Read an Excerpt

Physicians and Hospitals

The Great Partnership at the Crossroads Based on the Ninth Private Sector Conference, 1984

Duke University Press

The Cybernetics of Our System

ANDREW G. WALLACE, M.D.

It seems reasonable to assume that our topic for this conference, "Physicians and Hospitals: The Great Partnership at a Crossroads," would not have been chosen if it was not perceived as a problem. I am new to the ranks of medical administrators, while there are others at this conference who have been in the trenches longer and viewed our subject from a wider range of experiences. Some have actually developed strategies to relieve the stress that is implied by our title. For these reasons I have chosen to search my own background for experiences that might complement yours and provide each of us with a useful framework for thinking about our topic. There is a risk in this departure, but there is also a compelling need to orient our dialogue toward causes rather than symptoms, toward a process that seeks solutions as opposed to palliative procedures.

Consider that the interface between doctors and hospitals is just one special case of a much broader issue: how to understand and constructively steer the behavior of complex systems. As in any system made up of interdependent parts, the interaction of the components of the system will determine how we progress from the present to the future. Most systems that I am familiar with are characterized by change, growth, and goal seeking and are governed by feedback. For example, feedback is a characteristic of the operation of biological systems, of interpersonal relations, of corporate behavior, and of large social systems. I believe that there are lessons we can learn from those who have studied feedback, and that these lessons have application to the problem we are discussing and to the formulation of policies that will lead to real solutions.

My first exposure to the concept of feedback came in the study of biological systems. As a student in biochemistry, I struggled to memorize a host of biosynthetic pathways responsible for generating compounds essential to the function of living cells. I have forgotten many of the specifics, but I remember the important concepts, including that of end product inhibition. The principle is illustrated by figure 1.1. In this example the conversion of substrate A to intermediary B is catalyzed by enzyme E1, and the conversion of intermediary B to the final product C is catalyzed by enzyme E2. The important point is that the level of product C modifies the activity of enzyme E1. This intrinsically simple control mechanism recurs throughout biological systems and is, in fact, the basic process by which physiologic systems achieve equilibrium.

Claude Bernard, who is regarded by many as the father of modern physiological thought, wrote the following in the middle of the nineteenth century:

Science should always explain obscurity and complexity by clearer simple ideas.... There is an arrangement in the living being, a kind of regulated activity which must never be neglected, ... it is in truth the most striking characteristic of living beings.... It is as if there existed a pre-established design ... such that though considered separately each physiological process is dependent upon the general forces of nature ... yet taken in relation to other processes it reveals a special bond and seems directed by some invisible guide....

The "invisible guide" in Bernard's work and thought became the basis for the concept of feedback control in biology. Does this concept apply to the arena in which doctors and hospitals work? I think it does. Physicians are products of an educational system that catalyzes the conversion of students into medical manpower. Yet today it is primarily physicians, from both practice and academia, who are calling for a reduction in the rate of production of their own kind. Hospitals, too, are the product of forces that over the last twenty-five to fifty years have called for expansion of facilities and greater access. Yet today, either in the open or by subtle means, hospitals are exerting pressure to constrain their own further proliferation. The dynamic of feedback is with us.

For a period of time science was satisfied to look for evidence of feedback, or its lack, in various biological events. But soon the process of feedback itself became the subject of intense investigation. There are concepts derived from this effort to understand the regulation of biological systems that have their analogies in social systems. We have learned that feedback is often triggered not by the level of a product per se, but rather by the difference between the actual output of a reaction and either the genetically determined or evolutionarily desired level of output. We have learned that the compensating force in a feedback loop is related to this difference but not necessarily equal or directly proportional to it. We have learned that the change in the level of product over time, when a process is perturbed, is influenced importantly by the delay in the feedback loop (figure 1.2). In this figure A is the level of substrate. B is the level of product. S is a sensor designed to detect the difference between B and its desired level D. The compensatory force C is related to this difference, but delayed in time and quantifiably related to the difference by a transfer function or process T.

