Theory and Reality of Value-Based Purchasing: Lessons from the Pioneers

In contrast to all the research conducted on health care providers, managed care organizations, and health care systems, there has been very little systematic research on employers as purchasers of health care. To fill this need for information, the Agency for Health Care Policy and Research (AHCPR) has begun a research initiative focused on employers as purchasers of health care.

This report summary provides early lessons on the scope, patterns, barriers, and possibilities of value-based purchasing. It is drawn from in-depth discussions with nine employers and coalitions as well as written materials from many other organizations.

The 55-page report can be obtained from the AHCPR Publications Clearinghouse by calling toll-free 800-358-9295 and requesting AHCPR Publication No. 98-0004.


The Theory of Value-Based Purchasing
The Reality of Value-Based Purchasing
Measurement Activities of the Pioneers
   What and Who Pioneers Are Evaluating
   The Methodologies
What Are the Obstacles?
   A Sense of Complacency
   Lack of Dissemination Mechanisms
   Inadequate Staffing
   Credibility of the Information
   Information Not Being Used to Drive Change
   Employers Mainly Concerned With Cost Reductions
Research Agenda
Project Staff
About the Authors


Employers are central figures in the current market-based health system. In their push to contain health care expenses, employers have been a driving force behind recent dramatic market changes: The move from indemnity coverage to managed care, the flurry of consolidations and mergers of health plans, the emergence of new risk-sharing and contracting patterns, and the slowdown in the rate of premium increase. More recently, some of the more forward-thinking employers have begun to use their market power as a force to promote quality and value of health care services as well.

However, despite the undisputed role of employers in trying to shape the cost and content of health care, very little is known about why, how, or how well they exercise these powers, and we know even less about the impact of these purchaser activities on employees, employers themselves, and the broader health care market or community. We have a generation of research examining the behavior of physicians and hospitals, some more recent work on managed care organizations and health care systems, but very little systematic work on employers as purchasers of health care.

To begin to fill this gap, the Agency for Health Care Policy and Research (AHCPR) has begun a series of research initiatives focused on employers. In this report, Jack Meyer, Ph.D., Lise Rybowski, and Rena Eichler, Ph.D., present findings from one of these projects: an examination of the state of the art in value-based purchasing. The report, drawn from in-depth discussions with nine employers and coalitions and written materials from many others, provides early lessons on the scope, patterns, barriers, and possibilities of value-based purchasing. They find that most employers are focusing their attention exclusively on costs. Another group of "dabblers" has begun to collect some quality information from plans and providers but has not used the information to influence or mold purchasing decisions. Finally, a few employers and coalitions—the pioneers—are acting in a bold and innovative fashion to implement the principles of value-based purchasing.

The pioneers:

To take a strong role as a value-based purchaser, employers and coalitions must surmount many obstacles, including complacency, inadequate dissemination mechanisms, staffing shortages, data credibility issues, but the nine case studies provide evidence that these barriers can be overcome. Finally, the authors spell out a valuable list of research questions for the future—questions that will be very helpful to the Agency and the research community as a whole as we continue our study of the role of purchasers in shaping the way health care is financed and delivered.

Irene Fraser, Ph.D.
Director, Center for Organization and Delivery Studies
Agency for Health Care Policy and Research

Return to Contents


The purpose of this report is to describe some of the most promising examples of private business initiatives to build quality considerations into health care purchasing. In this overview section, we analyze the state-of-the-art of employer-sponsored quality measurement initiatives in the context of the strategy of value-based purchasing. In addition to laying out the activities and accomplishments of several innovative purchasers, the report elaborates on the obstacles keeping employers—even the innovators—from making greater progress in this arena. The report concludes with a number of suggestions for further research that would facilitate employer efforts to incorporate quality information into their health care transactions. Table 1 summarizes the activities of nine particularly innovative companies and business coalitions.

New Directions for Policy conducted in-depth telephone interviews with nine employer purchasers in the United States. We also collected and reviewed written materials describing the innovative activities. Five of the purchasers are business coalitions. Four are large corporations. We began by reviewing the activities of roughly 20 purchasers and selected the ones profiled in this report based on the sophistication and depth of their initiatives. Thus, this is a review of "best practices" from around the Nation rather than a "random sample" of employers buying health care.

