. Ten or fifteen years ago, managed care plans had the luxury of pricing their product under the spacious cost umbrella provided by completely unmanaged indemnity and Blue Cross plans that basically wrote checks after the fact for decisions made by professionals and institutions.

These forms of health care payment were primarily giant sluiceways for other people's money. If the Blue Cross plan in your area or Metropolitan Life raised its rates by 20 percent a year, your managed care plan could raise them by 15 percent and look like a hero. That has also changed dramatically. Health insurance premiums, at least in the private sector in the United States, are falling, and falling most rapidly in those communities with the highest managed care penetration. How long this continues is anyone's guess, but disinflation in the private health financing system is a very new and welcome trend.

So we have gone from a situation in which managed care was able to acquire a lot of additional revenue while caring for relatively healthy people, to an environment in which the per capita amount of dollars going to the plans is shrinking in real terms.

The fact that indemnity health insurance based on after-the-fact payment is disappearing creates a tremendous challenge for managed care plans to invent a new rationale for their existence besides "we are simply cheaper than conventional health insurance."

The third change in health care financing, which I believe is the most significant, is that we are moving from an event-driven to a risk-driven health care payment system. One could make the argument that until recently, managed care was not really about managing care at all. It was what I call event-driven cost avoidance (see Figure 1), which means trying to minimize the cost of caring for someone after the person has become sick by

reducing the amount of hospital care used and interposing bureaucratic barriers to access to practitioners. In this model, the health care system is arrayed at the bottom of the cliff waiting for people to arrive, while the managed care plan tries to reduce the cost of cleaning up the mess after the patient has crashed onto the rocks.

This model is rapidly failing as a viable definition of the managed care business, particularly in the many communities in our country where managed care has been established for a long time, because the first major layer of savings that managed care tries to achieve—reducing per capita hospital use—is nearing bottom in a lot of places. This is certainly not the case in the East, where a variety of regulatory forces, along with resistance from the medical community and from the labor unions, have retarded its growth.

In my home town of Portland, Oregon, inpatient hospital use for the entire community, including the elderly, is about 370 days per thousand people and falling. We used to argue in health planning about whether four hospital beds per thousand was the right standard. Managed care plans in my home town are running at 0.8 bed per thousand, including the elderly, and falling, 70 percent of the Medicare population have enrolled in managed care plans voluntarily.

The other thing that is happening to managed care in communities where it has been long established, which should not surprise anyone who thinks about the nature of price competition, is that price is becoming a less and less useful guide to selecting a health plan. In addition to premiums falling, what is happening in most of these communities is that the difference in price between health plans is narrowing to the point of perhaps a 5–7 percent difference between the least expensive and the most expensive plans.

As if that were not enough, if price does not serve to differentiate health plans from one another in these highly competitive markets, the networks that provide services under different health plans are becoming increasingly identical. So on two important points, price and access are far less useful than formerly for consumers and employers to decide which plan they want to choose. Thus, in addition to exhausting the easy savings from reducing hospitalization, managed care plans face an additional challenge—how to differentiate themselves from one another.

What you heard from the Jackson Hole Group this summer is that managed care is in the process of inventing a new rationale for its existence. That rationale has been called value improvement (see Figure 2). I believe that over the next 7–10 years, the activities that we have listed here are going to be the predominant preoccupations not only of managed care executives, but also of their medical directorates and the physicians who participate in them.

There are really two challenges in value improvement. Challenge number one is to try to control the resource intensity of the use of clinical services across an episode of illness, while also trying to improve consumer family satisfaction with the clinical experience. As has already been pointed out, this issue of improving the value of services is a data- and information-intensive enterprise. Trying actually to understand how costs are gener-

ated—how they are built up at the bedside, in the clinic, and in the physician's office—requires orders-of-magnitude better information about what clinicians actually do in treating patients, as well as systems that enable you to examine the variation in patterns of care across a particular type of clinical problem.

The most important issue of all, the issue that Dr. Jack Wennberg at Dartmouth and Dr. Robert Brooks of the RAND Corporation have been working on, is how to reduce the huge variation in resource consumption by practitioners for treating the same kinds of problems. There is obviously a tremendous amount of gold to be mined here as managed care plans seek, with the assistance of professional societies and their colleagues in academic medicine, to rationalize clinical decision-making—to create what has been called, in Don Detmer's presentation, an evidence-based framework that defines what constitutes best clinical practice for a patient presenting with a particular constellation of health risks.