Consider for a moment that this model might describe either the production of physicians or the availability or cost of some arbitrary unit of health service. Do we know the desired level of product? Do we have a sensor adequate to detect the difference between the actual and the desired level? What is the quantitative relationship between this difference and the compensatory signal? How long is the delay in the feedback loop—for example between the GMENAC report and a change in the rate of production of physicians that is equal to their attrition rate?

The range of possible dynamic responses of a feedback system when you perturb it in some way is illustrated in figure 1.3. We see that (1) shows what would happen without feedback, (2) is the desired response of a stable system with appropriate feedback, and (3) is a system with appropriate compensation but long delay in the feedback line. Which curve best describes the growth of physicians or of medical costs?

The word "cybernetics" is newer than the concept of feedback. Norbert Wiener, a mathematician from M.I.T., began to work on the theory of messages about forty years ago. He was interested in engineering, in language, in biology, and in psychology. He created the word cybernetics, which was derived from the Greek "Kubernetes," meaning "to steer." In 1950 he wrote a book, The Human Use of Human Beings, in which he developed the theses that social systems could be understood only through a study of feedback messages; that stability required that the results of one's actions be communicated back as a part of the information upon which we continue to act; that cybernetics is the only force that prevents the natural tendency toward disorganization.

Jay Forrester, who is also at M.I.T., applies cybernetics to social systems. About twelve years ago he wrote World Dynamics, a book based on studies sponsored by the Club of Rome and best known because of its doomsday predictions. Forrester is a member of a community of model builders—global models designed to capture and understand and forecast the behavior of complex social systems such as a corporation, a city, the economy, or the world. His thesis is that social systems belong to a class of multilooped feedback systems and that it is possible to construct in the laboratory realistic models that describe and predict their behavior.

Although simplified by excluding a multiplicity of factors that actually operate in complex social systems, figure 1.4 illustrates the nature of forces that ultimately change such systems from a growth mode to an equilibrium mode through feedback. In the illustration, the upper loops promote growth and the lower loops produce restraint. The idea is that people are attracted to an area of fertile land and their labor increases agricultural capability. Food per person increases and the rising food supply supports further increases in population. Growth continues until the marginal productivity of an additional worker fails to produce enough food to support growth. This stress may trigger a temporary investment in new technology to augment agricultural capability, and with this capability food per person is again lifted above the subsistence level and population growth continues. During growth, however, the population density increases, people begin to occupy the best agricultural land, and crowding creates pollution, social conflict, and a shortage of resources. The consequence of growth is to induce ever-rising pressures to restrain further growth. In time the forces of growth and restraint come into balance, and growth gives way to equilibrium.

Forrester's thesis would be that models of this type describe the growth curve of most organizations. By extrapolation, his model also predicts that for most social systems including world population there is a limit to growth. My reason for elaborating the Forrester philosophy is my belief that it has application to the health care enterprise and to the issue we are here to discuss—physicians and hospitals. Whether you view health care as a system unto itself or as a part of a larger social system, several generalizations can be drawn from the dynamic behavior of complex systems.

1. Together we have cultivated an expectation that our enterprise can promote health, increase longevity, and increase productivity. The accomplishments of the past give credence to that argument. We are now investing in technology as an expression of the continuing expectation of what can be accomplished through the health enterprise. The public has supported that view and so we are operating in the upper area of this particular growth curve, an area of positive feedback and growth. On the other hand, any thoughtful projection of what changes our enterprise is likely to produce in terms of health, longevity or productivity over the next fifty years compared with the last fifty years indicates that those changes are going to be smaller, more difficult to accomplish, and more costly.

In addition, we are bumping into a progressive series of ethical questions. We have technology today that we know how to use but not when to use. Nowhere is that dilemma greater in the operation of the hospital than at the two ends of the age spectrum.

My point is that we're beginning to see some of the forces of negative feedback. We're beginning to see some of the forces of restraints on our system and, from my perspective, the growth curve of the health care enterprise is beginning to turn to a slower rate of growth.

2. Forrester points out that the transition from growth to equilibrium is always characterized by stress. One reason is that problems previously solved by expansion now must be solved by setting priorities within the constraint of relatively fixed resources. Policies that served well during growth no longer apply. I suggest that the stress today between physicians and hospitals is a predictable symptom of a system in transition from growth to equilibrium. If that hypothesis is correct, let us not assume that the supply of new physicians, new entrepreneurial practice styles, and new hospital strategies will all be additive. There will be winners and losers, survivors and casualties in the new medicine.

