against competing requirements in the agencies. It is somewhat doubtful whether the various peaceful applications of nuclear energy could have been developed without being nurtured through their early stage by an agency which was specifically charged with expoiting nuclear energy in all its ramifications without limitation to particular fields of science or particular agency missions. In a sense the AEC was given a license to operate in the field of all the other agencies and of private industry, provided only its activities were interrelated among themselves by the common theme of nuclear phenomena. Thus we had an agency which was developed around a technological theme rather than a social mission. It could and did enter the areas of biology, agriculture, metallurgy, meteorology, civil engineering, energy production, ship propulsion, space propulsion, and waste management. It developed capabilities and promoted skills which touched on almost every aspect of American society.

The difficulty with the technological theme approach is that, once it is successful, there are few tools by which the theme can be kept in proper relationship with the various social goals which it serves. The continued development of the technology tends to become an end in itself to be pursued as a national goal in its own right. This is all very well when the technological goal is a small part of the national effort, and when it is pursued on a small scale in Federal laboratories and universities. But when it becomes a major national effort, then the problem of balancing with other social goals becomes acute. What is lacking, of course, is a feedback mechanism which provides a measure of how much is enough. The President and the Congress, for example, find it very difficult to obtain a really objective assessment of the relative benefits to be achieved through investment in nuclear power as compared with the development of hydroelectric resources or improvements in coal technology. It is similarly difficult to get a clear picture of the relative emphasis which should be placed on water desalting in comparison with, say, reprocessing of waste waters or large scale water importation. The expertise in the various technologies involved tends to be unevenly represented in the decisionmaking process, and there is no automatic mechanism like the market to redirect effort into the most effective or promising channels. Except possibly in the Defense Department, there is no place in the Federal structure where the President or the Congress can go to obtain a disinterested and systematic exposition of alternatives for applied research and development programs. In principle this can be done by the staffs of the executive office agencies, the Bureau of the Budget, and the Office of Science and Technology, but in practice these are forced to rely heavily on the analyses and arguments prepared in the agencies themselves. Furthermore, we have not as yet developed the techniques that would make such analyses possible on a sufficiently convincing and objective basis to command acceptance by the affected agencies and institutions.

In many ways the case of biomedical research has developed in a way which is intermediate between that of the technological theme and the social mission. On the one hand, the NIH has developed into an agency which is very broadly based in the life sciences, or-if you prefer-biomedical technologies. In this sense it has come to mean for the biomedical sciences what the AEC has meant for atomic energy,

or NASA for space science and technology. On the other hand, it is organized by disease categories, and the allocation of its resources tends to be more in terms of the perceived social importance of certain diseases rather than in terms of the opportunities for technological progress in the application of biomedical techniques regardless of disease. It is thus an interesting hybrid between the highly productoriented research of Agriculture and the highly technology-oriented research of AEC. The small feedback loops by which NIH controls its program are rooted in scientific criteria, but the large feedback loops which govern its budget and its broad allocations are primarily governed by the disease categories and health problems, and what is perceived as their social importance. This hybrid system has on the whole been remarkably effective.

There is no really satisfactory resolution of the issue of interagency coordination versus technological agencies. Probably the creation of a new agency becomes inevitable when support of a new technology reaches a sufficient fraction of the budget of existing agencies, or when the new technology is sufficiently revolutionary or has sufficiently broad potential ramifications. Certainly the breadth of application of the technology is an important criterion for the formation of a new agency. The technology orientation and the social orientation represent different cuts of the same complex of problems, and if we organize along one dimension we inevitably require coordination along the other. An agency organized around technology runs the risk of unresponsiveness to social needs, and this risk increases as the technology matures and the range of potential applications increases. There is no natural internal mechanism for ordering priorities among these applications. Even interaction with the political process tends to be ineffective in this regard because Congress matches its committee organization to the agency, so that a technological agency is supervised by a technological committee which has a natural inclination to expand its own area of responsibility. On the other hand, the technological agency does keep abreast of potential opportunities, while the socially oriented agency often tends to be well organized for obsolete problems and to be unresponsive to new technical opportunities. Its horizontal communications are too weak to adjust to technical change, and it is easier not to change. Furthermore, the mission of such a "social" agency tends to be defined in terms of the particular technologies which were most relevant to it when it was created. Even within the very progressive and forwardlooking Public Health Service I believe there is a tendency to underrate the future importance of the physical sciences and engineering in the achievement of health missions, and to define its missions too exclusively in terms of the life sciences, albeit broadly and imaginatively conceived. This is especially true with respect to problems involved in the organization, management, and mechanization of the delivery of medical care. Furthermore, as indicated by the studies of the Wooldridge committee, the quality of effort supported by NIH in the physical sciences probably falls short of the excellence achieved in the life sciences and in biochemistry. I suspect this is an institutional lag which is being rapidly overcome.