This is going to be a very difficult challenge, whose goal is to be able to say to the consumer or to the employer—with data to support it—that our plan is doing a better job of utilizing your clinical dollars than the other health plans available to you.

Now, it is worth asking ourselves what happens to this business when you have removed the excess hospital utilization, as well as a lot of the variation in clinical practice that does not necessarily add value in terms of patient outcomes. I have a somewhat unconventional answer to that question—an answer driven by the logic of managing risk itself. Managed care plans are going to discover that they are really in the public health business. So, to respond effectively to pressure from the large numbers of high-risk

people who will be migrating into managed care plans, improving the health status of defined populations ultimately is going to be what differentiates health plans from one another. These differences will provide essentially a noneconomic rationale for deciding which health plans to use.

The idea that market forces may be pushing private health plans in the direction of a population-based approach to health improvement is counter-intuitive. Familiar tools of public health, such as epidemiological modeling and forecasting, as well as the somewhat more invasive relationship between public health systems and the people they serve, will provide a guide for private health plans and participating physicians as they move into the next century.

What this represents is movement from an event-driven to a risk-driven framework for health care payment. Instead of diagnosis and treatment as its principal business, our health care system will have to predict health risk and try to manage that risk before it flowers into illness and cost. We will get a lot of assistance from the extraordinary developments taking place in both genetics and immunology. These advances will enable us to identify much more precisely the embedded genetic risks in populations and in individuals, while blood tests will identify the degree to which those risks have begun to flower into illness (see Figure 3).

I think the technological capacity in the year 2020 will be sufficiently powerful to enable us to frame the risks that an individual runs at any point in life—at almost any point beyond conception—and to present a challenge to that individual and that family, as well as society, to create the best outcome framed against that risk. What the health system will try to do is provide information about health risks to individuals and create a user-friendly framework that enables individuals to collaborate with the health system to manage risk before it flowers into illness.

Many questions are raised by this progression—a price-driven to a value-driven to a health-status-driven health care payment system. You can

see (Figure 4) that the focus of the health care system shifts from the hospital to the physician network to the risks embedded in the community itself.

The most important shift, however, is on the third line—the locus of control. Many cost management activities taking place in our health care system today are superimposed upon the doctor/patient relationship. Physicians find it necessary to have phone conversations with a far-distant nurse about what they can or cannot do to patients they have known all of their lives for any clinical decision that involves more than a few hundred dollars.

As health plans shift risk to physician organizations, responsibility for effective clinical decision making is going to devolve from a bank of nurses talking over 800 numbers to the communities of physicians who are at-risk. Much of the superstructure that has been built up to prevent physicians from abusing the fee-for-service system will become unnecessary. The real challenge will be for physicians in organizational frameworks that range broadly from independent practice associations to group practices. As risk devolves onto these physician enterprises, so will responsibility for setting clinical standards and policing them. Thus, we see the responsibility for intelligent, thoughtful, conservative clinical decision making devolving back onto the physician communities that are at-risk.

As we move into Stage III, we see a further devolution of responsibility onto individuals, as well as a tremendous societal debate over the appropriate division of responsibility between families, society at large, and the health system for the task of health improvement. The fact that 40 million people in the United States have no health insurance means that those 40 million people are 100 percent at-risk for the costs and consequences of their medical problems. This is obviously inappropriate at many different levels. Yet the idea that we also have individuals—in many cases, wealthy individuals—with no economic risk or responsibility for the cost of care is, in a way, equally intolerable to our society.

I think that many issues are raised by how we arrange things in our society, but that as citizens, a balance is needed between our right to health care and our responsibility to use the information available about our health risks to minimize them through our behavior.

There is no question that an increasing amount of our health is going to be mediated by the decisions we make as individuals, as well as by decisions made by society about how to organize our communities. There is something fundamentally unbalanced about the idea that we have a right to health care without a reasonably rigorously defined set of obligations to live in a healthy way. It seems to me that as we move into this third stage, we are going to have a dialogue about what we as citizens should be doing to manage our own health risk and the appropriate division of labor between the health system and individuals or families in managing that risk.

