The Roles of Academic Health Centers in the 21st Century: A Workshop Summary (2002)

WORKSHOP PROCEEDINGS

Uwe Reinhardt, Ph.D., Professor of Economics and Public Affairs, Woodrow Wilson School of Public and International Affairs, Princeton University

Academic health centers bring a unique and special set of values to health care. In the past, Dr. Reinhardt has often suggested that the private, for-profit sector could serve as a model for the management of academic health centers. The ability of the marketplace to discipline organizations and create wealth is evident in other industries and offers lessons for the management of complex organizations, such as academic health centers. However, in light of recent events in the business sector, such as Enron, he posited that while some techniques can be drawn from the business model, academic health centers need to meet a higher threshold of performance to fulfill their special role in society.

Under society’s “old” social contract with AHCs, the costs of conducting their missions in education, research and patient care were paid; in return, AHCs provided those services, as well as national pride and worldwide recognition. However, the management structure of AHCs has always been a “black box” of complex cross-subsidies, with little known about the actual costs of delivering these missions. Although the demand for such information was not requested in the past, pressures brought about by health care spending growth, imperatives for efficiency and demands for quality and accountability suggest this information gap may no longer be tolerable. Under a “new” social contract, questions will be asked as to how AHCs use the public funds they receive and how much funding is actually needed to conduct their work. Dr. Reinhardt believes that AHCs need to separate the accounting for their mission functions and that each should be explicitly financed; if there are cross-subsidies, they should be identified as such. The private and public good of AHCs should be identified clearly. Society should decide what it chooses to support, and to know that such support is being used for its intended purpose. One of the challenges in doing this is in defining ‘private’ versus ‘public’ good in order to determine what should be subsidized with public funds. Public goods (such as clinical research, public health measures and national reputation) should be fully financed with public dollars. Public goods with externalities (such as immunization and charity care for the poor) should be subsidized with public funds. Private goods (such as health care services, clinical trials and human capital) should be privately financed.

In Dr. Reinhardt’s view, medical education is no more a public good than other professional education and should be privately financed. Because the resident absorbs the cost of education through reduced salary, accrues the benefits of the education, and brings revenue to the AHC, it should not be considered a public good. In his view, there is no logic or reason for the variation in per resident amounts seen across the country, and it represents a significant lapse in accountability among AHCs. Furthermore, he believes that AHCs should no longer be willing to assume the role of caring for the uninsured. AHCs should either demand that Congress pay adequately for such care or AHCs should “get out of the insurance business.”

Finally, Dr. Reinhardt noted that, AHCs have traditionally performed their various mission functions while asking for additional resources to support them. Instead, society needs to determine what is required from AHCs and AHCs need to decide what they can do. An ongoing dialogue is necessary.

Jeff Goldsmith, Ph.D., President, Health Futures, Inc., Charlottesville, Virginia

Dr. Goldsmith emphasized three points in his comments. First, he spoke of the vulnerability of AHCs to economic cycles. The health of AHCs is dependent upon the vitality of the general economy and government. In times of economic recession, many effects are possible, all of which can potentially harm AHCs. To the extent that deficits occur at the federal and state levels, budget cuts may hurt the AHCs, that are safety net providers. At the same time, the number of uninsured may grow. Companies may change their benefit structures in response to rising health care costs. As a result, patients may have to bear a greater portion of health care costs and insurers may force profitable patients to shift their care to community institutions. It is a myth that AHCs are immune to economic cycles.

Secondly, Dr. Goldsmith spoke to workforce needs and his belief that a shift is occurring from an era of surplus of health resources (many beds, hospitals, and physicians) to an era of scarcity. Dr. Goldsmith believes there will be a shortage of health professionals, not due to an increased demand for health care services, but rather, because of the earlier retirement of physicians who are currently part of the “baby boom” generation. This group of professionals is “burned out” and additional increments in salary will not be sufficient to address the demands of practice today. Serious shortages exist for certain physician specialties, such as interventional radiologists, but also for pharmacists or for those with database expertise. As AHCs are faced with having to expand their training programs, a significant source of pressure will be exerted in recruiting, retaining and paying for faculty.

Dr. Goldsmith’s third point related to technology. Both information and biomedical technologies will require a rethinking and retooling of health care delivery. One of the major advances in information technology has been the maturation of the electronic medical record. This will be more than a “passive digital replication” of the written medical record. Instead, the electronic medical record will be a guidance and decision-support system that is able to find, acquire and deliver both patient-specific and comparative information to the bedside. However, according to Dr. Goldsmith, getting there is like “trying to rewire a car while it is running.” Unless people are taught to use these tools during their training, they are not going to be able to use them in practice. Another technological impact will be the rapid growth in the ability to screen people and pathogens for their genetic makeup. Eventually, our understanding about the basis of disease will enable us to assess those risks that are of a genetic nature and intervene strategically to modify them. The ability to genotype individuals and their pathogens, their cancers and their viruses will enable a more optimal mix of therapies and more effective use of expensive pharmaceuticals. Some have called this the emergence of personalized medicine. The challenge for AHCs is how to harness these technological opportunities and leverage the basic science being developed.