One of the byproducts of our departmental system of managing R. & D. is the creation of large Federal institutions devoted to the

missions of particular agencies and subdivisions within agencies. As the organization of the agency becomes obsolete, so does the mission of the laboratory-a process that may even be accelerated by success. The range of technological capabilities represented by these laboratories is extraordinary, but after a few years the magnificent machinery tends to be devoted to less and less significant problems, and it is extremely difficult to redefine their missions in response to the changing goals of Federal science. These laboratories-in-house and captiverepresent nearly $4 billion of annual national investment, nearly three times our support of universities proper, yet we seemingly find it very difficult to redeploy our national scientific apparatus to deal with new technical problems and missions without creating wholly new institutions or agencies or sometimes turning to the private sector when it is inappropriate. We do not treat our Federal laboratories as a common national resource to be used flexibly for many of the purposes of government. Rather we tend to regard each laboratory as the relatively inviolable preserve of the agency to which it belongs. This contributes to a high sense of agency responsibility and clear lines of management, but I wonder if it represents the best use of scarce national resources in scientific management and skilled manpower. I feel it is time we learned how to use these institutions more flexibly for national purposes with less worry about roles and missions. I admit there is a risk here an increased risk of that shibboleth of management "duplication and overlap." Furthermore, such institutions do not necessarily include the proper mix of skills to shift from, say, nuclear power to biomedical technology, but I believe that with better long-range planning we could redeploy some of these institutions over a period of years, and without merely adding appendages (and budget) to the existing laboratories. In fact, there is evidence that the laboratory managements themselves are approaching this area more imaginatively than the supervening Federal apparatus.

If we were to move in the direction I have indicated the management of Federal science might become considerably more complex, and the problem of dilution of agency responsibility more difficult. On the other hand, such a change of philosophy would lead to quicker identification of new missions and goals at the working level, and would have the virtue of bringing to our national thinking fresh minds, uncontaminated with the knowledge of what can't be done. Such organization, as Weinberg has said, would have the intellectual strength and range to develop new doctrine; that is, to reformulate the problems, and identify the underlying scientific issues. I think we could afford to pay the price of some initial naivete and fumbling for the sake of the fresh look, the breaking of obsolete communications patterns, and the formation of new feedback loops which might result.

I think also in this same connection, Government may have certain lessons to learn from private enterprise. If we circumscribed their missions less, I think we could encourage a greater entrepreneurial spirit in some of our laboratories, and permit some labs to grow and others to decline as they were able to sell their skills and their ideas to other Federal agencies. At least it might be worth experimenting with a kind of limited market economy within the Federal struc

ture and a modest amount of competition with some of its benefits and penalties. Such a scheme would foster the horizontal transfer of technologies to new missions which I believe to be increasingly necessary to innovation and progress within the Federal structure.

BIOMEDICAL RESEARCH IN THE VETERANS'
ADMINISTRATION

By Benjamin B. Wells, Ph. D., M.D.