This report begins by reviewing the theory of value-based purchasing and discussing the extent to which the reality of quality initiatives in today's marketplace matches or approaches this theory. The remainder of the report focuses on what will be required to move this reality closer to the ideal.

Return to Contents

The Theory of Value-Based Purchasing

The concept of value-based health care purchasing is that buyers should hold providers of health care accountable for both cost and quality of care. Value-based purchasing brings together information on the quality of health care, including patient outcomes and health status, with data on the dollar outlays going towards health. It focuses on managing the use of the health care system to reduce inappropriate care and to identify and reward the best-performing providers. This strategy can be contrasted with more limited efforts to negotiate price discounts, which reduce costs but do little to ensure that quality of care is improved.

The key elements of value-based purchasing include:

In a system based on value-based purchasing, employers and other purchasers gather and analyze information on the costs and quality of various competing providers and health plans. They contract selectively with plans or provider organizations based on demonstrated performance, or at least proposed approaches for improving performance. Ideally, quality information becomes a factor in the setting of plan prices, and employee contributions vary with each plan's "score," which reflects a combination of quality and cost indicators. In this manner, the best performing plans and providers are rewarded with greater volume of enrollees or patients.

Return to Contents

The Reality of Value-Based Purchasing

Our mapping of the terrain of value-based purchasing leads to the following conclusions:

Return to Contents

Measurement Activities of the Pioneers

What aspects of quality of care are the innovative employer purchasers measuring? This section examines what entities employer purchasers actually measure and the methodologies they employ to gauge quality of care. The pioneering groups whose activities are described in the report are the Dallas-Fort Worth Business Group on Health, Chicago Business Group on Health, Gateway Purchasing Association, General Motors, Digital Equipment Corporation, GTE Corporation, Pacific Business Group on Health, Iowa's Community Health Purchasing Corporation and Pacific Bell.

What and Who Pioneers Are Evaluating

The Performance of Health Plans

Most of the employers and coalitions that are attempting to measure quality are focusing on the performance of health plans. Of the employer purchasers featured in this study, the Chicago Business Group on Health, Digital, GTE, Gateway Purchasing Association, General Motors, and the Pacific Business Group on Health are evaluating the performance of health plans (see Table 1). This is, of course, one step removed from those who actually deliver the care: physicians, hospitals, and other health care providers.

Yet, the focus on plans has some important advantages. First, most employers are contracting with health plans, not directly with providers. This may change in the future as employers try to reduce costs associated with the "middlemen" between them and providers, and reap a greater share of the savings from new efficiencies in the delivery system (which is what employers in the Twin Cities are currently trying to do). But at the present time, the focus on plans is a practical response to the current environment.

Second, the focus on plans allows employers to aggregate the performance of many individual providers into statistically meaningful information. While much progress has been made, the science to measure the quality of individual providers is relatively undeveloped at this time. Finally, as employers become increasingly interested in obtaining information on health status and the prevalence of disease at the level of their populations or the community-at-large, large health plans become the logical collection point and repository for the data.

When assessing health plans, employers tend to focus on both administrative and clinical quality, where administrative quality includes such issues as the responsiveness of member services, waiting times in physician offices, and other factors important to enrollees. The clinical quality of health plans, on the other hand, is much harder to define. To the extent that they can get the information, most employers focus their attention on utilization and preventive care measures (i.e., process measures) that are relatively easy to understand, but not necessarily reliable indicators of quality. Most report cards, for example, contain very little clinical information. In contrast, the "pioneers" are trying to collect information on the outcomes of care, and identify the procedures and providers that are achieving superior results.

