Collaboration Among Competing Managed Care Organizations for Quality Improvement (1999)

Catherine Borbas, Ph.D.

The Health Care Education Research Foundation (HERF) is a nonprofit organization founded in 1982 as part of a hospital association. It became independent in 1986. The board comprises hospital administrators, physicians, health plans, purchasers, health department representatives, and other community organizations. It has projects that involve a considerable proportion of Minnesota's health plans, hospital systems, and provider organizations. Business coalitions and the health department also participate. The HERF mission continues to be to try to raise the quality of health care for the community. Although today's customers are increasingly clinics, clinic systems, hospitals, businesses, and health plans, the focus continues to be on the community, patients, and enrollees.

HERF has developed the Minnesota Clinical Comparison and Assessment Program (MCCAP). The products and services of MCCAP are based on community- and condition-specific evaluations. Since 1987 these projects have been focused on cholecystectomy, hip replacement, management of labor and delivery by Cesarean section, treatment of breast cancer, and acute myocardial infarction. The study involves large numbers of patients at a number of sites. All the data can be sorted by health plan, by clinic, and by physician.

MCCAP is not involved in the routine monitoring of care. The organization becomes involved when members of the Minnesota health care community agree that they have a common problem and would like to modify clinical practice or when the community would like comparative information to evaluate and compare the care provided by medical sites. Using practice guidelines as a blueprint for measurement and collaboration, HERF gathers data from sites using standardized data collection methods and makes the information widely available. The evaluated groups then have the opportunity to evaluate the results and any recommendations that have been offered. HERF then collects another round of data to see if there has been any change in clinical behavior and patient care.

This is difficult, despite the level of practice consolidation in collecting standardized data in Minnesota, because medical records do not reflect that consolidation.

Early on, HERF recognized that hospitals and the health plans were good at making physicians and providers aware of changes. However, HERF found that health care organizations and health plans had a great deal of trouble with the implementation of change across systems. For example, there is considerable acceptance of guidelines for the treatment of acute myocardial infarction by cardiologists, but there are many problems with fully integrating these guidelines into practice.

To overcome these problems, HERF identified opinion leaders in given areas by surveying physicians about to whom they turn with questions on the study topics. These physicians are recruited for participation in "spreading the word." HERF's opinion leaders are topic and site specific. Additionally, HERF recruits informal and formal administrative champions because system and process delays cause many difficulties in implementing change.

Many gaps between knowledge and practice can be explained by clinicians' previous negative experiences or personal attitudes and judgments. For example, in explaining why β-blockers may not be used as widely as is indicated, one cardiologist said that "in giving [a β-blocker] a physician cannot say he has saved a life, but if the patient has a stroke [as a result] he can say he caused it." It is very difficult for a group at a health plan level or even at a hospital department level to deal with these underlying attitudes and fears that can influence some clinical priorities on a day-to-day basis.

The value of HERF has been providing a neutral forum for community collaboration and economies of scale. If plans and providers choose to collaborated through a consortium like HERF, they will have access to the comparative standardized data increasingly requested by purchasers of care, some shared funding, and increased fund-raising capabilities. If they do not choose to engage in this type of activity and stay independent in quality measurement efforts, they will have more control and perceive that they will have a competitive advantage.

Donald M. Steinwachs, Ph.D.

The Employers' Managed Health Care Association (MHCA) comprises Fortune 250 companies and their personnel benefits staffs. After hearing Dr. Paul Ellwood speak about outcomes management at a conference in 1989 or 1990, several members became intrigued by the idea of trying to capture the patient's experience in terms of outcomes and trying to look at health status as an indicator of quality of care.

This subgroup, initially comprising 15 members of MHCA, came together internally to form an outcomes consortium. They recognized that to accomplish anything meaningful they would need to bring their managed care contractors into the effort. About 18 managed care partners joined the consortium. Over time, a few of these managed care organizations have merged with others, such that the consortium now has representation from many of the key actors in managed care in the country. Whenever they meet they are cautious about any potential for antitrust activity.

Although in the late 1980s, employers had focused on ways to cut their costs and to shift costs to employees, in 1991, the MHCA Outcomes Consortium began to examine ways to improve the health of its employees and to improve the quality of care provided by health plans through which they purchase care.

