National Symposium on Outcomes and Quality Assessment: State of the Art and Future Directions

The Pennsylvania Health Care Cost Containment Council was created in 1986 by the Pennsylvania legislature at a time of double digit health care inflation (1). Businesses were frustrated by the inability to manage health care costs. They were able to control other costs of doing business, but health care costs eluded traditional market cost control mechanisms. The axiom at the time, popularized by an executive at Hewlett Packard, was "We don't buy health care, we pay for it" (2). The strategy of some businesses to shift the costs to their work force put health care costs at the center of most collective bargaining disputes, especially in Pennsylvania.

Business and labor recognized that they had a common interest in controlling health care costs. They were also able to agree that information about the cost and the quality of health care would make them more informed purchasers and would enable them to identify value in health care-the best quality at the best price. Hospitals, physicians, and insurers were, at best, reluctant partners in this enterprise.

With the clear support of the business and labor communities, the Pennsylvania legislature passed a novel piece of legislation, the Pennsylvania Health Care Cost Containment Act. This legislation created an independent state agency which would collect data from hospitals and report to the public on the cost and quality of health care services. The objective of the legislation was to improve efficiency within the health care marketplace by providing information about the cost and quality of care and thereby, creating pressures to reduce the rate of growth in health care costs and improve access.

The underlying belief held by the law's supporters was that purchasers, equipped with information about the cost and quality of medical care, would stimulate providers to lower costs and improve quality. For that reason the legislation vested authority on the Council with purchasers. Sixteen of the 21 members represent the interests of those who are primarily responsible for paying for health care services. The other five members represent insurers, providers, and HMOs. Implicit in the Council's leadership structure, as originally formulated, was a recognition of the antagonism between the purchaser and provider communities. Ten years later, as a consequence of changes in the health care industry, it is safe to say that, in Pennsylvania at least, that antagonism has been reduced to constructive tension.

The law also permitted the Council to require the hospitals to install a severity measurement system developed by MediQuai Systems, Inc. The MedisGroups software (now evolved into Atlas^TM) was designed to assist hospitals in their internal quality assurance programs by providing them with an external database with which to compare their own performance. Hospitals are required to report the admission severity to the Council. This allows the Council to report observed and expected mortality rates.

During the next 10 years, the Council produced a number of ground breaking reports. The Hospital Effectiveness Reports (3) disclose observed and expected mortality rates for 57 illness categories at each of the general acute hospitals in Pennsylvania. The process of putting outcome information into the public domain has gradually led hospitals in Pennsylvania to accept the idea of public reporting of outcomes (4).

The Council continued to pursue groundbreaking approaches to public outcomes reporting. In 1992 the Council proposed reporting on physician-specific outcomes for coronary artery bypass graft surgery. Initially, the physician community voiced strenuous objections. However, several physicians persuaded their peers to work with the staff of the Council to develop a risk adjustment methodology which could account for differences in patient severity across institutions. The result has been a product and a process that has been widely praised by both providers, purchasers, and consumers (14).

As the Council approached its 10th anniversary, much had changed in the health care environment. The Council continued to be a leader in disseminating public information about the cost and outcomes of hospital services. Nevertheless, important trends were emerging that threatened to marginalize the efforts of the Council. Outcome reporting was becoming more widely accepted but there was also a desire to standardize outcomes measurement tools for psychiatric and rehabilitation care. Hospital inpatient services were no longer the sole focus of attention. With increased penetration of integrated care delivery systems and managed care organizations there was a greater need to include quality and outcomes assessments within the scope of outpatient services. In addition, information technology and outcomes assessment tools had improved significantly since the mid-1980s, creating an environment in which the ability to actively assess health system performance was expanded, adding both to the capabilities and challenges of processing information.

As part of an effort to chart a course into the next decade, the Council decided to sponsor a symposium. The symposium was organized with several goals. First, it was to showcase results of the Council's experience in making outcomes data publicly available. Second, it was to survey severity of illness and outcomes assessment methodologies available in both commercial and academic arenas for patients requiring services other than those provided by general acute care hospitals. Third, it was to introduce to the participants some of the possibilities and challenges offered by emerging information technologies and trends toward health system integration. An overriding goal of the symposium was to bring together the purchaser and provider communities within a common framework to learn about developments in outcomes technology and initiatives to manage heath care costs. In such a setting it was hoped that the two communities would become better informed about their common and competing goals and interests regarding health outcomes information.