If there was a "hole" in the Jackson Hole paradigm, it was that the sole role of the consumer in their model was to select the less expensive health plan. That role does not seem a sufficiently active one to enable us, particularly during an epidemic of chronic disease, to use effectively the societal tools—the science and technology—that will be available to help us manage the risk. So when I refer to a contract with the family, I am not sure where that contract will evolve or be developed. It may well be that the appropriate locus for dialogue about the balance of risk and responsibility between society and the individual is first of all, the health plan.

What is really going on in Congress right now, is that the federal government and state governments are getting out of the business of paying hospitals and doctors and into the business of delegating risk and responsibility for future costs to managed care plans. They are also delegating the very messy tasks associated with the transition in our health care system that will occur as managed care increases in influence; the reduction in the disparity of income between primary care physicians and specialists, and liquidation of the huge surplus of bed capacity and technology that we have in our health care system. These issues are also being delegated to health plans and, unfortunately, so is the issue of defining the appropriate relationship between the individual and the state, and between the individual and the community.

One could perhaps argue that there is a fourth stage that does not appear in Figure 4. Stage IV is probably what has gone on in Oregon. There are very real moral hazard-related limits on how much we can expect health plans to do in the way of limiting access to clinical services, particularly for the terminally ill and the handicapped. There is a very real limit to how much people will be willing to tolerate managed care plan involvement in the hospice movement or in the right to die a dignified death. If these issues are resolved by the narrow economic interests of the health plan, there may be tremendous tension between individual wishes and the economic interests of the plan.

The issue of an appropriate allocation of health dollars across a community and a region is going to be begged by the growth in managed care and will eventually result in a community dialogue that, with hope, will address them in a systematic and broad-based way as was the case in Oregon. Are we getting the maximum value for our societal investment in health care services?

The purpose of this discussion is not to minimize the costs or the difficulty of this transition, but to point out that there are opportunities involved in the idea of managing health risk as a central activity in our health care system. It is an integrative idea and something that competing institutions will not be able to do by themselves. I think we are going to see collaboration among competing health plans, and between private health plans and the public health sector. I think we are going to see bridges built between the health establishment and other institutions in our communities that are implicated in the economic risk associated with illness. These include our educational system, our criminal justice system, and our family and social services networks. To accomplish this transition in an intelligent and thoughtful way is going to require building such bridges. It is going to require creating a virtual community to support our country's health.

John M. Eisenberg, M.D., M.B.A.

Chairman, Department of Medicine, Physician-in-Chief, and Anton and Margaret Fuisz Professor of Medicine, Georgetown University Medical Center

I feel as though Dr. Goldsmith has brought us to a Tom Cruise film festival. As I think about his comments about the role of medicine in modern society, Top Gun comes to mind—high technology at work. We are doing the best that we possibly can, and we are delivering. Yet he also seems to describe a world of health care that is like The Firm: that is, we are working for an organization that is incredibly unscrupulous but we have not figured it out yet. The future of the health care enterprise seems to be so uncertain, according to Dr. Goldsmith, that it sounds like another entry in the film festival— Risky Business.

As I think about the future of health care, this idea of entering a world of risky business is a frightening one. I remember the adage that 10 percent of the people incur 70 percent of medical care costs. If that is true and if we the providers and the patients are going to be responsible for decreasing 70 percent of the costs that are incurred by 10 percent of the people, the question is, ''What do we know about where those costs are coming from? Can we discern the risks so we can alter them?''

I came into medicine at a time when the most popular books about health care were written by Ivan Illich, Thomas McKuen, and Rick Carlson, who said that health care providers did not make any difference. We might as well close up shop because we were having no impact on health. It is very reassuring to note that pundits now believe that we make such a tremendous difference in the outcomes of care for patients, and the costs, that we have so much control, that we should assume all of the risk.

Yet perhaps that is not what Dr. Goldsmith really means. Perhaps he is saying that there is much variation in what we do and that, since we aren't sure which parts of what we do make any difference, society is going to give the risk back to us, to deal with in any way we want.

No, I do not think that is what he really means either. Yet, it is of concern at a time when we do not have good measures of risk stratification, when we know that measures of severity of disease have lagged behind what the market demands, that we are being told—those of us in academic medicine, the Agency for Health Care Policy and Research, NIH, and the entire health enterprise—that we have really failed to provide the public with the information it needs to decide what kind of health care it wants.