They saw this drive as a feasibility experiment to examine (1) whether it is possible to use the patient's experience (as captured through surveys and other instruments, such as health status measurement) to look at quality, and (2) if it is feasible, whether it is useful. Can health plans use this information to change care? Can the information be translated into clinical indicators of outcome as well as health status and functional status indicators? The initial feasibility test included coronary angiography and adult asthma and sought ways to capture changes in functional status from a baseline and follow-up assessment.

This partnership existed only because the employers brought the managed care organizations to the table and the plans participated to satisfy them. This project resulted in benefits for both the employers and the plans. The plans gained economies of scale in instrument development and in the pooling of data and basic analysis and had the opportunity to measure their progress against the progress of other plans. This provided a way to hold the managed care plans accountable for quality. The employers were able to direct plan attention toward the outcomes that most concerned them. Assessments always included measures such as numbers of work days lost, missed activities, and other measures that represent employer concerns.

Such pooling of data, even when employers did not have much control or input into how the information was used by plans, has been a useful process. It kept the plans' attention on quality. For the plans, this effort was useful in helping them build quality improvement activities. At least one plan in this effort uses the MHCA process to satisfy some NCQA accreditation requirements.

Thomas J Davies, J.D., M.P.A.

Several collaborative efforts on the West Coast offer examples of strategies, philosophies, and approaches that make them successful, as well as some of the pitfalls and barriers to be avoided. The primary example is the California Coop-

erative Healthcare Reporting Initiative (CCHRI), which among other activities provides an annual Health Employer Data and Information Set (HEDIS) report card. CCHRI is designed to collect, verify and report HEDIS data. This collaboration, initiated by the Pacific Business Group on Health (PBGH) has been very successful despite its size, complexity and $2 million annual cost.

In California, there is a high degree of overlap of practitioners in more than 20 health maintenance organization (HMO) networks. This overlap reaches 70 to 80 percent in some cases. This poses a practical problem because each HMO would need to send its own reviewer into each practice site to examine medical charts unless the data can be generated by the HMO's administrative systems, which few of them can do. This avoids the chaos and disruption that would be caused by having so many reviewers descend on so many practices at about the same time.

PBGH initiated the collaborative with 22 HMOs and a broad range of provider organizations. Its purpose was to devise a methodology for independently collecting and verifying HEDIS data on an annual basis on behalf of all participating HMOs and reporting to PBGH's 35 member companies and the public. Without the coalition, it is unlikely that the HMOs would not have launched this collaborative.

There have been numerous difficulties in the technical details, timing, and timeliness of the data. What makes this collaborative successful, however, is (1) the perceived value of the product to the HMOs and physicians; (2) the economies and efficiencies derived from the collaborative effort, such as a single independent party doing the chart reviews; (3) the explicit agreement by everyone about the common goals; and (4) a purchaser-driven component, which is probably the most important ingredient. The purchasers in this coalition are not the entire market, but they are very influential participants in the market and they are trying to drive the market in a direction that emphasizes quality of care.

The philosophical and strategic elements that go into making a collaboration successful are the following: First, collaborate to establish objectives, standards, measures, metrics, approaches, and goals. These are the critically important activities, and they are concentrated at the beginning of the project. Second, each party remains responsible for cooperating in implementation, according to the standards and the plan developed with the support and assistance of the vendor doing the on-site chart review. Although considerable work must be done by the organizations individually, there is a need for cooperation within this framework. Finally, these efforts are meant to support marketplace competition that is based on quality. So, the PBGH approach is to collaborate on the planning, cooperate on the implementation, and compete on the basis of quality and performance. PBGH is committed to making quality count in the process of purchasing health care and to driving a quality agenda in the marketplace.

The major pitfall or barrier to the success of this project that has yet to be overcome is the tendency for the managed care plans to have deep-seated competitive instincts. Some are inclined to function independently, using data in a

proprietary fashion, that does not serve the interest of employers, purchasers, consumers, or physicians.

Rachel M. Rowe, R.N., M.S.

The Foundation for Healthy Communities was founded by the New Hampshire Hospital Association in 1996. The reason it has been successful is that it offers an opportunity for employers, consumers, plans, and hospitals to come together to focus on the care of patients. The foundation is a statewide effort including the seven health plans that do business in New Hampshire. None of these health plans owns hospitals. The New Hampshire legislature recently outlawed exclusive contracts, so provider networks are established primarily on a contractual basis, and the networks overlap.