The session about the effort by Hershey Foods, Inc., to use PHC4 clinical outcome data to help to build its own managed care network, perhaps best illustrates the intent of Act 89 (19). Labor and management cooperated to bring about .the change. Both the health insurer and the providers followed rather than led in the process. But since then, both the insurer and the provider have refocused their organizations on demonstrating and providing value to their customers. Hershey, however, is not a typical example of how the Council's data have been used. The examples provided by Vantage Health Care and Forbes Medical Systems may be more commonplace (15). In these examples it is the provider who has forcefully used the data for strategic planning and marketing, as well as helping to develop practice guidelines.

Even more typical are the many anecdotal stories that filter back to the Council. These stories belie the public position of some providers. Although sometimes publicly critical of the data or the underlying methodology used to assess patient risk, privately they are engaging in activities to improve the quality of the product they deliver and to lower the price using information at least partly derived from Council data.

This illustrates one of the recurring tensions expressed by Dr. Chassin (15) in the keynote address and by Dr. Goldfield (16) and others during the symposium sessions. Is outcomes data the first step in a quality improvement process, or is it to be used to publish report cards on providers with the intent of "punishing" poor providers? It is desirable to think that the purchasers can work collaboratively with providers to improve the process of care for the benefit of all. But it is the public measurement of performance in Pennsylvania through the Hospital Effectiveness Reports and the Coronary Artery Bypass Surgery Report, and the widespread awareness that these databases are publicly available, that has helped to stimulate providers to understand and use data to improve the process of care.

Quality assessment and outcomes reporting have changed dramatically over the last 10 years. Drs. Chassin (5) and Gaus (6) emphasize how health services research has been able to determine guidelines for effective clinical practices. Implementation of these guidelines has saved providers and payers millions of dollars. To identify effective clinical practice there must be a measurable outcome. Goodman and Green (7) point out that the inability to measure outcomes has lead to uncertainty about clinical practice. This uncertainty can be measured by the wide variation in certain clinical practice patterns from one area to another as measured by small area analysis. Understanding this variation is a first step toward improvements in the process of care. Dr. Patterson describes one effort to understand and act on this variation in Pennsylvania (8).

Several symposium speakers discussed the ingredients that are essential to improving the process of care. First is the ability to measure outcomes. Measurement of outcomes needs to adjust for severity of patient illness. Measurement of outcomes must have clinical validity. Lisa Iezzoni (17) emphasized that these words, risk and outcome, should be treated carefully.

Second, in order for the process improvements to take hold, providers must understand and accept the methodology used to measure performance. They must understand that measurement will lead to process improvement and not be used punitively. Speakers from Vantage Health Care discussed the use of Pennsylvania Health Care Cost Containment Council's data in close collaboration with the physician network to develop practice guidelines.

Third, without internal or external pressure to change practice patterns, providers, as a group, are not likely to take the lead. They need the leadership and the support of those responsible for paying for services, including business, government, and consumers. This is evidenced in the comments of several session speakers who describe purchaser initiatives in Hershey, PA (9), Cincinnati (10), Cleveland (11), Florida (12), and New England (13).

Fourth, improved information technology has enabled complex risk adjustment methodologies to be available at individual work stations in formats that are relatively easy to understand by unsophisticated computer users. Several speakers discussed applications of new technologies to the measurement of outcomes. Minnesota Blue Cross has installed an outcomes measurement system in each of its participating hospitals. This provides information which is used to develop a reimbursement methodology that rewards better than expected outcomes. Roger Boll, of EDS and Neil Marcuson, of Ameritech discussed issues pertaining to storage, retrieval and analysis of health care data. Important issues regarding ownership, access and confidentiality have yet to be fully resolved. Nevertheless, integrated data systems facilitated by continual improvement in telecommunication technology, will dramatically expand our ability to measure the performance of health care delivery.

Almost 10 years after its creation, the Pennsylvania Health Care Cost Containment Council can reasonably argue that the experiment in Pennsylvania is successful in reaching many of its goals. The public availability of information has prompted closer collaboration between purchasers (business and labor) and providers. Purchasers took the lead in bringing outcome information into the public domain and into the deliberations about health care purchasing decisions. Hospitals and physicians have responded positively, albeit not always enthusiastically. They are using information, both public and internal, to identify areas for improvement and they are developing strategies to address these areas.

There is now a much greater understanding of each other's concerns and problems. This opens the possibility of creating collaborative strategies to continually improving quality of care and access while maintaining control over costs. The Symposium highlighted the challenges that remain to be solved, not only by the Council, but by all those in both the public and private sectors who are working toward improving the efficiency of the health care delivery system by measuring outcomes and improving processes. It also highlighted the dynamic and creative tension between purchasers, providers and the public, to use information as a tool to bring out a more efficient and equitable health care system.