Davies, Stephen M. MA, MSc; Bennett, Anita MA
Partnerships between a university and its affiliated clinical institutions produce outputs which have profound effects on society. Analysis of these effects typically focuses on the primary mission activities of such partnerships: clinical care, education, and research. The concept of the tripartite mission recognizes the interdependence of these three streams of activity within an overarching mission that binds partnerships together. But the effects of such partnerships on the communities they serve extend beyond the tripartite mission. Combined, the university and its clinical partners are powerful economic and social actors through their roles as major employers and as procurers of goods and services. In this article, we present a model for conceptualizing the full spectrum of economic and social effects, and we describe how we have applied this model in a number of center-specific studies. The setting for these studies is the United Kingdom, but the model is likely transferable to other national settings. The article is intended to be a contribution to methodology, rather than a systematic report of research undertaken.
The British Context
Before describing the model and its application, it is first necessary to briefly outline the British context for the study and to clarify some issues of terminology. The United Kingdom is home to some of the world’s leading universities for biomedical research. Medical schools, life sciences departments, and schools for nursing and allied health professionals work in close partnership for research and teaching with their affiliated university hospitals. University hospitals as well as universities are publicly owned, but they are accountable to different ministries for health and education. Accountability for research spans both ministries in a complex set of arrangements that policy makers are currently reshaping as a result of recent recommendations and government reorganization.1 The publicly owned National Health Service (NHS), which is funded through taxes and provides free health care coverage to all citizens of the United Kingdom, owns most UK health care institutions and dominates both clinical research and clinical education, with the private sector, an option for payers, playing only a minimal role in these areas. Partnerships between universities and the NHS are thus always built on nonintegrated governance and separate accountability, and they require effective liaison mechanisms to counter these attributes.
From an international perspective, the United Kingdom is unusual in that it has not, until very recently, regarded these partnerships as distinctive organizational forms meriting any special attention in policy making. The most obvious contrast here is with the United States of America, where the academic health center (AHC), the umbrella organization comprising the medical school and its affiliated clinical institutions, figures large in policy and academic discourse. An obvious explanation for this contrast might be that the United States allows models of common ownership and integrated governance that are not present in the United Kingdom. But this cannot be the whole explanation. Because of the national diversity of ownership, organizational form, and governance, some AHCs in the United States share the UK characteristics of nonintegrated public ownership. British university–NHS partnerships also hold in common with U.S. AHCs the tripartite mission, and they produce outputs such as specialist care and research that are also directly comparable with those of U.S. AHCs.2 Furthermore, other countries have public ownership and accountability arrangements similar to those in the United Kingdom, yet policy makers describe these in terms similar to those seen in the United States. For example, Canada uses the name academic health sciences centers, and The Netherlands uses the term university medical centers.
To explore why health policy in the United Kingdom has had so little regard for the topic of medical school/clinical hospital structure and governance is not possible in an article of this length, but it is apparent that the explanation does not lie primarily in differences of organizational form or outputs but, instead, in history, politics, and culture. Regardless of national policy, the leaders of British medical schools and universities have often looked to other countries and, in particular, to North America, for alternative models of governance that might enable them to more effectively pursue the tripartite mission. At the time of this writing, the country’s largest medical school, Imperial College, and two of London’s most illustrious university hospitals have advanced unprecedented developments in this area in the form of proposals to create the United Kingdom’s first “academic health sciences center ” through a merger and integration of governance. In parallel, stakeholders have begun to realize that they need a plurality of models to pursue the tripartite mission, even in a system as uniform as the UK NHS.3 Whether or not the academic health sciences center will become an enduring component of the British health care system remains to be seen. In this article, we use the term academic clinical partnership (ACP) to describe the combined endeavor of universities and their NHS partners in the United Kingdom. This descriptor is more neutral than the terminology used in other countries, which can sound both too centrist and too medical in a national context that values highly integrated systems and multidisciplinary clinical practice.
For the leaders of British ACPs, learning from the United States is not confined to matters of structure and governance. What is surprising from the British perspective is that U.S. AHCs are so very diverse in this respect, yet this diversity does not seem to undermine the utility of the AHC construct in the public mind. Emphasis on mission, rather than on form, is one means by which AHCs have achieved public understanding, regardless of the practical difficulties of precisely defining these entities. For example, a recent Institute of Medicine committee, when contemplating the diversity of the AHC sector, “chose to focus on the roles performed by AHCs and how they fit together, rather than on the AHCs’ organizational components.”4 This approach followed earlier leads in arguing that what is important is a focus on mission rather than on ownership, structure, and governance.5 The lesson for leaders of British ACPs is that communicating mission and impact might be more important to their collective interest than arguing for structural reform.
