cost of equipping a suitable operating suite runs about $30,000, and the personnel required include four physicians, with intensive training and skills, one to three registered nurses, and one to three skilled technicians. An AMA survey indicates that the cost of hospitalization for open heart corrective procedures averages $1,732.15 Apart from the special equipment and the physical plant necessary to support the treatment of patients with new technology, hospitals also bear the expense of providing personnel with high degrees of technical skills.

A concomitant result of an increased level and intensity of hospital technology is the increase in pressure from the patient who expects the latest advances in medical technology to be applied to his particular case. At the same time, as a greater segment of the population becomes able to pay for hospital care, hospitals are experiencing an increased demand for implementation of new and expensive techniques. University medical centers, especially, are expected to provide the newest and best in medical technology. The likelihood is that the rate of medical advance will continue to be rapid, especially in light of the current explosion of medical and technical knowledge.

The Future of Hospital Costs

A number of factors related to the major components of hospital cost are beginning to cause a change in the rate at which hospital costs are rising. These factors will have their effect primarily on hospital use, the cost of hospital labor, and the cost of hospital facilities. Federal legislation, such as the Medicare program, the Minimum Wage law, and the Heart, Cancer, and Stroke legislation will influence these factors in various ways.

Centralizing Role of Hospital-The effect of the Medicare program and other socioeconomic forces of this century indicate that the hospital will continue to occupy a central role in the nation's medical care system. Physicians and patients increasingly prefer the hospital as a setting for medical care. Clinical education is conducted primarily in hospitals. Many physicians prefer to associate with hospitals which have facilities available for the furthering of their own specialized educational needs. Additionally, the recent trend in federal legislation leaves little doubt that the hospital is the place from which medical care is most likely to be disseminated. The Heart, Cancer,

15 Commission on the Cost of Medical Care, American Medical Association, Report of the Commission Vol. 3, Significant Medical Adences (Chicago: The Association, 1964), p. 62.

and Stroke legislation of the current administration not only increases the likelihood that new technology will be implemented in medical centers but also serves as an impetus to community planning for health facilities. Secretary of Health, Education, and Welfare, John W. Gardner, has recently announced that $4.1 million have been awarded to ten states toward establishing regional medical projects to combat these three major killers.16

Consumer Attitude-Tied in with the concept of the hospital as a centralizing agent in the medical care system is the nature of the consumer's attitude. Adequate medical care is increasingly viewed as an essential rather than as a privilege. This view of medical care can only increase the demand on hospitals created by unmet needs existing in the population at large. As these needs come to the surface, those who provide care will come under increasing pressure.

Service Component-Since the major part of what the health consumer buys is specialized personal service, large productivity gains are not immediately foreseeable. This does not mean that operating economies are not possible. But it does mean that the hospital system is highly resistant to the type of cost reductions which have been brought about in business and industry through the introduction of mass production, labor-saving devices, and increasing centralization of controls.

Wage Rates-Most of the legislation of the last decade having to do with the acute general hospital of the United States has meant increased income for participating institutions. Public Law 89-97 (both Titles XVIII and XIX) will provide a steady stream of income to hospitals that were formerly pressed to come up with sources of money to pay for care given to medical indigents. This fact has not been lost on hospital personnel who feel that their claim to higher wages no longer threatens the hospital financing matrix. A fair portion of the total income is more readily demanded now that any feelings which may have been associated with undermining a charitable institution have been dissipated with the rise in hospital income. Any doubt about the reluctance of organized nursing to demand higher wages would have been dispelled after the recent policy statement by the ANA that starting salaries for nurses should be a minimum of $6,500.

The recent legislation for application of federal minimum wage standards to hospitals will also increase the cost of hospitalization, particularly in those states which do not now meet federal

16 Benjamin Wells. "U. S. Plans Inquiry Into Medical Costs," New York Times (August 24, 1965).

standards. Although southern institutions will be particularly hard hit by this legislation, every hospital will be affected to some degree, since many hospital workers are in the semi-skilled and unskilled categories-those most likely to be affected by the minimum wage legislation.

More Personnel-Another effect of Public Law 89-97 will be an increase in the demand for all kinds of health field workers. Even though the long run effect of Medicare will be to increase the number of workers in this field, the availability of personnel already lags behind the need.

Unions, too, have had and will continue to have an influence on labor supply, particularly through demands for a reduced work week. Partly because of the reduced work week, from 48 to 40 hours a week, there were 148 employees per 100 patients in 1946 and 241 in 1963. Any increase in the organization of hospital employees in unions or union-like group advances the prospect that hospitals will be pressured into paying higher wages for shorter hours.