Finally, Dr. Goldsmith pointed out the need for better management of these institutions in the future. There is a need to sort out the products of AHCs and hold organizations accountable for producing them.

During open discussion, a number of issues were raised that can be grouped under three areas:

The first area of questions asked about the ability to define accountabilities for AHCs within the present system of cross-subsidies. Larry Lewin noted that although it may be desirable and advisable to understand each mission individually, many activities in AHCs are jointly produced. Therefore, the allocation of these joint costs to programs is subjective. Ralph Snyderman indicated that at Duke, they have indeed been able to account separately for each mission. However, some complexities and unpredictability remain because of narrow margins, unpredictable volume, care of the uninsured and lags of several months in the data needed for decision-making. Edward Holmes noted that although a better understanding of costs for each mission is needed for accountability, we also know that the education and research functions are not self-supporting. Even in a tightly managed organization, research requires approximately 15 cents for every dollar spent. Education is also not sufficiently compensated.

In response to this series of issues, Uwe Reinhardt responded that if AHCs are unable to separate their product lines and identify the cross-subsidies, the alternative is a budget approach for financing by which AHCs are given a budget to be used to maximize health. However, it is not clear how "health" could be measured for purposes of payment policy.

Jeff Goldsmith responded that one of the most troubling aspects of AHCs is the lack of accountability for their performance of the missions. He also believes that an organization can be "mediocre" in all three missions and suffer no consequences. The efficient use of resources as a goal is not a high enough bar for AHCs. Furthermore, he noted that mission accounting does not mean that every mission at every institution must be self-supporting, but each must function efficiently. The problem is that the decision-making for the missions is distorted if, for example, 20 percent of the clinical mission is carved out for research. The people focused on each mission need to be accountable for resource use, supported by an explicit strategy for funding the subsidies that may be required.

Different views were expressed in reaction to Uwe Reinhardt’s comments about what constitutes a public good. John Rowe questioned: 1) why AHCs should receive tax exemption for providing public goods if, as Dr. Reinhardt suggests, students pay for their own education and 2) the AHC role in caring for the uninsured should be re-examined. Robert Dickler suggested that education is indeed a public good and must be viewed in the context of education and the nature of medical education; however, he also agreed that this will be a topic of increasing debate. Roy Wilson challenged the assumption that for-profit AHCs are less involved in caring for the poor than non-profit AHCs. Although this is a common assumption, he suggested the need for better evidence. In response, Uwe Reinhardt expressed his belief that the only public good aspect of a medical education might be a particular ethic that physicians, as professionals, bring to their jobs, and perhaps that is the public good of medical education. Beyond that, however, it is hard to for him see the public good. Education is “human capital that sits on the skeleton of the person who has it and they can deploy it in any way they wish. They can do boutique medicine or they can do indigent care.” The objection of economists to education as a public good is that it is stated as an axiom rather than something that has been justified. In terms of tax exemption, Dr. Reinhardt suggested an alternative approach in which all institutions become taxable entities, but each time that charity care is provided to a patient, a tax credit would be issued as a credit against taxes owed. If a hospital provides enough charity care, no taxes would be owed. If no charity care is provided, taxes would be paid.

Paul Ramsey noted that there is a great deal of variation across AHCs. For example, some are research-intensive institutions that may require greater support for their research infrastructure. However, Nicole Lurie noted that there are likely variations in expectations as well. James Curran noted that all AHCs measure themselves differently and, as a result, identify themselves as being in the top quartile on some dimension. He asked how many AHCs are necessary, and what components are needed. Uwe Reinhardt responded that in his view, national policy should not support regional variations. It is a contradiction and represents a loss of control in setting standards and direction.

Ralph Snyderman noted that the health care delivery system is not designed effectively to improve health. For example, if diabetes could be identified as it happens, early intervention could be introduced to avoid complications. However, the system is unable to prospectively deal with health problems; it can only react. He asked whether AHCs will be drivers or passive participants in the progression toward more prospective care. In his view, AHCs have the potential to be leaders in developing best practice models for more prospective care because they have the intellectual capital, access to clinical information systems, and access to patients and other resources. However, he also noted that a profound cultural change is necessary to organize academicians. John Rowe suggested that although AHCs may be the best place for doing the "research and development" (R&D) function to develop models and pilot programs for improved delivery of care, he questioned if they are the best place for designing implementation strategies. For example, in his experience, when Aetna wanted to purchase disease management programs, the AHCs could not compete with private vendors on efficiency, price or other factors, even though the vendors were probably using models developed at AHCs.