Reflecting on this lesson, the Association of UK University Hospitals (AUKUH) and the Council of Heads of Medical Schools (CHMS) commissioned from us in 2005 a study to identify and, where possible, quantify the impact of UK ACPs.6 After this initial national study, we carried out detailed studies of centers in three cities of the United Kingdom (two in England and one in Scotland). The purpose of this article is to describe the conceptual model developed for these studies and how we applied it in practice—not to present the individual impact studies in detail, although we have included illustrative findings drawn from these studies.
Institutional Impact Assessment
Traditionally, economists have used impact assessments to quantify the economic effects of an institution or sector on a defined geographical area so as to support a case for investment. Standard methodologies for measuring the economic impact of a sector or institution on local or national economies use input–output models. This approach analyzes the monetary flows to and from industries and institutions, including multiplier effects, to quantify the net economic effects of the unit of study on the economy of interest. This methodology has been used to impressive effect in the United States where, for example, a study commissioned by the Association of American Medical Colleges concluded that its membership (which includes medical schools and teaching hospitals/systems) had a combined economic impact of more than $326 billion and employed 1 out of every 54 wage earners in the U.S. labor force, either directly or indirectly.7
The primary criticism of this approach is that it is conceptually limited, reflecting a traditional economic viewpoint and ignoring more recent approaches such as endogenous growth theory, which emphasizes the role of education, training, and knowledge in producing long-term economic growth. Some have argued that, in higher education, for example, a primary output of the sector is the production of human capital, which is both longer-term in its impact and more difficult to quantify. Even more difficult to measure, but nonetheless important, is the social return from education, or the extent to which society as a whole benefits from an increase in the overall level of education. In response, researchers have refined models of impact assessment for higher education to try to capture this wider range of effects.8 Other limitations to traditional input–output models include the static view of the economy inherent in conventional impact analysis and its decreasing relevance as the spatial scale of the impact assessment is increased. In summary, conventional economic impact analysis is a useful tool, but it can neither provide a model that captures the full range of economic and social effects nor describe these in a holistic manner.
Developing a Holistic Model
The goal of the AUKUH- and CHMS-sponsored study was to describe and, where possible, quantify the societal effects of British ACPs. As a first stage, we developed a model to capture the full range of ACP outputs and the domains in which society might feel the effects of these outputs. The subsequent stages involved identifying and collating or collecting metrics for these outputs.
Figure 1 presents the model finally adopted. We considered, in the development of this model, a number of sources, including AHC economic impact assessments, approaches from studies of UK higher education (commissioned by Universities UK),8 work on the impact of UK NHS organizations beyond health care,9 and the UK government’s sustainability policies.10 The inner ring in the model represents the institutional partnership. The middle ring shows a range of outputs from the partnership which include, but are not confined to, clinical care, education, and teaching (discussed below). The outer ring represents the societal effects of these outputs and is divided into five domains: economic, human capital, social capital, knowledge, and place. These effects express themselves on a varying spatial scale according to the nature of the domain. So, for example, knowledge outputs may be global in their impact because the market for biomedical research is global, but social capital and place effects will generally be very localized because they mostly follow from the way in which the ACP engages with its host community.
Application of the Model
Using this model, we undertook detailed studies of three ACPs, two in England and one in Scotland. One of the English partnerships is based in a major conurbation and involves six NHS providers (hospitals and mental health care providers) operating in a network model with the medical school. The other is based in a provincial city in one of the United Kingdom’s most rural regions and is a one-to-one partnership operating from a shared campus. The Scottish ACP is also a one-to-one relationship but in an urban setting and in a context of major urban redevelopment, including redevelopment of the university hospital itself. In the following sections, we briefly describe the methodology adopted to operationalize the model for each domain of impact. We also report selected findings from the urban partnership in England to illustrate the application of the model.
We assessed the economic effects of the study ACPs at regional (typically about 5 million people in England11) and subregional levels (i.e., the English urban conurbation of approximately 2.8 million people). Defining an appropriate spatial scale is particularly important in the UK context because 87% of health care activity (2005)12 and 61% of university activity is publicly funded (2004).8 At the national level, it is likely that the government would spend much of this funding on other public goods were it not used for health. At the level of a region or subregion, this is less likely to be true. Many ACPs are situated in deprived urban areas and make a major contribution towards local economies that could not necessarily be easily substituted. Thus, any findings about economic impact will have most validity at the level of a region or a subregion.