Future Hospital Costs

Increases in the total cost of hospital care are reducible to three components: population increase, increases in the utilization rate, and increases in hospital cost per day. As the component that is increasing most rapidly, hospital cost per day has received the most attention in recent months.

The population of the United States has been increasing approximately 1.5 percent per year. The Bureau of the Census population projection for 1970 of 207,127,000 is a continuation of that rate.17

Utilization, annual hospital days per thousand population, has been increasing at approximately 1.7 percent per year since 1950, and, during the most recent four years, at an average of 2.4 percent per year. The hospital cost per day has increased 7.2 percent per year since 1950 and 6.2 percent per year for 1961 to 1965. Since 1950 the increase has ranged from a high of 9.4 percent in 1952 to a low of 4.5 percent in 1956. Estimates of the American Hospital Association indicate a 9.8 percent increase in hospital costs for 1966. The ability to predict future hospital costs from past trends, then, hinges primarily on the ability to predict the increase in hospital cost per patient-day.

A prediction of the future may be made in a number of different ways. It can be a simple

17 "Revised Projection of the Population of the Largest Standard Metropolitan Statistical Areas," Current Population Reports Series P-25, No. 329 (U. S. Department of Commerce, Bureau of the Census (Washington, D.C.: U. S. Government Printing Office, March, 1966]).

"best guess," or it can be a fairly precise projection based on an extensive understanding of the situations which are likely to effect a change. In predicting future hospital costs a simple guess is not good enough. On the other hand, while we have a good deal of knowledge about the forces changing hospital costs, we are not fully able to understand the way in which all these forces will act in the future. This is a common problem in prediction and is often resolved in business and industry by projecting the future on the basis of past trends. In the absence of the ability to read the future, the projection approach is better than simply conceding that no prediction can be made. It gives a fairly precise projection of the future based on assumptions about the past to which present knowledge can be related and allows us to gauge other predictions about the future in terms of deviations from the trend.

There are a number of ways in which a projection of the future can be made on the basis of past information. In general, the best projec tion is the one that best fits the data. Hospital cost per day, as Chart II shows, have increased over the past by an increasing amount. Such a change can best be described as a parabolic trend.

Two prediction models have been developed, both parabolic trends, to describe the change in hospital cost per day. The first trend is based on experience from 1950 and has the advantage of describing the hospital cost per day over a rather long period of time. The second trend is based on experience from 1961; while this provides a fairly short base period, it does appear from the data that a change in the trend may have occurred about 1961 which is not adequately reflected in the 1950 based trend line. These two projections are shown in Charts III and IV. Both charts, in addition to showing the parabolic trend from the base date to the present, also show three separate projections for 1967 to 1970. Each of the three projections on the charts reflects a different assumption about recent changes which are occurring in hospital costs.

The lowest projection through 1970 on both charts is based on the trend through 1965. This projection assumes that the nature of the forces affecting hospital costs from the base dates has not changed in quality or extent and will continue to operate in a similar manner in the future. This trend line assumes, for example, that salaries, personnel per patient, skill level, and medical technology will continue to increase at about the same rate that they have between 1950 and 1965 (Chart III) or between 1961 and 1965 (Chart IV). While these trends result in the lowest projections, they

show a future increase in cost per day of between 5.4 and 7.9 percent per year through 1970.

When 1965 is used as the end point for projection, however, the 1966 cost per day is different from what one would expect and to a greater degree than is desirable. Using the trends from 1950 through 1965 and 1961 through 1965, the predicted 1966 cost per day was $46.93 and $47.84. This was $1.90 and $.99 less than the American Hospital Association estimate of $48.83 per day for the year ending September, 1966.18 Both amounts are above the $.78 and $.12 variations which might be expected as a maximum from the trend lines ending in 1965, based on prior variations. Based on the variation in cost per day in past years, there would appear to be new factors present this year, or factors of a different magnitude this year than before.

Even since the American Hospital Association's estimate for the year ending September, 1966, the trend of increasing hospital cost per day has continued. The cost per day for the months of September through December were 10.0 percent, 11.0 percent, and 13.7 percent more than those same months of 1965.19

18 "Hospital Indicators," Hospitals, Vol. 41, No. 2 (January 16, 1967), PP. 21-22. "Hospital Indicators," Hospitals, Vol. 41, No. 6 (March 16, 1967), pp. 23-24. (All cost per day figures are based on the year ending in September which is the standard AHA index. Figures based on a year ending in December, in a climate of increasing cost, would be slightly higher.)

