Medicare has successively curtailed the benefits provided in extended-care facili ties. Unexpectedly, Medicaid furnished a cushion when Medicare denied widely anticipated long-term care benefits to the aged.

EXAMINING THE AVAILABLE DATA

Suitable data on changes in population, use of services, unit cost, and program expenditures have been published for California.1 My own reanalysis of the data shows the following percentage distribution among the sources of increase in expenditures between 1964-65 and 1969:

When Congress and the state legislatures acted to curtail rising Medicaid expenditures, they took a variety of steps. In some states the range of benefits was narrowed; elsewhere limits were put on the amount of hospital or physician benefits per episode of illness or period; in other states fees paid to providers were reduced or co-payment by the medically indigent was introduced; finally, criteria of eligibility for Medicaid were restricted everywhere. The result was that in states or cities with a history of liberality toward the medically indigent, certain segments of the population were worse off after a few years of Medicaid than previously. Not only had they lost access to free or part-free care, but along with the rest of the population they were confronted with much higher price levels for medical services.

In my judgment it was a combined reaction to the loss of eligibility for free care by some and to the high prices facing all that led to a revival of interest in national health insurance in the late 1960's. It was not so much the attraction of the remedy, which took varied forms and was sometimes nebulous, but rather a growing feeling that government had an obligation to try to undo what government had helped bring about-high prices.

I noted previously the tendency in the late 1960's toward equality of use of health services by income class. Use data are not the same as information on access, which is more akin to the opportunity to use services and comprises several dimensions, such as distance, waiting time, comfort and convenience, and so on. Conceptualization and measeurment of access have been unduly neglected. Moreover, as use is equalized, the tendency is to proclaim equal health status as the goal, implying inequality in the use of health services to offset need. Equality in health status by income class does not strike me as a realistic goal for the health services industry, since medical care is not always the key to improved health.

PROGRAM FOR HEALTH CARE PLANNING

The two new programs in haelth planning-RMP and CHP-had no connection with urban planning in general or with the two new programs in financing, Medicare and Medicaid. Nor did they relate to the single ongoing health planning activity in this country that had emerged from the 1940's. Under the Hill-Burton Act passed in 1946, federal grants to the states for hospital construction were conditioned upon the preparation of an annual inventory of facilities and a state plan indicating priorities for the coming year. Federal support for hospital planning was expanded in the early 1960's when direct federal grants became available to areawide planning agencies within the states. Mostly these took the form of hospital councils under voluntary non-profit auspices operating without official mandate but with support from philanthropic capital fund raising agencies and from corporate donors. Sometimes sanctions were brought to bear by the refusal of large purchasers of care, such as the local Blue Cross plan, to pay for care provided in facilities not approved by the planning council.

The impetus for the new planning programs came from two disparate sources in the health field. The progenitor of RMP was the medical education and research community, which through the intercession of Mary Lasker persuaded President Johnson to establish the President's Commission on Heart Disease,

1 Foline E. Gartside. "Causes of Increase in Medicaid Costs in California," Health Services Reports, Vol. 88, No. 3 (March 1973), pp. 225-35.

Cancer, and Stroke, with Dr. Michael De Bakey as chairman. The point of departure was the belief that technological devices and procedures had come into existence that were not being widely applied outside the great university medical centers. As Dr. De Bakey put it, by having funded medical research, the federal government had built up an equity in having the fruits of medical research applied. The De Bakey Commission recommended the establishment of regional centers of excellence throughout the country, but neither the Administration nor Congress was prepared to finance or otherwise undertake such a venture. What Congress passed in the spring of 1966 was a program to encourage regional cooperative arrangements in the diagnosis and treatment of heart disease, cancer, stroke, and related diseases. It was left to each region, with boundaries defined locally, to plan for developing an RMP organization and then to propose specific operating programs for support by the federal agency. There was no national agenda, no set of nationwide priorities-other than to encourage cooperative arrangements at the regional level.