This gets back, of course, to the question that has been a recurring theme throughout our discussion: Where is the information that we can apply in accepting this responsibility for the risk and variation in health care costs and outcomes?

There is a wonderful line from T. S. Eliot in which he writes, "Where is the wisdom that we lost in knowledge, and where is the knowledge that we lost in information?" He wrote that early this century, before there was any idea of a silicon chip. The meaning was that information alone will not do it—that we must take the information in all of these databases, create knowledge from it, and from that create wisdom. I hope we can serve the public as well as possible in that regard.

Let me turn to a second theme in response to Dr. Goldsmith's comments. I am fascinated by the idea that risk would be shifted to providers and to the public. If that is the case, the public will probably ask, who are these third-party payers and what have they done for me lately? Why am I paying overhead to these people who then pass my money on to doctors and hospitals and ask them to take the risk? In some ways, we are led to the same conclusion as one major entrepreneur in health care, who said recently about his hospital chain, "We will be the K-Mart of the health care system." In essence, we will be the wholesalers. We do not need these retailers. We do not need these intermediaries.

As I think about Borders moving in and shoving out the little mom-and-pop bookstores, and about Giant and Safeway moving in and displacing small grocery stores, it is not just a matter of accepting risk but a matter of accepting size, integration, and networking and what they mean to the health care system. I think, in general, that the results will be salutary. Yet, as a physician who relishes the intimacy of the primary care relationship, I wonder what that size and the assumption of risk will mean to the doctor–patient relationship. Is there a way that we can accept that risk and yet retain the intimacy of our relationship with our patients?

I think about calling up my travel agent for an airplane ticket. I do not relish the idea of my travel agent working for an airline any more than I like the idea that my doctor has large incentives to do either more or less than is appropriate for me. I would like for us to develop ways in which we could

assume risk, adopt integration and networking to take advantage of size, and yet retain that intimacy and trust that is so important in health care.

Let me finally ask a question about the assumption of risk, whether we as a society—not just one health care system or another, or doctors, or small groups of physicians—will assume risk? The more we learn about genetic predispositions and genetic markers for disease, the more we learn that risk is not merely whether or not you drink and whether or not you smoke. It is not just whether or not your physician has advised you about healthy habits. It may be uncontrollable, essentially the hand we were dealt.

If we reject Thomas McKuen and Ivan Illich and assume that we can make a difference, we still are left with the fact that the people who come to us are in very disparate risk categories. There is the jeopardy that we will reject as our patients people who are BRCA positive or people who have positive family histories of disease without even knowing what genes are involved, figuring that some gene is lurking that increases the risk of our patients' incurring costs. This is going to test our community and our societal values; it will also test the ability of members of the health care system to share responsibility for one another.

PARTICIPANT: We had, I would say, a rather optimistic picture of the future, at least from our Dr. Goldsmith. I wonder if there is not more to the dark side when I think about the fact that the major for-profit concerns have their stock market value determined by the ratio of their promotion and administration expenses to their expenses on actual health care—the higher the latter, the lower the former. Are we headed in quite the opposite direction, where we will not get to Stage III or Stage IV but will end up in a swamp unless we make it impossible by law, I suppose, for nonphysicians to practice medicine? That is a very eschatological and unhappy view, but it seems to be a view that we perhaps ought to hear a little more about.

DR. GOLDSMITH: Well, to be rigorous about that, their stock price rises and falls in accord with per-share earnings momentum, not the loss ratio. Let's be clear about what the market is valuing.

PARTICIPANT: No, but I understand that market analysts use that figure as a shortcut.

DR. GOLDSMITH: Well, these folks are my colleagues. A lot of them are poised right now over the sell button on their computers because they already know what many of the people who are worried about for-profit medicine dominating health care fail to remember, which is that health insurance is a brutally cyclical business. On the downside of that cycle, you do not want your capital tied up in an enterprise that is losing money on underwriting medical risk.

I do not know how many of you saw the market response to U.S. Healthcare's missing its earning by a penny or whatever it was. The company lost 40 percent of its market value. A lot of the money in for-profit stocks, which I follow closely, is "hot money." As Mexico and Brazil have taught us, you cannot really build a solid foundation on hot money. When the hot money leaves, the executives of those firms are going to have to ask themselves if they can create a noneconomic rationale for their subscribers and for the physicians that work with them to continue participating in their system. So perhaps a bloodless University of Chicago person sees slightly more market discipline than someone who is looking narrowly at this allegedly inexorable trend of the gobbling up of not-for-profit medicine.