3. A third concept that can be drawn from the Forrester view is the "counter-intuitive behavior of social systems." The basic idea is that in the absence of a conceptual model that describes and quantifies the relation between all components of a system, people of good intentions will establish and follow policies designed to optimize the performance of their own component believing that in so doing they will help to improve the operation of the entire system. When a system is in trouble and these operative policies are identified, combined, and allowed to interact, the consequences usually mimic the actual difficulty of the system.

One of my fears is that as doctors, hospitals, government, and business seek new ways to optimize their own conditions, in the absence of an adequate understanding of the negative consequences of their actions for other components in the system, the health care system as a whole becomes vulnerable to a worsening condition. The capacity of doctors to optimize their practice style without regard to hospital operating costs is legend, and the capacity of hospitals to adopt strategies to cut costs without appreciating the adverse impact on doctors is also legend. What we seem to lack is a model that defines these relationships, that allows us to predict the short-term and long-term consequences of a particular action and to seek policies that may require compromise of individual goals in the short run for the sake of overall system performance in the longer run.

I will conclude this brief and risky venture into the world of cybernetics with a few thoughts.

1. The principle of feedback is fundamental to understanding the behavior of simple and complex systems. It applies to the interface between physicians and hospitals and to the interface between health care and other components of our social system. It is our responsibility to explore relationships that cultivate common objectives, promote feedback, achieve synergy, and avoid adversarial postures.

2. If doctors and hospitals are interdependent, then the basic question is whether or not they are coupled by policies that facilitate feedback and optimal achievement of a shared goal. I believe the linkage leaves much to be desired, and the perception that our goals differ contributes significantly to the stress between doctors and hospitals.

3. Even if the feedback loops exist, there is reason to ask whether the phase lag between a particular action and its consequences isn't so long that the feedback is ineffective. The best example I can think of is, how long will it take to bring the rate of production of physicians into balance with the rate of attrition?

4. I confess I do not know what the relationship between the vast range of physician practice styles that is emerging and the equally vast range of hospital styles should be. My current view is that the plethora of styles is more a symptom of stress than a solution. They smack of encircled wagons, or self-optimization without a clear vision of their effect on other parts or on total system functions. I admit that some surgery can be done well and at less unit cost in a surgicenter, but I haven't been convinced that total health costs of a community will be lowered. A "doc-in-a-box" may do well today, but what will the quality of his practice be ten years from now, how will he know, and what will we do?

5. I return to Forrester's model because I believe we are, or shortly will be, headed for an excess health care capacity relative to resources, a host of negative feedback forces designed to reduce the rate of growth, and predictably a period of increasing stress. We seem to lack a conceptual model of the whole health care system, one that allows us to predict the distant consequences of a given policy. We simply have not yet faced the reality that to simultaneously seek quality, equity of distribution, freedom for providers, and cost control is impossible. A group such as this one gathered here has the clout to recommend policies that will constrain growth and the proliferation of dysynergy within the health care system. A group such as this has the capacity to create a model that defines the relation between components of our system and to estimate the consequences of a given policy on other components and on the performance of the total system. A conference such as this should set for itself the goal of understanding the relation between doctors and hospitals and formulating strategies that will facilitate their interaction toward the broader social purpose both serve. Physicians and hospitals are interdependent, and they are mutually dependent on achieving that broader purpose.

Norbert Wiener, The Human Use of Human Beings (Garden City, N. Y.: Double-day, 1954).

Jay Forrester, World Dynamics (Cambridge, Mass.: Wright-Allen Press, 1971).

Jay Forrester, "Churches at the Transition Between Growth and World Equilibrium," Zygon; Journal of Religion and Science 7(1972):145. Reprinted with permission of the author and publisher.

Jay Forrester, "Counterintuitive Behavior of Social Systems," Technology Review 73(1971):52–68.

---

Excerpted from Physicians and Hospitals by Duncan Yaggy, Patricia Hodgson. Copyright © 1985 Duke University Press. Excerpted by permission of Duke University Press.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.

---