The impetus for CHP was an unusual conjuncture of preferences and opportunities. State health officials had wanted for a long time to do away with the numerous categorical public health grants-in-aid to the states. Congressional leaders in health were opposed to a consolidation of grants on the ground that federal monies should be spent in accordance with priorities set at the national level. However, the same leaders were somewhat interested in promoting health planning at the official level. A compromise was struck, with some of the cate gorical grants merged into a general public health grant to the states in return for the creation of a health planning mechanism in the states. The amount of money involved was of the order of $84 million in 1965, but on this modest base a new era of "Partnership in Health" was proclaimed between the federal government and the states and between official agencies and voluntary organizations. An elaborate structure was designed to plan for the optimum provision of comprehenvsie health services, consisting of "A" agencies to plan statewide, "B" agencies to plan locally or areawide, and "C" agencies (universities) to train staff for the A and B agencies. Provision was made for majority participation by consumers.

CHP was enacted in the fall of 1966. Since their sponsorships and constituencies differed and their mandates were far from concrete, the two programs were not linked in implementation except fortuitously. They were, of course, connected in the labor market, as the two new programs created a boom demand for top-level health planning staff, of whom there was a scanty supply.

The two programs were adopted independently of one another, and each developed in its own way. The federal administrators of RMP thought that the absence of nationwide goals was an advantage. Let each geographic area organize to mobilize its resources and then let each agency propose what it wishes to do, in light of its capabilities, priorities, and opportunities. Money for RMP was adequate relative to the proposals made. With the medical school in a region typically playing a central role in organizing the cooperative arrangements, the most common program was provision of post-graduate medical education for physicians.

DIVERGENCE OF INTERESTS

It has taken CHP much longer to get organized and functioning. In some parts of the country problems of substance still await consideration. The notion of comprehensiveness, which ultimately came to encompass environmental health as well as personal health, has led to a scattering of effort. Agencies that had gained experience in hospital planning in the 1950's and early 1960's were either submerged or discarded when the earmarking of Hill-Burton funds for areawide hospital planning ceased. Where planning agencies have passed from the organizational phase and begun to deal with substantive matters, they are often immersed in reviewing individual proposals for construction. Little time is left for dealing with the conceptual, technical, and public policy aspects of health planning.

A vital omission is the failure to consider the essential nature of a desirable community health plan and thus the very basis for health planning. Most often the view adopted by hospital planning agencies in this country is that a sound community plan is the sum of the best possible plans developed by individual institutions. That would of course hold true if the interests of the individual institutions and of the community always coincided. In that case, the planning agency could furnish information on socio-economic developments in the area and

on the future outlook that would help improve the planning performance by individual institutions.

There are good reasons to believe, however, that a divergence of interests frequently exists between an individual institution and the public at large. One example of such divergence is the desire on the part of the individual hospital to attract the best possible medical staff, while the community continues to receive care from physicians without a hospital staff appointment as well as from physicians with one or more appointments. Lack of a hospital appointment bars a physician from the most ready and likely source of continuing education in medicine.

A second source of divergence of interests is the inability of the physician to care for patients in a hospital in which he does not hold a staff appointment. This means that if a special facility for the care of a disease with infrequent occurrence were confined to Hospital A, physicians in Hospital B would lose out. It would be an act of self-denial for the latter to support such a proposal. The problem of regional coordination of health serviecs in a metropolitan area cannot be resolved in the absence of appropriate policies on the extent and types of multiple staff appointments for physicians.

A third example of divergence of interests, which has arisen since World War II in connection with the growth of prepayment, is conveyed by the proposition generally associated with Milton Roemer. Roemer's Law says that under conditions of prepayment, hospital beds, if built, will be used. In the early 1960's a good deal of controversy surrounded this proposition, but in the late 1960's it gained increasing support from scholars and almost unanimous support from active hospital planning consultants. The case of England, where there had been no hospital construction for 25 years and longer, was conclusive evidence that geographic areas did differ in the ratio of beds to population but were fairly uniform in rate of occupancy. It is worth spelling out the serious policy implications of Roemer's Law. A high rate of occupancy is taken as a sign of full use of resources. It is beneficial to the community, as well as to the hospital to which it assures more revenue, as long as a low rate of occupancy denies revenue and threatens the financial survival of a hospital. However, if prepayment assures a high rate of occupancy to a hospital, and the bed-to-population ratio in the area makes no difference, any deterrent to continuing expansion by individual hospitals has been removed. It then becomes incumbent upon the community to decide how many hospital beds it wishes to support. So far, within the range of bed-to-population ratios observed, there is no reason to base that decision on the effects that a change in the number of beds presumably exerts on the health status of a population.

