The United States has no overall structure for the delivery of healthcare. It is difficult to organize improvement of health and healthcare in a way that engages all hospitals, clinical practices, and other health institutions. Moreover, the various regions of the country have different preferences for how care is organized and delivered. To cope with these features of American healthcare, voluntary regional associations or partnerships for health and healthcare improvement have been suggested since the 1960s.1 During the past 20 years, interest in regional collaboratives has grown steadily.2 There are now more than 40 collaboratives across the country.3
Over the past 7 years, the Robert Wood Johnson Foundation (RWJF) has nurtured a regional approach to health and healthcare improvement through its Aligning Forces for Quality (AF4Q) initiative.4 The RWJF’s term for a regional multi-stakeholder partnership is an alliance.
Aims for Alliances
In designing an alliance, the first question to answer is what the alliance will do to improve health and healthcare. Suitable aims include: (1) convening stakeholders to identify the community’s health needs and set goals; (2) promoting healthy behavior in the community, for example, exercise and avoidance of tobacco; (3) public reporting of healthcare providers’ performance; (4) improving the quality of care provided by hospitals and clinical practices; (5) controlling costs; (6) improving health information technology; and (7) reforming payment for healthcare. No alliance has the resources to pursue all of these activities; choices will need to be made. From the beginning of the AF4Q initiative, grantees have been required to pursue public reporting of performance, improve the quality of care, and engage the general public in these activities.
Designing Governance for Quality Improvement
Let us suppose that those forming an alliance have decided that improving healthcare quality will be one of its aims. In order to succeed, the alliance will need to engage hospitals, physicians, and other care providers in systematic quality improvement work. One possibility for the governance of an alliance is a multi-stakeholder board consisting of representatives from hospitals, the medical community, employers, health plans, government, and the community at large—or some subset of these constituencies. But a board with this composition will hinder engagement of hospitals and physicians because an organization governed by employers, health plans, government, and consumers is not friendly territory for providers. For many years, employers, health plans, and government have criticized hospitals and physicians for the high cost and uneven quality of American healthcare. Representatives of hospitals and physicians will be wary as they arrive at this board table. They may take their seats at the table in order to defend themselves, but they are not likely to join in wholeheartedly. Also, it may be difficult to determine whether they are truly engaged, since they will have good reason to appear engaged even if their engagement is only superficial.
It is a management truism that commitment is engendered by giving people authority.5 This insight provides the answer for fashioning governance for quality improvement— put the hospitals and physicians in charge. In other words, design the board so that it is dominated by providers or so that they have control by virtue of the positions they hold. If they control the endeavor, they will trust it. This governance arrangement is seen in highly successful collaborative improvement programs, for example, the California Quality Collaborative, the Institute for Clinical Systems Improvement in Minnesota, Maine Quality Counts, and the Wisconsin Collaborative for Healthcare Quality. At the same time, it is important for employers, health plans, and the general public to be represented in a minority so that the quality improvement program does not become focused solely on issues of high importance to the healthcare professionals, neglecting patients’ perspectives and the costs of care.
Designing Governance for Other Alliance Aims
If the alliance is a single entity and if governance is placed in the hands of the providers to secure their commitment to quality improvement, then, hypothetically, the providers would also control public reporting and any other activities of the alliance. This arrangement would not serve the whole community well. For example, an organization dominated by providers is not likely to be enthusiastic and thorough in publicly reporting the performance of those same providers. Provider-dominated governance will not serve all aims well. We need to consider some arrangement other than a single multi-stakeholder board—which would interfere with quality improvement—or a single provider-dominated board— which would promote quality improvement but hinder vigorous public reporting. The path to a solution begins with setting aside the notion that the alliance must be a unitary, multi-stakeholder entity.
A second line of reasoning leads to the same conclusion. Some alliance aims are not compatible when pursued by a single organization. For example, public reporting and improving quality do not fit well together. An alliance will have difficulty serving as both the public scorekeeper and the coach for healthcare providers. If the alliance serves the public assertively in reporting performance, it will be regarded with caution—perhaps even with suspicion—by the providers. If the alliance has strong working relationships with hospitals and physicians, and is thoroughly supportive of their improvement efforts, it will be hesitant to report performance that might embarrass or discomfort the providers.
Instead of using a single-entity model, one can envision the alliance as a cluster of closely connected organizations, each with governance suitable to its purpose. A multi-stakeholder entity is ideal for surveying the community’s health needs and setting goals. It is also well suited for other public health purposes, for example, promoting healthy behavior in the community. In contrast, a provider-dominated organization is not well suited to public health tasks because providers are typically not oriented to the whole population, focusing instead on caring for 1 patient at a time. Multi-stakeholder governance also fits well with public reporting, assuming that there is representation from employers, health plans, government, and consumers, and that providers are present in a minority. And multi-stakeholder governance will work well for those seeking to lower costs, to reform payment, and to deal with health policy in general. Again, it will not work well for improving quality.