DR. HELMS: Bob Helms with AEI. I just wanted to get you to talk a little more about how this market will evolve over time. I do not disagree with your view that managed care may play itself out, but I wanted to ask what the market response will be to traditional fee-for-service medicine—if it exists at all now.

Dr. Eisenberg has talked about the concept of physicians becoming agents. Is it likely that the market will demand more of that as people want more information? It seems that in a certain sense, physicians have the knowledge base to become those agents. I wonder if there is going to be a market reaction to managed care? Is the market waiting to take advantage of it?

DR. GOLDSMITH: That is an interesting question, because I see two distinct responses to the shifting of risk on the part of physicians. Response number one—from perhaps 80 percent of the medical communities in most places—is a flight from risk into salaried employment by hospitals and health plans. Physicians are essentially saying that the business risks associated with their continuing to be individual entrepreneurs are unmanageable economically or psychologically, and they want shelter in a large enterprise.

What I have been saying to my health system CEO colleagues is that the risk is shifting to them. I know enough about large health care organizations to know what happens when you get up to $1 billion or $2 billion. The person who is really at risk is the CEO, and that individual cannot push the risk down far enough into the organization to get the kind of behavior change on the part of professionals on the frontlines, who are actually getting blood on their shoes, to really affect the cost or value of services.

The remaining 20 percent, which is just now beginning to be visible, is the flight toward risk on the part of physicians who are organizing a variety of risk-bearing, risk-seeking enterprises. They are saying, "Give us a percentage of the premium, give us the responsibility, and we will figure out how to organize medical care for the population that we represent." This is the most exciting thing happening in medicine right now—even more exciting than a lot of the outcomes research and evidence-based medicine—that is the organizational response to the shifting of risk.

I really believe that physicians have the power to manage this transition. If they throw the power away—if they give it to Columbia HCA or the

Massachusetts General Hospital—obviously these organizations will take it. We are seeing in California and elsewhere that physicians are able to organize risk-bearing enterprises successfully and to accept the risk and responsibility, not only for the economics, but for the health outcomes as well.

DR. HELMS: Are these responses in conflict—the idea that a small group practice of three or four people does not want to accept risk, but that the same people are willing to accept risk if they are part of a larger corporation?

DR. GOLDSMITH: Well, it is not necessarily a corporation. Some people have given the independent practice association (IPA) up for lost; that is not an accurate read. Many private practice physicians have found what might be termed "virtual relationships" to one another that enable them to bear and manage risk without becoming part of a large medical bureaucracy. A lot of people looking at the medical care marketplace right now are saying that the Mayo Clinics, the large regional group practices, are going to dominate. I think they are wrong. There are diseconomies of scale and coordination in medical practice that are going to severely hamper the 800-person or 1,200-person physician organization because it just cannot get the discipline or achieve the degree of change in values and behavior necessary to manage risk in an effective way.

DR. SHINE: Where does the capital come from?

DR. GOLDSMITH: There is no shortage of capital. The Mayo Clinic is sitting on, in round numbers, $2 billion. So is Kaiser. There is money lying around all over the place. I do not see capital as the constraining element here. This is not a capital-intensive business; it is a knowledge business.

In terms of the issues that Dr. Detmer has raised—of how much money it takes to create an information system in which people can talk to one another and access databases—vendors will tell you that hundreds of millions of dollars are needed. If we pay close attention to what Dr. Detmer has said, it may not be capital that constrains us as much as the ability to create the connections and to achieve what is a more difficult challenge—professional consensus about appropriate standards of treatment. That is where the real gap is. I do not think capital constrains physicians to the degree that people marketing capital or systems would have them believe.

PARTICIPANT: One of the key words in Figure 4 is "contract." What kind of innovative work is going on in the area of contracts and contract development as you move toward Stage III? It is not just the contract between the purchaser and the managed care organization, it also involves the managed care organization with the provider and possibly with the patient. As I look at the kind of services we are going to need, I am worried about that contract link with the managed care organization, which in immature markets may be selling lives every three to four years as they move in and out.