If potential divergence of interests lays the substantive basis for health planning, it should constitute the core. The aim to reduce the supply of beds entails mandatory controls, exercised by seizing the opportunities that present themselves, such as proposals for rebuilding and relocating individual institutions. The other examples of divergence of interests entail the operation of more subtle. factors, such as the gamut of physician-hospital relationships, and may call for complex actions which are best left to voluntary arrangements until they are better understood.

THE INFORMATION PROBLEM

An important contribution of planning agencies in the past has been their service as repositories of information. The question persists: What types of data are useful and necessary for health planning? The first answer is that planning in accordance with so-called health need is fruitless, since people always use more or fewer services than indicated by purely biological and technological factors. To collect a good deal of information to measure the need for services is bound to be an idle exercise. Beyond that, however, it is helpful to recognize that the most voluminous information routinely collected is unlikely to illuminate every problem that arises. It is best to rely on a combination of a modest, routinely collected data base and an organization capable of collecting special purpose data with competence and speed.

Data pertain to the past, and all best efforts do not guarantee an accurate description of the future. Uncertainty is greater for small areas than for large ones; yet for the most part the market for health services is local. Let us therefore acknowledge the following: It is a statistical property of small numbers to display greater variation or instability than large numbers; many projections for small areas are bound to be in error; the best potection against erroneous forecasts lies not in technical improvements in forecasting but in flexible opera

tion; provision for flexibility costs something extra; and therefore it may well be the principal business of planning organizations to develop and disseminate devices that promote flexibility.

Three other changes seem necessary. First, in order to be effective, a planning agency must focus its efforts. To try to plan everything is to accomplish nothing. It is useful to ask not only what the most important problems are, but which of them are most susceptible to solution at present. Second, the preoccupation in health planning with structure and process has been excessive and may reflect in part the new requirement of consumer participation, which is sometimes translated into attempts to secure local control. The sources of decision making in a society are diverse and subtle, and it is difficult to articulate precisely what makes for legitimacy of authority. Continuous preoccupation with the decision making structure is unsettling and unlikely to yield action on substantive matters. A third requirement for progress in health planning is a reversal of the tendency to dam the flow of information. Without access to systematic information, we lack knowledge of the dimensions of most of our problems and of trends.

Certain organizations, governmental and other, are collecting more and better information than ever before. However, they are holding on to it for internal processing and analysis, while other agencies are barred from collecting similar data. In this instance the quest for efficiency in information gathering has created data monopolies. A monopoly over information exercised for an extended period does not serve the public interest, because some problems go unexplored. The work done by insiders escapes criticism, and outside scholars work in areas where information is more readily available.

NEW ORGANIZATIONS

The increased flow of public funds into health in the mid-1960's led to the purchase of additional services from the private sector at large. Production by government facilities did not increase and there was only a modest expansion of facilities aimed exclusively at government beneficiaries.

The latter facilities were the neighborhood health centers, most of which were launched by the Office of Economic Opportunity (OEO) and later supplemented with centers funded by the Department of Health, Education, and Welfare. Neighborhood health centers were meant to exert a direct influence on the way in which medical care was provided to persons in low-income areas. They were to provide comprehensive, or a broad range of, services, including outreach services by indigenous workers and legal services that do not usually fall within the purview of a health facility; their services were to be continuous, in contrast to the fragmentation and discontinuity that characterizes services in hospital out-patient departments. Accordingly, they could assume a measure of responsibility for the health care of the registered population.