Once we had established the appropriate spatial scale, the modeling of economic impact followed methodologies established for other sectors and included direct, indirect, and induced effects.8 We calculated the direct effects as the salaries and wages paid to employees as well as the goods and services procured by the organizations in the ACP. The purchase of goods and services stimulates, in turn, further demand for goods and services by suppliers to ACPs, creating a rippling out through supply chains. This ripple creates indirect effects. Further effects follow from the payment of salaries and wages to employees, who spend some of their income on consumer goods and services. This spending also creates a rippling out (known as induced effects) as it creates wage income for employees in other sectors, who also then spend their income, and so on. In addition, we considered two further types of indirect income: the impact of student expenditure and nonlocal hospital visitor expenditure on the local economy.
To quantify these effects at the appropriate spatial scale, we performed detailed analysis of where suppliers are located and where staff live. We constructed an input–output model, using multipliers derived from other studies, matching for both sector and location, to arrive at indirect and induced effects. In the case of the urban partnership, we calculated the economic impact on the subregion (including direct, indirect, and induced effects) as being in the range of £525 million to £565 million per annum in 2004–2005 and the impact on the region (population 6.7 million) as being between £750,000 million and £1 billion per annum.
Human capital effects
In classical economics, human capital is one of the factors of production and incorporates the knowledge, skills, creativity, and physical attributes of individuals. In the studies of three UK ACPs, we employed the concept in a broad way to capture the health care and educational missions of the ACPs. Preventive and curative health care maintains and restores, respectively, individuals’ ability to function both socially and economically. Education, training, and development broaden and deepen that ability and are required for the future delivery of health care.
For this domain, there is available a wealth of routinely collected data on the health care missions of university hospitals and the educational missions of medical schools. These emphasize the central role that university hospitals play in the British NHS. The 33 university hospital trusts in England, for example, deliver just under 30% of all inpatient and day case activity nationally. Throughout the United Kingdom, 32,000 undergraduate students and 13,000 postgraduate students are enrolled at medical schools. Interestingly, university hospitals now provide only 45% of all clinical teaching placement time for undergraduate medical students, a figure which has fallen from 62% as recently as 1999–2000. This reflects the continuing trend towards greater involvement in teaching by other parts of the NHS, such as local hospitals and community settings. University hospitals also employ a third of all doctors in training grades.6 These statistics illustrate the vital contribution of ACPs in maintaining and developing human capital through health care and education.
University hospitals are additionally important providers of education, training, and development across a broad spectrum from basic skills training to continuing professional development. Much of this activity is multidisciplinary and provides accreditation for a wide range of skills. Our studies found a range of activities in addition to undergraduate medical education that include postgraduate medical education, postgraduate multidisciplinary education, clinical skills training, information technology training, vocational training, specialist skills training, general training, and staff development.
Social capital effects
Social capital has become a dominant concept in policy making for international development, because of the growing evidence that it is a key determinant of sustainable development and economic growth. In developed countries, there is evidence that social capital is a strong determinant of population health, which is consistent with evidence on the impact of psychosocial factors on health.13 There is no single authoritative definition of social capital, but there is a broad consensus that it is a multidimensional concept which incorporates social networks, shared values, norms of behavior, and trust among individuals. Instruments for measuring social capital do exist, but these are all designed for use in neighborhoods rather than in institutional settings. It follows that no routinely available data are available for outputs in this domain, so we used qualitative methods for this part of each of the detailed studies. We interviewed stakeholders to identify ACP initiatives that seem likely to promote social inclusion, community participation, social networks, and shared values.
We found a range of areas in which the study ACPs are contributing to the building and maintenance of social capital, without necessarily conceptualizing their activities in these terms. These included participation in urban regeneration projects, support for volunteering, local recruitment initiatives, patient and public involvement initiatives, and making education facilities available for community use. Medical schools are actively engaged in projects to widen participation and work with schools and colleges to provide routes into medicine for those from less-advantaged social groups. The English urban center in this study, for example, runs a program of active outreach into local high schools to reach children from disadvantaged backgrounds who have the potential to study medicine. This program links a mentoring project, workplace placements, scholarship programs, and the provision of a foundation degree (a diploma-level qualification) through partnership with a local college.
This domain captures the contribution of ACPs to the stock of knowledge, both through research and through the application of new knowledge to create innovation. ACPs are uniquely positioned to undertake clinical research because of their ability to provide access to patients, their concentration of academic expertise, their research infrastructure, and their range of services. Research activity is the aspect of the mission that most distinguishes UK university hospitals from other health care providers. Research activity in university hospitals in England alone attracted more than £500 million of external grant funding in 2005, or more than 70% of all external grant funding received by the NHS.6 In practice, research funding agencies allocate much of the remaining 30% to networked projects which are led by principal investigators employed by university hospitals.