Assistant Chief Medical Director for Research and Education, Department of Medicine and Surgery, Veterans' Administration

It is quite impossible to understand the present position or the future potential of the Veterans' Administration in the field of biomedical research unless you first have in mind the overall purposes, the dimensions, and the general nature of VA's total medical program. At no time can we overlook the fact that it is the prior existence of the VA patient care mission and the needs arising from it that have created, maintained and determined the character of our biomedical research efforts. For better or for worse, the VA research program is hospital based and health service oriented. These circumstances provide certain unique opportunities, but they also pose certain limitations within which we must fit both the content and the extent of our research work.

In January of 1946, following World War II, a Department of Medicine and Surgery was created in the Veterans' Administration by Public Law 293. This law put the new Department under the supervision of a Chief Medical Director and specified, among other things, an Assistant Chief Medical Director for Research and Education. Also in January of 1946, the VA took its most crucial and decisive step toward the assurance of high quality medical services by the issuance of Policy Memorandum No. 2. This administrative policy established the Deans Committee mechanism under which more than half of the hospitals in the agency have achieved a valuable working relationship with medical schools and universities throughout the country. Although many adaptations have been made within it, the fundamental structure of this organization has not been changed in the past two decades.

In more specific reference to the VA research program, let me give you just a fragment of its historical background. Prior to 1946 there had been no significant research in the agency. During the last months of World War II, Dr. Michael E. De Bakey recognized the tremendous potential for biomedical research that was coming into existence because of the millions of returning veterans who had medical problems which were being carried over from military into civilian life. Dr. De Bakey appreciated the scientific value of long-term clinical records and the relatively high degree of patient control that could be exercised in the veteran population. By enlisting the support and cooperation of the armed services, the Veterans' Administration, and the National Research Council, and by bringing these interests together in the Com

mittee on Veterans' Medical Problems, Dr. De Bakey became the father of medical research in the Veterans' Administration system. For entirely different reasons, VA was soon to realize that research was urgently needed in support of the patient care mission. Two needs. were especially compelling: first, there were many diagnostic and therapeutic problems more or less unique to the veteran population, and, second, opportunities to engage in research provided the best possible inducement for the recruitment and retention of high quality physicians and scientists. Thus, medical research in the Veterans' Administration began and has remained eminently practical in its objectives.

In the years immediately after World War II the VA was seriously lacking in laboratory facilities and technical personnel, and the agency hospitals and clinics did not have a proper administrative environment for the encouragement of medical research. To begin with, therefore, most of the work was done outside the agency through contracts with medical schools and university groups. As in-house capabilities were subsequently developed, the extramural contracts were rapidly phased out in favor of intramural programs. Finally, our research programs became exclusively an in-house effort, and VA has made only token use of its contract and granting authority for research since fiscal year 1953.

There was never to my knowledge a deliberate administrative decision to remove VA from participation in extramural research programs. The change came about simply because the research needs of full-time VA personnel exceeded the available funds. There can be no doubt that this pattern of evolution limited the total growth of our program, greatly inhibited the use of our hopsital and our patient resources, slowed the attack on certain clinical problems that are highly characteristic of the veteran population, and led to what we now recognize as a degree of deterioration in VA-medical school relations. One of the imperatives of our time is to reverse this process, to make VA resources and VA needs both more visible and more accessible to the academic medical community. Nothing can be more clearly evident than the fact that VA hospitals and clinics cannot sustain their high quality of patient care services if any degree of professional isolation is allowed to endure. Herein lies a difficult problem where statesmanship must prevail over management, where professional qualities must take precedence over organizational structure, and where cooperation must be given priority over institutional identification. Although this problem can be expressed with Biblical simplicity, the solution will no doubt require force and ingenuity almost beyond comprehension.

The VA medical research programs grew very slowly until 1956 at which time we had an annual research budget of about $5 million. Expansion was more rapid in the subsequent decade, so that by 1966 our budget had reached the level of about $40 million a year. The dimension of research in VA is not fully reflected in budget dollars. In the first place, we employ relatively few personnel on research funds. Practically all of the more than 2,000 physicians who conduct projects in our hospitals and clinics receive their entire salaries from patient care funds. Many others are detailed to us on a "without