The New Hampshire Hospital Association established the foundation because the association recognized that there was an opportunity to improve care. However, it was clear that it would be difficult to do this as a trade association. The effort needed to extend beyond hospitals, and the foundation was created as a neutral space where there would be more diverse participation and all parties could come together with an equal voice and influence.

Fortunately, the hospitals have had a long track record of collaborating with one another on a number of quality improvement initiatives. The health care plans also had some experience. In 1991, three of the state's major health plans had initiated and carried out plans to finance the cost of immunizations for children, regardless of their insurance coverage.

With the assistance of Paul Cleary and his colleagues at the Picker Institute, the foundation developed a statewide (now New England-wide) survey process using diagnosis-specific patient feedback and functional status measures to be used across the continuum of care. These measures focused first on two categories of patients: those with myocardial infarction and postsurgical patients. It used a three step survey process to look at aspects of care from the time of hospitalization on. Another component was added to each of these surveys to measure functional status. In the last year, maternal and newborn care have been added. Every hospital in New Hampshire, along with several in Maine, Vermont, and Massachusetts, participated in this effort. During 1997, 16,000 mothers are expected to report on their experiences with prenatal, childbirth, and postpartum care.

Because of the scope of this work, six of seven plans being studied not only agreed to help finance the project, they also agreed to be active in the foundation's work. They were interested in participating and being able to share jointly developed information because of concern about the care and coverage of maternal and newborn care expressed in the political and media arenas. The foundation afforded them an opportunity, as well as the credibility, to participate in a

study of the care of these patients. The results of the initial studies have shown that mothers in managed care plans and mothers throughout the state show no significant differences in their perceptions of the quality of the care they receive. The problems identified appear to be everywhere.

Over the last year, these projects led plans and providers to collaborate on numerous other ventures. Providers were concerned most with who received the generated information and how it might be used internally for quality improvement efforts. With the assistance of Paul Batalden and his colleagues, the foundation is working now to give physicians clinically focused feedback about patients' experiences of care so they can tailor their processes of care. The foundation is midway through this process.

The foundation has been successful with these efforts because of its caution in identifying and avoiding potential barriers. External competition, as well as internal competition, was a concern when the foundation initially looked for funding from the plans. External competition among plans and internal competition for improvement and research funds means that any effort must compete for limited resources. When the foundation asked for funding from the plans, it faced resistance from the administrations because of external competition and from the providers because of internal competition.

Like competition, ownership also presents two potential barriers. First, some of the participating health plans have corporate offices outside of New Hampshire, and they have their own agendas regarding quality improvement. Currently, they are more interested in a national focus than in working with New Hampshire efforts alone. Second, changes in the ownership of plans can make the continuity of these efforts very difficult.

Tim Size, M.B.A.

The Wisconsin Rural Zones of Collaboration Initiative grew out of a series of open-ended discussions among HMOs and rural providers in the central and southern parts of the state initiated by the Rural Wisconsin Health Cooperative (RWHC) and the Community Physicians Network. The context was (1) the already high market share held by HMOs in rural counties, (2) the fact that rural providers typically contract with multiple HMOs, and (3) a desire by RWHC that rural health not become primarily defined by competition among regional HMOs.

These exploratory talks led to the development of a shared vision: that rural communities have a strong interest in cooperation among HMOs and other payers on issues directly affecting local care and public health and that some agreement among regional plans with rural providers will reduce duplicative and fragmented interventions within and among rural communities.

Desirable outcomes were identified in three issue clusters: (1) inefficiencies related to multiple HMOs working with the same rural providers, (2) public health, and (3) education and research. Fortunately, just as funding was being sought to initiate activities in the first area of HMO-provider relations, the Health Services Resource Administration's Rural Network Development grant program was announced. RWHC along with the Wisconsin Rural Zones participants wrote and successfully competed for a three-year grant at full funding ($200,000 per year). The focus of the grant is as follows:

1. Increase the effectiveness and utilization of RWHC's regional credentialing service for multiple practitioners, hospitals, health plans, and direct purchasers and more effectively coordinate with neighboring regional credentialing services.
2. Develop and implement a model for providers, plans, and direct purchasers to collaborate on data collection, site visits, and other administrative audits required of practitioners in a rural network.
3. As a rural network, centralize health plan customer satisfaction surveys to achieve a rural database that will be large enough to provide meaningful information for improvements in local areas,
4. As a network, create a common clinical practice guideline review and adoption process and demonstrate rural practitioners' ability to attain the desired clinical practice outcomes.
5. Identify and implement clinical quality management projects in which multiple practitioners, hospitals, health plans, and direct purchasers share a common interest at the local level and develop uniform performance objectives and outcomes measures to improve the health status of the populations served by multiple health plans. (The RWHC Quality Indicators Program is a newly accepted performance measurement system for the Joint Commission on Accreditation of Healthcare Organizations' ORYX Initiative.)