The Performance of Providers

Some employer purchasers are looking at the performance of providers. In most cases, purchasers focus on the providers because that is where the contractual relationship lies, i.e., through a preferred provider organization (PPO) or similar intermediary, the employers purchase care directly from physicians and hospitals. Iowa's Community Health Purchasing Corporation is taking bids from competing integrated card systems. Employers are tying health premium contribution to "benchmark" care systems, creating incentives for employees to select them. The Dallas-Ft. Worth Business Group on Health (DFWBGH), for example, is collecting information on both office-based and inpatient care for pregnancy and delivery. The group also plans to evaluate provider performance in four other disease group categories (see Table 1). In addition to a major effort to assess plan performance, the Chicago Business Group on Health (CBGH) is also trying to evaluate individual providers. CBGH's PPO, called EPIQual, uses some quality measures to screen hospitals for inclusion in a network available to member companies. The coalition also worked with hospitals to develop a critical pathway for coronary artery bypass graft (CABG) surgery.

There are a few purchasing organizations that are evaluating provider quality despite the fact that their contractual relationships are with health plans. Given the tremendous overlap of provider networks in most markets (i.e., most providers belong to several plans), it is very difficult, if not impossible, to use information at the plan level to distinguish differences in clinical quality. These groups recognize the need to have information about the quality of individual practitioners and providers organizations, as that is the point at which care is delivered to patients.

The Pacific Business Group on Health (PBGH), which contracts with major California health maintenance organization (HMOs) on behalf of nearly 20 large employers, is engaged in several activities to assess provider quality, including patient surveys and data-intensive analyses of claims and medical records. For instance, in partnership with the California Office of Statewide Planning and Development, PBGH is developing the California CABG Mortality Reporting System. This system is collecting and reporting standardized, hospital-level mortality data for all hospitals in the State that perform coronary artery bypass grafts. General Motors, which has to date focused on its health plans, is also exploring ways to use its extensive claims-based database to derive HEDIS (the Health Plan-Employer Data and Information Set)-like quality indicators for providers in its PPO networks.

Return to Contents

The Methodologies

As in any evolutionary process, there are basically two factors separating the pioneers in quality measurement from those merely dabbling in this arena: the sophistication of the tools they use and their ability and willingness to absorb and apply the information they receive.

How Purchasers Get Information

In general, purchasers collecting information on quality rely on the following tools, or methodologies:

1. Enrollee or Employee Surveys

While surveys usually focus on the question of "satisfaction" with care or service, they are beginning to incorporate questions that elicit reports, rather than just rankings, about care. For instance, in addition to asking a patient how satisfied he or she was with waiting times, the survey would ask how long the wait was (e.g., 10 minutes, half-hour).

The more sophisticated survey tools also gather information about health-related behaviors and health status. Knowledge about behaviors, particularly risky behaviors such as smoking, provide health plans and employers with useful information for designing educational and behavior modification programs. Data on health status are also useful for this purpose; in addition, these data are critical factors in adjusting survey results to be comparable among competing plans. Since poor health status (for example, chronic conditions such as asthma) is known to correlate with lower satisfaction with care, it is critical to adjust scores to reflect the health of respondents; otherwise, health plans may be unfairly penalized for serving a relatively "sicker" population.

While there are a number of organizations, including many consulting firms, that develop and conduct enrollee/employee surveys, three specific instruments are currently drawing the attention of sophisticated purchasers: the Consumer Health Value Survey, the National Committee for Quality Assurance (NCQA) enrollee satisfaction survey, and the Consumer Assessment of Health Plans Study (CAHPS®) survey. All three of these surveys offer a standardized set of questions and analytical tools so that health plan results can be fairly compared within and across markets, and can be assessed over time.

The Consumer Health Value Survey—which has been used by a number of sophisticated purchasers, including Xerox, Digital, and GTE—was originally developed by a team at the New England Medical Center's Health Institute under the direction of Harris Allen, Ph.D. (the instrument is available through Dr. Allen, who is now at Coopers & Lybrand's Boston office). The NCQA survey was developed in an effort to provide employers with comparable patient-derived information that would complement the health plan-derived measures in HEDIS. Gateway Purchasing Association in St. Louis, working with the Missouri Consolidated Plan representing State employees and over 300 local public purchasers, is using the HEDIS 3.0 Member Satisfaction Survey in 1997 as part of an effort to evaluate the performance of plans.