In examining data from the Federal Employee Program, a major change in the rate of increase in cost per day can be seen beginning in March' of 1966. Prior to that time, the average change in hospital cost for each month over the same month one year prior was 6.5 percent. Since March of 1966, the average change has been 9.1 percent, with peaks of 11.1 percent for March, 1966 over March, 1965 and 13.3 percent for December, 1966 over December, 1965.

The highest projection by 1970 on Chart IV and the middle projection on Chart III represent a trend in both charts based on a parabola computed through 1966 rather than 1965. These projections, especially the 1961 based projection, give a great deal more weight to recent changes in hospital cost. In essence these projections assume that the recent demands of nurses for higher wages and shorter hours, as well as all other forces which are changing hospital costs, began to make a major change in the nature of the trend lines themselves in the last year. On the basis of the 1966 ending date, and depending on the base date, a prediction of 6.3 or 9.4 percent per year increase may be made through 1970.

These two trends come closer to the 1966 cost

19 Statement of Mr. Mark Berke in behalf of the American Hospital Association on H. R. 5710, "Social Security Amendments of 1967," to the Ways and Means Committee, U. S. House of Representatives. March 8, 1967 (Processed), p. 8.

estimate of $48.83 than do the projections based on the trend to 1965. In Chart III the cost per day projection is $1.10 below the American Hospital Association's estimate and in Chart IV it is $.12 below. While $1.10 difference is fairly large relative to a standard error for the long trend of $.47, the $.12 difference is less than one standard error of $.22 for the short trend.

The remaining projection through 1970 is not, strictly speaking, a trend projection. On both charts, this projection is made by assuming that the recent changes in hospital cost do not represent the beginning of a radically different trend line from that which has been evident over the past, as does the projection through 1966. Nor does it assume that these changes have not had a real influence on hospital costs. These two points assume that the recent demands for higher wages have created a shift in the position of the trend. At the same time, they assume that the basic trend through 1965 will be resumed once wage demands have been met.

It has been estimated that increases of 10 to 20 percent across the country in nurses' salaries over the level of their 1965 salaries would bring

them fairly close to the salary level of teachers. This is significant. Only about one nurse in four has comparable academic training, and the average working life of a nurse is not much more than five years. Assuming, however, that nurses were to get general increases of the magnitude of 20 percent, it may be reasonable to expect other nursing personnel to demand and receive salary increases of as much as 13 percent over their 1965 salaries. With this type of increase for nurses, all other personnel might get 10 percent increases. If this were to happen, we could expect hospital costs by 1967 to go as high as $53.43 per day on the basis of our long-term trend, an increase from 1966 to 1967 of 9.4 percent and an overall increase from 1965 to 1967 of 20.1 percent. On the short-term trend, we would predict a cost of $55.30 per day, an increase of 13.3 percent from 1966 to 1967 and 24.3 percent from 1965 to 1967. This projection assumes, of course, that the same type of forces operating on personnel per patient, changes in technology, and other factors affecting hospital cost on the trend up to 1965 will continue to operate, in addition to the trend-projected increases in salary levels through 1970. These two

projections represent an average yearly increase of 7.0 and 9.1 percent from 1965 to 1970.

Charts III and IV show six trend lines. Under the assumptions of these trends, six projections can be made for hospital costs per day both in 1967 and in 1970. For the trends ending 1965, a cost per day of $49.69 in Chart III and $51.56 in Chart IV would be projected for 1967. This is 1.8 and 5.6 percent above the 1966 estimated cost of $48.83. By 1970, on these two trends, cost per day would be up to $58.56 and $65.03 respectively. For the trends ending 1966, the cost per day in Chart III by 1967 will be $50.70, an increase of 3.8 percent over 1966. In Chart IV, the cost per day will be $53.13 in 1967, an increase of 8.8 percent. By 1970 the cost per day on long-term trend will be $60.31 and the short-term trend, $69.79. The cost per day by 1967 for the final projections have been given. By 1970, these projections should show a cost per day of $62.30 in Chart III and $68.77 in Chart IV.

It should be remembered in examining any of

these projections that they represent "point" estimates. The probability that the actual cost at any time in the future will fall precisely on one of the points of the charts approaches zero. Depending on the standard deviation of any one of the trends, however, the range within which the actual future figure is likely to fall may be predicted with a fair degree of accuracy. One could be about 95 percent certain, for example, that the projection for 1967 based on the 1961 to 1966 trend line (if the assumptions of this trend are realistic) would fall between $52.69 and $53.57. The size of this range, of course, depends on the extent to which actual cost has followed the trend line in the past. These projections present a possible range in hospital costs of $11.23 by 1970, from a projected low cost per day of $58.56 on the long trend through 1965 to a projected high of $69.79 on the short trend through 1966. In all likelihood the future increases of hospital cost should put them somewhere between these two extremes.