Knowledge effects can be measured using conventional metrics such as number of grants provided, number of articles and reports published, number of funded projects undertaken, as well as innovation income and number of doctoral students recruited. However, such approaches to measuring research impact, even when factoring in refinements such as journal impact and peer esteem, arguably fail to capture the full contribution of applied research. A recent UK forum concluded that medical research produces a wide range of socioeconomic benefits and that multiple methods are needed to demonstrate these.14 In applying the impact model, therefore, we supplemented the collection of conventional metrics of research and innovation output with detailed case studies of research groups using a multidimensional model adapted from a study for a leading medical research charity.15 The case studies examined effects under the headings of
* knowledge production;
* contribution to research capacity;
* contribution to new products, diagnostics, and therapies;
* contribution to changes to clinical practice or to service delivery and organization; and
* wider economic and social benefits, such as gains in health care productivity and safety.
This qualitative approach also allowed for the explanation of other important issues, such as the model of knowledge production in the ACP, where the research agenda is shaped by daily interaction with patients and their needs.
“Place” is a social construct which is of growing importance in UK public policy and is linked to goals such as sustainability and “livability,” or the extent to which localities have the features, ranging from public safety to green space, that make them desirable, healthy, and safe places to live. The concept of “place making” is now central to the debate on the future of local government in the United Kingdom.16 Place, in the sense of distinctive contributions of physical and social attributes, is also increasingly viewed by some researchers as a determinant of population health that operates powerfully at the neighborhood level.17 The domain of place impact may become especially important in the context of a hospital redevelopment, as was the case with the center in Scotland, or where the clinical facilities are placed in a setting of urban renewal.
To some extent, place cuts across the other domains. It cuts across the social capital domain because it includes attributes such as local governance—that is, the extent to which communities are well run with effective and inclusive participation, representation, and leadership. In the United Kingdom, for example, there has been a search for new models of neighborhood governance as part of a broader set of policies for addressing small-area inequalities. As yet, the policies of localism have had little impact on the NHS, but this may change as NHS Foundation Trusts become more widespread. NHS Foundation Trusts are semiautonomous organizations, still under public ownership, but under the direction of an independent regulator rather than of the Secretary of State for Health. Most importantly in this context, their constitution includes a public membership which controls the appointment of the Board of Directors. It is government policy that all NHS hospital providers should adopt Foundation Trust status by the end of 2008. Positive place impact also includes the promotion of a strong local culture with shared community activities, in which ACPs, as major employers, can also have a role.
The concept of place effects converges with the economic domain because it incorporates the strength of the economy at a very local level. In one of the studies, we found the university hospital acting virtually as an “economic island.” Complex supply chains, the absence of local employment projects, and a lack of penetration of the hospital site for local businesses meant that that the economic benefits in the immediate community were tiny compared with the hospital’s budgets. This represented a missed opportunity for the hospital to achieve wider community benefits that would have been reciprocated in strong local affiliation and support.
Finally, there are some more obvious place effects. Through their role as huge generators of people movements, ACPs can strengthen local communities with good transport services and communications linking people to jobs, health, and other services. They can also contribute through their impact on the environment, for example, through energy, travel, and waste-management policies. Finally, by commissioning high-quality design, they can make a positive contribution to the built environment.
We have outlined a conceptual model for assessing all aspects of the impact of ACPs on the societies that they serve. We describe how the application of this model through a mixed-methods approach can produce a spectrum of evidence on societal impact that will vary from highly quantified (e.g., economic effects) to qualitative (e.g., social capital and some aspects of place effects). What might be the value of this evidence? At the national level, it can be used to demonstrate the importance of ACPs to national economies and their contribution to wider social goals. This, then, strengthens the case for policies that sustain these partnerships rather than those that might weaken them as the unintended consequences of reform to funding streams for individual aspects of mission. At the more local level, evidence of positive social and economic effects will strengthen the perception of centers as key partners with local authorities and other agencies responsible for economic development, regeneration, employment, and other social objectives. This, in turn, may lead to success in obtaining grant funding, physical development permissions, and other practical means of support.
The main limitations of the model proposed at present are twofold. First, some could argue that it simply aggregates the impact of university and clinical enterprise without conclusively demonstrating the added value of partnership. To test this, some form of paired comparison between academic and nonacademic hospitals might be attempted. Second, it would strengthen the model if methods could be developed for quantifying effects in some of the domains, for example, a measure of the institutional contribution to the formation of social capital.
We developed the model described in a British context, using British terminology and reflecting UK public policy goals. This raises the question of its transferability to other national settings. ACPs exist in most developed countries, sharing the tripartite mission as well as common concerns.18 As a result, although some local adjustments may be needed, we are confident that the approach we have described is likely to be highly transferable to other national settings.
The authors acknowledge the support of the Association of UK University Hospitals and Council of Heads of Medical Schools, who provided funding for the study on which this article is based.