The RWHC has found that some of the following principles enhance the likelihood of successful collaborative behavior:

1. Respect the need to affect one's own future. The preference for autonomy needs to be respected through the promotion of collaborative solutions that enhance the local delivery of health care and the health of the community.
2. Involve the community in the planning process. The planning is interactive, with the plan for the joint enterprise being the result of and feeding into the plans of local and regional participants.
3. Assure all participants know they are needed. All participants must know that they are needed for the success of the joint enterprise.
4. Share the big picture. Participants need to know where the joint enterprise is headed.

5. Agree on methods of accountability up front. Participants must always know up front what the rules and what is expected of them.
6. Ensure a fair system of arbitration is available. A clear, nonthreatening arbitration mechanism in case of contractual or other disputes should be agreed to before disputes arise.
7. Design approach where participation makes sense. Organizations may start participating in a group to explore a group's potential; they remain in a group only if they perceive that they are receiving a good return on their investment of time and money.
8. Make yourself a partner who can be trusted. Develop a relationship based primarily on mutual trust so that the collaborative effort is not limited to the minimum performance inherent in written agreements.

9. A number of factors are expected to be helpful:

   • Perceived win/win/win for insurers/providers/public regarding the:
   • • demonstration of quality possible in rural areas;
     • increased efficiency in the flow and use of data; and
     • reduction of provider and plan ''hassles."
   • • Cooperation is not new in Wisconsin.
     • The public preference for balance between cooperation and competition.
     • The initiative builds on existing relationships and efforts.
     • Rural providers are needed by expanding HMOs.
     • Rural providers have a "home court" advantage.
     • Rural providers provide a neutral forum.

   Still other factors might be problematic:

   • The complexity of multiple relationships.
   • The inherent difficulty of the quality related goals.
   • The participants may be unable to "leave their guns at door."
   • Rural providers may become or remain fragmented.

   Barriers to Collaboration

   As of May 1, 1998, RWHC had had six months of very positive experience with implementing the grant-funded goals stated above. Several barriers have been identified and are noted below, along with strategies for addressing them.

   • Lack of consensus regarding delegation of tasks relevant to NCQA among health plans. Currently, plans are not willing to commit to full-scale delegation for their involvement with rural practitioners.

• Strategy: Collaboration can be viewed at different levels, with increasing benefit accruing to rural practitioners as the collaboration increases.
• The vertical structures of large organizations make communications difficult and diffuse responsibility for effective decisionmaking. Middle managers are reluctant to give firm responses, and frontline staff do not receive the history or the actual documents related to a project.
  
  

  Strategy: Use detailed agendas and meeting minutes, and disperse them widely; include staff members from multiple levels within the organization whenever possible; publicize significant accomplishments.

• Frequent turnover of participants.
  
  

  Strategy: Prepare a plan for orientation of new participants. Determine the information needed to bring new staff up to date on projects. To the extent that systems and procedures are in place and documented, personnel changes should be less disruptive.

• Different forces drive each organization to begin using clinical guidelines and outcomes such as the requirements of NCQA and JCAHO and bottom-line finances.
  
  

  Strategy: Use pilot programs to demonstrate how one project can be used to satisfy various needs.

• The existence of different levels of knowledge regarding clinical guidelines and outcomes research.
  
  

  Strategy: Provide educational opportunities for participants as well as encourage participants to learn from one another.

• Statistics: sample size and ceiling effects.
  
  

  Strategy: Projects must be carefully selected to ensure adequate sample sizes for analysis of characteristically smaller rural data sets. Projects must be carefully selected by taking into account the possible influences of ceiling effects. For example, as an outcome measure, "smoking cessation" may be a poor choice if most of those who will stop smoking have already done so.