The results achieved by neighborhood health centers have varied, and the conclusions drawn from the findings depend to some extent on the priority assigned to the several goals. The extent of comprehensiveness depends on the size of the budget. Continuity, in the sense of seeing the same physician in successive visits, is difficult to achieve, so continuity is often sought through the patient's chart. The idea of serving a designated population is similar to a bounded catchment area, which is easier to implement in rural areas than in a large city. But even apart from feasibility, it is necessary to spell out the circumstances under which it serves the interests of a group to limit their options. Published figures show that the neighborhood health center compares favorably in cost with other methods of delivering personal health services. I have reservations concerning these findings. Allocations of overhead can make a big difference in allowing for services that some programs provide and others do not. More important is the fact that a "registered" population is not the same as a population enrolled in a health insurance plan, in that members of the former lack the financial incentive of the latter to obtain care within a particular program and not outside. The concept of a “registered population associated with a neighborhood health center serves to overstate the denominator in any per capita computation and thus understates per capita cost. These technical reservations seem to receive some support from the reluctance exhibited by neighborhood health centers to change from lump-sum funding to financing through the sale of services to patients under the usual financing programs, such as Medicaid and Medicare. It is now recognized that this method of delivering health services is likely to be extended to large numbers of people for reasons of cost and

because of recurring doubts about the desirability of maintaining separate health services systems for the poor. What reason is there to believe that a separate system for the poor will provide health services equal to those received by the rest of the population?

To many people the concept of the Health Maintenance Organization continues to pose a mystery. Depending on the context, it seems to change in form. At times the HMO seems to be prepaid group practice dressed in a respectable legislative cloak. Sometimes the HMO is either prepaid group practice or the medical foundation, which is an association of solo practitioners. To those who formulated the concept, the HMO is prepaid group practice or the medical foundation or any other organization that meets the requirements of a basic definition: Subscribers will receive contractual assurance of access to a specified, fairly broad set of health care benefits in return for a stated premium. The form of HMO ownership, non-profit or otherwise, is left open, as is the method of paying providers of service.

The HMO concept is most closely associated with Dr. Paul Ellwood, who has developed, modified, and promoted it in the course of serving as consultant to both the Nixon Administration and a Senate subcommittee chaired by Senator Edward Kennedy. Disenchanted with the course of events in the late 1960's, particularly the continuing rise in expenditures and cost without a sign of improvement in the geographic distribution of health resources, Dr. Ellwood placed the blame on the ineffectiveness of government regulation and planning, and sought relief through competition in the market place. The aim in promoting the HMO and fostering its growth is to obtain diversity; HMO's would compete with one another, as well as with the more traditional health services system. The interent incentives of the HMO were toward efficiency; any tendencies toward underserving the consumer could be mitigated by feeding him information on the quality of care.

It is worth acknowledging the political ingenuity displayed by Ellwood in bringing under the HMO umbrella both prepaid group practice and the medical foundation, arch-rivals in California. While prepaid group practice has long enjoyed support from liberals, a medical foundation is the creature of the local medical society. Extending the HMO even further, without restriction as to form of ownership, serves to attract support also from other health insurance plans and business corporations.

BRINGING COSTS UNDER CONTROL

The main interest of the Nixon Administration was to do something to bring health expenditures under control, and prepaid group practice has a record showing lower per capita expenditures than other forms of health financing and delivery. The major source of savings in prepaid group practice is in hospital use, and the most widely quoted figure on the magnitude of the savings is 20 per cent, popularized by Donald Straus, consultant to the Rockefeller Panel Reports. It is noteworthy that by 1960 it was no longer taken for granted that more hospital care for a population is necessarily better.

Given the desirability of savings in hospital use, three questions arise concerning the reports on prepaid group practice. First, prepaid group practice is usually associated with a tight bed supply, as in California, or with difficulty in access to a hospital, as in New York City. Is the effect attributed to prepaid group practice confounded with these other control mechanisms? Second, how are the savings attributed to prepaid group practice to be explained? Thus, why is the length of stay the same in prepaid group practice as in other insurance plans? If the admission rates are equal, a longer average duration of stay is to be expected for the population with a more selective hospital admissions policy. Furthermore, the 20 per cent savings figure in New York is over a matched insured population and in California, over the population at large. Should it not be higher in California, where physicians in prepaid group practice share in hospital economies? Third, prepaid group practice in this country has been limited to five million persons and to a few thousand physicians. It has been out of the mainstream of medicine, attracting doctors and clientele with a special taste for this form of medical organization. Is it proper to extrapolate past experience to much larger numbers, such as the 60 million subscribers projected by the Committee for National Health Insurance?

Proponents of the HMO have also claimed possible savings due to economies of size in the medical firm; at present there is no evidence for this. It is also argued that solo-practice fee-for-service medicine is inherently inefficient, so that