Christine Seidman suggested there are several sides of the AHC safety net function other than caring for the uninsured. One is caring for the overly insured—those patients who choose a higher cost option and more care, perhaps in the face of a terminal illness. The second is support for the community physician caring for clinically difficult patients. The latter may become more important over time as biomedical technology creates a specialized knowledge with which the average community physician may need support. In response, Uwe Reinhardt stated that if the overly insured desire more care, they should pay for it, but society should not subsidize it. He also believes that in their own way, AHCs are part of the uninsured problem, in that the AHCs have always cared for the uninsured, “bailing out” Congress from having to deal with it. AHCs should be the most powerful lobby today in getting coverage for the uninsured so their organizations can be properly run.

Linda Aiken noted that care delivery and workforce issues intersect in terms of how patient care processes are designed and how the workforce is used. She asked whether AHCs have an “R&D” role in how the workforce is utilized and deployed to deliver care. Specific to this issue, Jeff Goldsmith said that resolving the workforce issue will require that AHCs become humane, efficient and thoughtful places to work. A generation of health workers has become exhausted because we have not been able to provide a “satisfying, intellectually stimulating work environment that is consistent with the values that brought them into clinical practice in the first place.” In his view, the workforce issues are huge and are central to the design and management of work.

Sara Rosenbaum, J.D., Harold and Jane Hirsh Professor of Law and Policy, George Washington University

Professor Rosenbaum spoke to the expectations of AHCs from the perspective of low-income populations. From her perspective, the issue is more than how much uncompensated care is provided by AHCs, but should include how AHCs reach out to poorer communities. She identified three approaches to community benefit. The first is the nature of the educational process itself and its ability to train people who can relate to the patients in their communities. Despite the growth of technology, much of medicine still comes down to the intangible relationship between physicians and patients. Current education may have the effect of “hardening” attitudes rather than helping students see the social and other needs of their patients. The education of a health professional is not complete unless they also understand that a family in Washington, D.C. earning twice the poverty level cannot find affordable housing. Second, community service should not be considered an “add on” to education, but a more central part of it. It should include time working in real community settings in addition to working in the hospital’s clinic. Third, while continuing education programs have traditionally focused on retaining technical skills, there should also be a focus on community outreach and understanding the social context in which care is provided. AHCs should take a broader perspective on continuing education. Professor Rosenbaum tells her students that the quality of health care is shaped not only by a slip of the knife, but also a slip of the tongue. A physician who is technically brilliant, but completely impervious to his or her surroundings, cannot practice good medicine.

Charles Cutler, M.D., Chief Medical Officer, American Association of Health Plans

Dr. Cutler spoke to the expectations of AHCs from the perspective of health plans. In preparation for this presentation, Dr. Cutler polled a number of health plan medical directors. They agreed that although the roles of research, education and patient care are expected from AHCs, the approaches needed within each role should vary from what is done today.

In terms of the research role, scientific advances are important, but health plans also need AHCs to take leadership in: 1) describing what makes care safe, effective and efficient; 2) developing processes of care that are more patient-centered, efficient, effective and timely, and defining how these processes can be implemented; 3) defining what are important and relevant measures of quality of care and practice; 4) identifying effective methods of health promotion and disease prevention; and 5) understanding what works and what doesn’t work in basic and applied clinical science to shorten the estimated 17 years it can take between a scientific advancement and implementation into practice.

In terms of the education role, health plans need AHCs to train health professionals in the skills and knowledge needed in practice, with training in using information technology to manage practices and support clinical decision-making. The education role should place greater emphasis on training people to work within the types of teams that should be used in practice and that can

improve efficiency and quality (as is done in the manufacturing and aviation industries, among others). As more patients are faced with managing chronic conditions, they need to be involved as partners in their own care, and health professionals need to be trained to communicate with and support patients in that way. Training in population-based care should train health professionals how to manage not only the patients they see in person, but also the patients that are not seen. Additionally, education programs for health professionals should include self-evaluation that supports reflective practice, lifelong learning and accountability.

In terms of the patient care role, AHCs should become role models for the attributes necessary for practice, including team care, use of measures to evaluate and improve performance, active management of relationships with community providers and interdisciplinary coordination.

