Several recent studies offer compelling evidence that American health care is not as efficient nor as evidence-based as it should be, and that Americans suffer from underuse, overuse, and misuse of care. Many experts believe that greater (and different) delivery system organization is fundamental to improved quality and efficiency. A redesigned delivery system would require an infrastructure largely absent from the cottage industry form of physician practice today.
Some analysts suggest that a model already exists for providing that infrastructure: the organized delivery system. One challenge faced by advocates of this model, however, is the difficulty of clearly describing its structural elements. Due to confusion over the definition of organized delivery systems, a literature search to determine how such systems perform is not straightforward. One study may address a specific structural attribute of the organized delivery system, while the next may address another, etc.
This report addresses that challenge by proposing a framework for synthesizing an emerging but disparate body of literature linking specific attributes of organized delivery systems (specifically, physician practices) to improved quality and efficiency, so that research addressing one attribute can be usefully considered in light of research addressing other attributes. The report also summarizes recent literature, highlights areas for further research, and discusses the role of policymakers in promoting physician group attributes linked to quality or efficiency.
Studies were selected for inclusion in this report based on an informal literature review, discussion with experts working in this area, and review of references in key articles. Emphasis was placed on more recent studies (post-2000), as the intent was to capture emerging knowledge.
A Framework for Synthesizing the Literature
Experts use many terms to describe the same general concept of "organization" among physicians, including integrated delivery systems, organized systems of care, organized delivery systems, and accountable care systems. Each of these terms stakes out a territory that is a point along the organizational continuum, from none to complete vertical integration.
At what point along that continuum is there "enough" organization to meet the quality and efficiency needs of a 21st-century health system? To address that question in the literature this report first identifies a number of measurable characteristics that represent the core elements of organization. The common denominator of the many concepts of organization referenced here is the physician group practice. A few specific physician group attributes are both key to most of these definitions of organized delivery systems and have been extensively studied in the research literature. These attributes are:
- Cohesion. This term describes the degree to which physicians practice collaboratively in a group, with shared purpose, performance measures, and often finances. Because no bright line separates cohesion and its absence, in the literature this quality is often defined by delineating between "true" medical groups and independent practice associations.
- Scale. Separate from the degree of cohesion within a practice, a minimum practice size may be required to support necessary infrastructure for quality and efficiency improvements. (Some diseconomies of scale above a certain size may exist.)
- Affiliation. This characteristic situates the practice in a larger context. Is the practice part of a system that can provide infrastructure support? Such a system might be created, owned, or supported by a health plan, hospital, physician group, or independent entity.
Most studies of these physician practice attributes measure quality in a handful of ways, including HEDIS (Health Plan Employer Data and Information Set) scores; assessment of the application of specific evidence-based practices; and the presence of care management guidelines; electronic medical records or information technology capabilities, and other quality improvement activities. Researchers have hypothesized that various attributes of organization are positively associated with quality measures, which are, in turn, hypothesized to translate into actual quality of care.
Physician Groups, Quality, and Efficiency
This report summarizes several studies exploring the relationship of physician group cohesion, scale, and affiliation with quality of care. These attributes appear to contribute to quality, although the research is not entirely conclusive. Instead, findings synthesized here begin to buttress with evidence the theoretical case for delivery system reform through physician group organization. Today, the state of that evidence is not great, but it is good enough to be intriguing and to prompt further study.
The report also summarizes the somewhat-limited literature on the efficiency of organized delivery systems or physician groups. Efficiency of health care delivery varies greatly. Work by Elliot Fisher and colleagues indicates that if health care providers in all regions of the country were as efficient as those in the most efficient regions, Medicare savings of up to 30 percent would be possible. The challenge for efficiency-seekers is to identify which 30 percent of care is unnecessary and could be eliminated safely. It would be naïve to suggest that any health care provider has the key to doing this correctly. Some evidence indicates, however, that multispecialty and/or prepaid group practices use fewer resourcesor get more for the resources they do expendthan do other providers.
Although research suggests a link between group practice organizational attributes and quality or efficiency, researchers dont know exactly why these links exist, nor the direction of causality. Most likely, the attributes of cohesion, scale, and affiliation are proxies for other, more difficult to study, characteristics. Leaders of high-performing integrated delivery systems suggest several characteristics that are key to their performance:
- Strong Physician Leadership. Many of the best-known integrated delivery systems and large multispecialty medical groups were founded by strong and charismatic physician leaders.
- Organizational Culture. Shared vision, values, and sense of mission around stewardship for both individual patients and populations is critical to performance.
- Clear, Shared Aims. Clarity of aims allows for meaningful performance measurement and encourages internal, transparent sharing of performance data. Shared aims also ensure that different parts of the organization are not hampering one another's attempts to improve quality and efficiency.
- Governance. As used here, governance refers to an organizations ability to set goals purposefully and implement a plan to achieve them. Someone or something (e.g., a board of directors) can cause the organization to act collectively and intentionally to improve quality or efficiency.
- Accountability and Transparency. Accountability to employers and patients, coupled with transparency of information, can help improve quality of care. Research shows that groups with external incentivesfinancial or otherwisefor improving quality tend to score better on quality indices.
- Selection and Workforce Planning. In organized delivery systems, leaders can select providers for participation, excluding those who do not meet standards. Organized systems also can be more intentional about the mix of providers they include (e.g., primary vs. specialty care, physicians vs. ancillary providers), targeting them toward the population's health needs.
- Patient-Centered Teams. Multidisciplinary teams of providers may provide higher quality care than individual providers. As physicians organize and affiliate with other parts of the delivery system, their one-on-one relationships with patients can be leveraged to connect the patient to a team of providers and to the delivery system as a whole. Alternately, rather than being a key to the success of systems, teams may detract from patient-centeredness (or the human scale of care), as the relationship with a single provider becomes less important.
It is the attributes listed above (and likely others not listed), rather than cohesion, scale, and affiliation per se, that are hypothesized to create a causal link to quality measures and, ultimately, to quality itself.
Evidence increasingly shows that improved "systemness" drives quality and efficiency. Until a better understanding is reached of how specific organizational attributes contribute to systemness, however, policymakers should strive to create an environment that rewards quality itself (rather than tying incentives to organizational attributes). An important area of focus is the payment system. No amount of evidence of the superiority of systems will encourage providers to join group practices if payment incentives work in the opposite direction, as some do today.
The pure fee-for-service (FFS) payment model can discourage the organized, integrated care that is the hallmark of systems. Under FFS, physicians and hospitals are rewarded for taking actionsdoing procedures, prescribing drugs, performing tests, etc.regardless of whether the best evidence calls for such actions. FFS may also stand in the way of cooperation and collaboration across the delivery system, as each provider has an economic interest in providing more services for the patient, rather than in collectively determining how much and what mix of care is ideal.
Changing payment systems to reward quality and efficiency requires action on two fronts, both of which are examples of value-based purchasing. First, payments should reward better care. Schemes designed to do this include prepayment (coupled with quality measurement and reporting) and pay-for-performance, which builds on FFS. Second, the unit of payment should be large enough to encourage providers to seek efficient combinations of resources. A bundled payment for a complete episode of care, for example, might encourage coordination of inpatient and post-acute care.
As policymakers and purchasers focus on quality and efficiency outcomes, researchers should continue studying high-performing health systems to understand how they produce value. This work would provide a foundation for understanding how the best attributes of organized physician groups can be adapted for use in the broader, less-systematized health care mainstream.