Funded by the AHCPR, the Consumer Assessment of Health Plans Study (CAHPS®) is a national initiative to develop a standardized survey instrument that can capture the experiences and perspectives of consumers using the health care system at any level. Underway since early 1996, it is a joint project funded as a cooperative agreement between AHCPR and three consortia led by RAND, Harvard Medical School, and Research Triangle Institute. These grantees are producing questionnaires, a detailed manual on how to implement the survey instrument and analyze the results, and consumer reporting formats (print and electronic), all of which will be released into the public domain. The official public release of the instrument took place in the summer of 1997.

Several employers, business coalitions, and Medicaid programs are planning to pilot test the CAHPS® survey in 1997. The survey will be administered by independent venders, who will be able to receive some technical assistance from Westat, which has a contract under the CAHPS® grant to provide this service. In addition, the developers of CAHPS® are currently working with the NCQA to develop an instrument that combines their survey with the one now required as part of HEDIS 3.0. This blended instrument will be available in 1998, and is expected to become the new requirement for health plans reporting their HEDIS results. The Health Care Financing Administration (HCFA) will require the use of a Medicare version of CAHPS® for plans offering managed care services to Medicare beneficiaries.

According to its developers, the CAHPS® instrument offers several improvements over the existing array of survey tools. First, the researchers designed the questionnaire to be relevant across a variety of delivery systems and populations. In contrast, most surveys address specific insurance circumstances (e.g., enrollment in managed care plans) and demographic groups (e.g., elderly, employed adults). The survey accomplishes this latter goal by presenting core survey questions that can be supplemented by items for specific populations, such as Medicare beneficiaries, Medicare risk plan enrollees, Medicaid recipients, children, and people with chronic conditions or disabilities.

For instance, in addition to the standard set of questions, consumers who have indicated a medical condition can be asked about their particular experiences with the health care system, with an emphasis on the coordination and continuity of care. Among the questions asked in this situation would be items regarding:

This structure allows the administrators of the survey to collect a set of items that will be comparable across all users, as well as any items that meet their unique needs. It also enables surveyors to address population groups that have not been surveyed consistently in the past.

The CAHPS® survey also takes a different perspective than the surveys commonly in use today. Rather than focus on what purchasers want to ask, this instrument emphasizes consumers: what they care about and what they can report reliably. Thus, it does not ask consumers to report on the technical quality of care they receive, which can be more reliably obtained from other sources, but on their access to care and their interactions with providers. However, the CAHPS® researchers are working toward adding items that will make the survey more relevant to purchasers over time.

2. HEDIS (the Health Plan-Employer Data and Information Set)

HEDIS is a set of standardized measures developed by the National Committee for Quality Assurance (NCQA) to facilitate employer efforts to assess health plan performance. The wide availability of these indicators has enabled purchasers to receive comparable sets of quality-oriented data from multiple plans. To date, the HEDIS measures have been criticized for emphasizing preventive and process measures, such as mammography screening rates, rather than outcomes of care. They have also focused primarily on the health issues relevant to the employed population.

In the winter of 1998, the NCQA will be releasing the latest version of HEDIS (known as HEDIS 3.0), which was designed to address many of the criticisms that have been raised. In addition to posing questions relevant to employers, HEDIS 3.0 starts to address the concerns of Medicaid and Medicare as purchasers of health plan services. It also contains some indicators of outcomes. HEDIS is intended to evolve over time to meet the changing needs of purchasers for information.

While there are probably several hundred employers and coalitions (plus consultants on behalf of employers) requesting HEDIS results from health plans, few are actually looking at the results and using the data in their purchasing strategies. To be fair, this reluctance to act on the data is in large part due to the poor quality of the information that is received; many health plans are struggling to provide all of the measures, and some fail to produce any data. Moreover, when employers have audited the data, they invariably find wide discrepancies between the reported and the audited results. Until recently, it was also difficult for a typical employer to translate the data-intensive HEDIS results from multiple plans into usable information. However, the NCQA has created a database called Quality Compass in order to collect HEDIS data from plans across the country and produce user-friendly reports for employers.