To be included in health plan networks, AHCs will need to demonstrate performance in each role. Relative to patient care, AHCs should be able to demonstrate that they are centers of excellence based on performance measures, including costs. In terms of research, they should be able to demonstrate the conduct of cutting-edge research that is not available in other settings. In education, they should be able to demonstrate the availability of expanded training opportunities in areas such as primary care, in which community practices may be more effective. Finally, AHCs should seek partnerships with managed care, as appropriate. Some currently exist (e.g., Harvard Pilgrim Health Plan and Harvard University, or Group Health Cooperative of Puget Sound and the University of Washington). More partnerships should be pursued and developed.

During open discussion, a series of questions were raised, which can be grouped within two areas:

Larry Lewin raised the issue of competition in health care and erosion of the ability of cross-subsidies to support AHC functions. Brian Biles noted that prior research has shown that in past years, indemnity payers paid AHCs approximately 30 percent more per case (case mix adjusted) compared with community hospitals. In the mid-1990s, research suggested that managed care plans were only going to pay an additional 10 percent at best, introducing a major financial change for AHCs.² John Rowe suggested that health plans are “balanced” payers that help balance the budget for academic health centers. He noted that data developed by staff at the Medicare Payment Advisory Commission (MedPAC) suggest that as Medicare funding goes down for hospitals, health plans pay more. Medicare is both the largest payer and the most powerful discounter. In response, Charles Cutler suggested that if health care premiums are to be kept affordable, more tiered networks are likely. That may not be favorable for AHCs and may have unintended consequences for patients, but the market will need to resolve that issue.

Both Uwe Reinhardt and Jeff Goldsmith raised the issue of how to balance public and private responsibilities in supporting AHCs. Uwe Reinhardt commented that there is a perception that private payers are the “residual budget gap closer,” addressing the shortfalls from Medicare. To the extent that private payers believe they have to fulfill that role, then there is no constraint and the private payers are unable to discipline the market. If so, the public/private mix needs to be re-examined to determine each party’s responsibility, particularly for academic health centers.

Jeff Goldsmith stated his belief that there has been a flight from competition, both in health care delivery and health insurance during the mid- to late-1990s, and that in most metropolitan areas, provider cartels are negotiating with health insurance cartels. In the typical metropolitan area, 3 plans control 70 percent of the private insurance market. In his view, it is absurd to think there are market forces operating in this field. This raises the serious question of the government’s role.

Both Sara Rosenbaum and Charles Cutler noted the importance of paying AHCs for the outcomes they produce. However, Sara Rosenbaum commented that medicine is also driven by judgment, and financing should consider the “fungible” side of the product. Charles Cutler reinforced the concern about how to pay AHCs for their outcomes. This includes not only clinical outcomes, but also supporting the desired outputs from medical education. For example, if the needed skills include population and public health, those should not be incremental add-ons.

Several people commented on the types of activities desired from AHCs. Linda Aiken asked about the need for a broader perspective in education that includes health, knowledge management, economics and other disciplines. Larry Lewin asked about the potential for greater integration of public health and medicine, asking if the separation and isolation of those two areas made sense. Paul Ramsey raised the issue of workforce distribution, especially the serious maldistribution of the workforce in rural areas. As the need for multidisciplinary teams and information technology grows, he asked whether AHCs should play a greater role in addressing the needs of rural health. Claude Bennett commented on the apparent disconnect between what AHCs think they are doing and what they are doing. He noted that AHCs are at the cutting edge of research, yet may not adequately care for the more common health care needs of people or consider the impact of research on health care costs. Education attempts to provide students experiences in community settings, but comments raised in earlier discussions suggest that such efforts may not be as effective as desired.

Sara Rosenbaum suggested that there may be some similarities between legal and medical training in that legal training has developed techniques to teach students to be empathic and to be able to advocate for their clients. There is as much emphasis placed on the teaching of these skills as there is on teaching torts. This type of educational approach requires more than just rotating students through a community setting, whether that be a public defender’s office or a community health clinic. It requires that such skills be taught as a core part of the training experience.

Ralph Snyderman noted that there are many barriers to change, including payment, culture, and cross-subsidies that confound the operational understanding of AHCs and the need for more effective management. But the organizational complexity of the AHC itself is a problem. It is not clear how many have a single office or authority that is able to direct the organization as a whole. Half of its business takes place in an aggressive marketplace; half of its business is embedded in a university. Linda Aiken asked to what extent AHCs need to be integrated within a university in order to integrate across the disciplines and address social determinants of health, but Ralph Snyderman noted that getting closer to the university may provide certain benefits, but can also create even more complexity in decision-making.

John Porter noted that although it has been difficult to achieve and remains a work in progress, the federal government’s Government Performance and Results Act (GPRA) may offer one model for setting measurable objectives and assessing whether results are being attained.