Quality improvement is not the only aim poorly served by multi-stakeholder governance. The other aim that is better served by different governance is improving healthcare information technology (IT). The reason for separating this activity from the others is different from the reason for separating quality improvement. IT professionals have a culture that does not fit well with community governance or with the culture of healthcare delivery. IT culture is hierarchical. Also, because of understandably high levels of concern about protecting intellectual property, IT culture is secretive. The hierarchical mind-set fits poorly with the egalitarian atmosphere characteristic of multi-stakeholder, community-based organizations. The secrecy is foreign to hospitals and physicians, who may even regard IT secrecy as unethical because keeping medical or surgical technology secret is contrary to healthcare professionalism. For these reasons, improvement of healthcare IT is best not mixed with the other aims of an alliance.
A Structure for All Aims
Despite the differences in governance, pursuit of all of the aims needs to be coordinated. The Figure shows a structure that provides for coordination while at the same time
affording beneficial separation of activities. Three entities are shown in the Figure, with the regional health improvement alliance intended to be first among equals. Consider the structure shown in the Figure to be a model design for community-based health and healthcare improvement.
In the model, the alliance itself handles goal setting, public health outreach to the general public, and payment reform. It also serves as the forum for dealing with health and healthcare concerns that need to be processed in a communitywide setting. These discussions may lead to new initiatives for
the alliance. Or, they may lead to new initiatives for other organizations; for example, a campaign to educate the public about immunization and autism led by a public health agency working with the local medical society and news media. The alliance could also handle public reporting. A separate performance reporting agency is shown in the Figure only because this function could be handled by a fourth entity if, for reasons of historical development or local preference, using a separate entity were more attractive in a given geographical region.
The healthcare delivery collaborative needs to be dominated by providers, as already explained. It is their home in this overall design, and its separation from the other entities enables providers to focus on improving the quality of their care. Its most important job is to marshal peer challenge to motivate hospitals and physicians to improve. For this function, it uses the data reported by the performance-reporting agency as well as knowledge among providers—generated by the collaborative—of what other providers are doing to improve care. In addition, the collaborative trains and coaches hospitals and physicians in quality improvement methods. It also runs group projects that bring providers together to work jointly on quality improvement topics (eg, patient safety, disease-specific care [diabetes, depression]), patients’ experience of healthcare, efficiency of care, and culture change to foster improvement. In some regions, the collaborative could also produce clinical practice guidelines or evidence-based medical technology assessments.
The regional health information exchange has the task of providing electronic connectivity for all of the healthcare entities in the region, including hospitals, medical offices, and health plans. In some cases, employers might also participate— with due care for confidentiality of medical records. Advanced exchanges could also provide connectivity for patients. The exchange is also responsible for promoting the adoption of electronic health records in hospitals and medical offices.
All 3 entities in the model need to be connected in several ways. However, none of the 3 operates under the direction of any other entity in the trio. The lines connecting the boxes in the Figure represent interaction and coordination, not authority of any entity over another. The 3 boards should have overlapping membership, and there should be connections at various other levels. For example, there should be a standing committee of people from the delivery collaborative and the information exchange to design new data interchange capabilities to support quality improvement. In sum, the structure is similar to the divisionalized form described by Mintzberg except that the headquarters or central authority is replaced by a process of continuous consultation among the 3 entities.6
Multi-stakeholder regional alliances have generally been successful with public reporting, but they have struggled with quality improvement.7 Regional health information exchanges have generally not used multi-stakeholder governance and have often been poorly connected with public reporting and quality improvement in their regions. The 3-part structure described here promises to advance quality improvement by putting hospitals and physicians in control of this activity. At the same time, it promises to enable coordination of quality improvement, healthcare IT improvement, and all other activities needed to improve health and healthcare in a given region. Partial implementations of the model structure already exist in California,8 Maine,9 Minnesota,10 and Western New York.11 Research on the feasibility and effectiveness of this 3-cornered approach—compared with other approaches—would benefit community and healthcare leaders in those states as well as other states and metropolitan areas where this model might be considered.
The author thanks Professor James W. Begun for helpful comments on an earlier draft of this article.
Author affiliation: Division of Health Policy and Management, Schoolof Public Health, University of Minnesota, Minneapolis, MN.
Funding source: This supplement was supported by the Robert Wood Johnson Foundation (RWJF). The Aligning Forces for Quality evaluation is
funded by a grant from the RWJF.
Author disclosure: Dr Mosser reports board membership with the Minnesota Community Measurement and serves as a consultant for HEALTHeLINK, P2 Collaborative of Western New York, and the RWJF. He also reports receipt of honoraria from the RWJF and meeting/conference attendance with the Annual Colloquium of the Institute for Clinical Systems Improvement and the RWJF.
Authorship information: Concept and design; drafting of the manuscript; and administrative, technical, or logistic support.
Address correspondence to: Gordon Mosser, MD, School of Public Health, 1913 E River Pkwy, Minneapolis, MN 55414. E-mail: firstname.lastname@example.org.