That said, the pioneers are encouraging plans to submit "cleaner" and more complete data, in some cases by providing financial incentives to do so. Perhaps more importantly, they are beginning to use the results to determine whom to contract with, how to price the plans for employees, and what to communicate to employees.

3. FACCT Measures

Over the past year, the Portland, Oregon-based Foundation for Accountability (FACCT) has been identifying and organizing new and existing measures that would allow purchasers and consumers to assess quality of care for specific conditions and populations. One of FACCT's primary goals is to bring together sets of measures that can be used to evaluate the delivery of care regardless of organizational structure. Unlike HEDIS, the FACCT measures should be relevant regardless of the type of payer involved or the level of provider integration.

The initial priorities for FACCT measures are breast cancer, depressive disorders, asthma, low back pain, and diabetes. While no purchaser is currently using these measures, of which only a few have been published so far, several of the pioneers interviewed for this study explicitly indicated their interest in participating in pilot studies of this tool over the next year or so.

There are a number of tools currently being used to assess provider quality, although most of them focus on utilization rates, i.e., the frequency with which a provider performs a given procedure, such as a C-section or bypass surgery. Some purchasers are looking more closely at severity-adjusted outcomes of care, such as rates for mortality, infections, complications, and readmissions, which is much more expensive, complicated, and time-consuming to do, but results in information that is more useful for both the providers and the purchasers. The Dallas-Fort Worth Business Group on Health, for instance, is currently collecting data for obstetric care, including such specific outcomes as low birth weight, unplanned neonatal readmission, and uterine rupture rate.

Sophisticated purchasers are paying attention to outpatient care as well, both by collecting data from physician offices and by conducting detailed surveys of patients who have visited the doctors or hospitals. However, this is rarely a task that purchasers take on by themselves. While many employers use the services of consulting firms with technical expertise in this area, they also often need to work closely with the providers themselves in order to ensure that they are getting the "right" information.

How Purchasers Use the Information

There is little question that what separates the pioneers from the dabblers is their willingness to put the information they receive to use, which creates a meaningful incentive for the plans and providers to provide credible data. These purchasers are taking advantage of the leverage they have, particularly with the plans, to encourage their participation in quality measurement and improvement initiatives. GTE, for example, recognizes the top 15 percent of its 140 plans by publicly designating them "Exceptional Quality" plans. General Motors is designating certain plans as "benchmark" plans based on a combination of HEDIS information, NCQA accreditation, and site visits. These companies price these plans more favorably, creating an incentive for employees to enroll in plans with the best performance ratings relative to both quality and costs.

Other purchasers use financial incentives for health plans to provide data; the Gateway Purchasing Association (GPA) in St. Louis withholds 4 percent of premiums from the health plans, which can only receive full payment if they provide the quality information requested by the coalition. In 1996, payment of the premium pool was linked to participation in a satisfaction survey and the provision of 20 HEDIS indicators. Over time, sophisticated purchasers like the GPA want to link payment of the premiums to improvement in performance, not simply the delivery of performance data.

Finally, a number of purchasers on the forefront of using quality information are beginning to share this information with their employees and families. Their goal is to educate enrollees to become "value-based purchasers" themselves, i.e., to identify and select the plans that offer the best combination of cost and quality—not just the ones that are most convenient or offer a particular doctor. While these employers do not expect enrollees to act upon this information immediately, they are hoping that people will become more sensitive and responsive to the differences among plans and choose their plans accordingly, which will in itself create an incentive for plans to improve their performance.

Venders Offer Support, Databases

The pioneers are receiving a significant amount of assistance from a number of consulting firms around the country that specialize in employee health benefits, such as William M. Mercer, Hewitt Associates, Foster Higgins, and MedStat. In addition to consulting services, several of these firms offer their clients access to extensive national databases that allow for comparisons of health plan or provider cost and quality performance. Some firms are specifically trying to bring groups of employers together to participate in national projects to contract with health plans based on their ability to meet certain quality criteria, which is one way that these venders are helping to disseminate the accomplishments of the pioneers to other less advanced purchasers.

Return to Contents
Proceed to Next Section