DR. GOLDSMITH: Well, this is an area that I am raising as a policy issue now but do not see people formally addressing yet. Political scientists talk about a social contract—that is really the kind of contract that I mean.

A University of Chicago person would think about the insurance contract as a mix of incentives, a set of signals that society is sending to people about what it wants them to do. It seems to me that we are going to begin redesigning that contract to reward individuals who make health-conserving decisions and to place some economic risks in front of individuals who have a measure of control over the amount of health risk they are creating for themselves.

Dr. Eisenberg is correct. As we get more and more information about genetic risk, more and more questions are going to be begged by the role of the individual agency in promoting that risk and flowing into illness. The insurance contract by itself is a blunt instrument in moving that risk, but we at least ought to be trying to move the margin. We should be telling people through the insurance contract that we want them to make conservative, thoughtful decisions in managing their own health risks. I am not sure how long it will take before people think about this. The $50 deductible or $5 copay for prescriptions is not what I am talking about. We are going to have to decide what behavior we want to invest with some degree of economic risk. We must figure out how to protect individuals with limited resources from having too much risk placed on them. We have to begin thinking about what we want people to do to help improve their own health.

PARTICIPANT: You say very little about the role of government in this evolution, other than that it is shedding risk and will continue to do so. Surely there has to be a major role for government to ensure that the market will work as you envision it. Could you speak to that?

DR. GOLDSMITH: There is a huge role. I do not know how many of you saw Matthew Miller's piece in the December 11, 1995, issue of New Republic on the Medicare program. It is an absolute bafflement to me why we should use cost-based payment to pay HMOs to take care of Medicare recipients, which is essentially the framework we use now. We have built up this huge body of cost that varies threefold from community to community, and we now give health plans a license to "mine" that variation instead of, as Miller suggested, putting the contracts out to bid and letting health plans competitively bid for and accept the risk for the Medicare population.

On the issue of the loss ratio, one of the reasons HMOs are really nervous right now is because that loss ratio is an enormously tempting regulatory target. I could see a slightly different administration, a somewhat less risk-averse, more courageous one than the one we have in office now, saying that it is not going to contract with a health plan that spends less than 90 percent of the premium dollar on health services. I think that government has an enormous role to play here; I am just daunted by how difficult it is for our federal government to make rational decisions in health financing.

DR. EISENBERG: Let me add something to that. The idea of contracting for managed care organizations and having bids was described by Miller in that New Republic article as having come out of the Physician Payment Review Commission (PPRC). So I must state that it was developed beautifully by the PPRC staff and is described in last year's annual report. We will see if anyone decides to take it up. It is a risky but interesting idea.

DR. JOHNS: I am Mike Johns from Johns Hopkins. I am interested in understanding how the marketplace is going to accommodate the ever-expanding numbers of uninsured. Although there seems to be a sense that shifting government health dollars to the states will allow them to spread those dollars further, I expect that more people will drop off the roles and there will be an increasing number of uninsured. Somebody ultimately bears the cost of that, generally the middle class in some way.

How will this system deal with that? Will there be any sense of accountability? For example, these physician groups that you talk about—the new entrepreneurial physicians that we see coming together—will they be willing to take on the responsibility for some of these populations? Will anybody be able to afford to take on those responsibilities? Who is going to step up to the plate?

DR. GOLDSMITH: The system that we have now gives academic health centers and urban public hospitals the quasi-governmental responsibility of taxing the rest of the health care system to pay for services provided to the uninsured. I agree that there is nothing in the current round of "reforms" that is going to do anything other than increase the number of people who are not covered. I think this is fundamentally irresponsible, flawed social policy.

You could argue from a strategic point of view that the federal government is now the principal driver in health cost inflation in the United States. President Clinton's spirited defense of a double-digit rate of increase in public spending for two public health financing programs is not adding anything to the debate over how to get more affordable care for the population that is uninsured. I do not see the private sector leaping forward to take responsibility for these folks, and if public costs continue rising at the present double-digit rate, we won't get an affordable federal or state response to the problem of the uninsured.

Yet I do think there are opportunities to pool the purchasing power of individual and small group health insurers and give them the same kind of per capita cost advantage that large employers enjoy. We are not going to get very far in solving this problem without revisiting the health policy debate. With hope, we will not waste as much time as we have in the last two years.