Introduction

Health care systems throughout the developed world face ‘crises’ of access, quality, financing and sustainability. These ‘crises’ have developed, in large part, from advances in health science and technology, as well as changing demographics, increasing demands for new services, rapidly rising costs of new drugs and technologies, highly publicized cases of denial of services and wait lists as well as calls for greater transparency, inclusiveness and accountability in public policy. During the same period when virtually all industrialized nations developed health care systems committed to providing public coverage for medically necessary services, the power of modern medical science and technology grew dramatically, expanding the concept of medical need. This power, coupled with access to medical information and consumer choice, finds every country with publicly funded health services reeling to respond in a fair and fiscally responsible manner as health care threatens to consume the budgets of other programs, including those essential for health.

These pressures have led governments to look for more efficient and equitable ways to allocate public resources [8, 9]. The definition of baskets of core publicy covered services and the prioritisation of health care services for public funding have been some of the strategies used by decision makers to reconcile growing demands with limited resources. By definition, a priority is “of highest or higher importance; something given special attention; the right to take precedence in obtaining supplies, services etc” [15]. Priority setting, then, is a process to determine what is important and should be given special attention, particularly in the allocation of resources. As such, it is an intrinsically normative/ethical process. While resource allocation and priority setting are often conflated, they are distinct but related processes.

Prioritisation occurs at all levels of the health care system [35]. At the macro-level, governments determine the relative priority to be given to different public programs (e.g., health vs. transportation vs. education) and the extent to which these priorities will be met through public or private revenue sources. [37]. The meso-level is concerned with both the allocation of resources across organizations and institutions and services and programs within an organization [40, 50]. At the micro-level, health providers decide the type and amount of care to be offered to individual patients [52, 60]. These levels of priorities are inter-related.

Our concern is for priority setting at the macro level since all other levels are dependent on the fundamental decisions regarding public or private coverage of health services. Further, prioritisation involves many actors at each level including politicians, bureaucrats, health care managers and providers, health profession unions, patients and families, the public, advocacy groups, and the medical industry. All present competing demands on available resources.

Prioritisation can be implicit or explicit. Implicit prioritisation, the historic norm, is characterized by a lack of clarity regarding the priorities; failure to identify the strategies and criteria used to determine priorities and an absence of public accountability for decisions [33]. Proponents of implicit prioritisation believe that public policy is best achieved by “muddling through elegantly” [41]. They claim that explicit priority principles limit organizational capacity to respond to complex and changing information in a sensitive and timely way [21]. Nonetheless, implicit prioritisation has been subjected to mounting criticism as arbitrary, inequitable, and lacking in the transparency and accountability necessary for public legitimacy. Moreover, implicit prioritisation insulates decision-makers at every level from blame by allowing them to avoid difficult decisions as they devolve authority for determining who receives shared resources to others, such as physicians at the individual patient level [22].

In contrast, explicit prioritisation is concerned with establishing clear priorities; making transparent the rationale for these priorities and basing resource allocation decisions on agreed-upon priorities. It focuses on principles, norms and values (e.g., need, effectiveness, cost-effectiveness, justice, solidarity) as criteria to guide the decisions or methods, strategies and processes to make the decision (e.g., deliberative, transparent, inclusive and accountable). Theoretically, explicit prioritisation improves the quality of decision-making, increases allocative efficiency and gives legitimacy to prioritisation decisions [8, 11].

The ethical issues inherent in the setting of public priorities in health care and their implications for shared public resources have received increasing academic and policy reflection [13, 26–29, 36–38]. Traditionally, ethics has been concerned with theoretical reflections on the nature of the good and what we ought to do, dealing in values, principles, and interests. Policy sciences have been concerned with the practical pursuit of the good, dealing largely in technical criteria and evidence. In recent years, scholars in each field have come to recognize that evidence is not value-free; values and interests underlie all policy choices [4, 47] and “implicit normative judgments are often hidden in technical criteria” [26]. Moreover, ideas about the good are conditioned by historical, economic, social and political contexts [24]. So, policy analysts are paying increasing attention to making transparent the values inherent in policy goals and processes and health care ethics is expanding from its clinical and research orientations to become more involved with questions of public policy development and evaluation [13, 18]. Every policy priority presupposes an underlying normative, ethical justification. However, despite its fundamental importance, explicit, normative policy analysis remains a neglected element in the field [27, 36].

In this paper, we briefly describe the approaches to explicit prioritisation at the macro level in a set of publicly funded health care systems; analyze the ways in which normative, ethical concepts are presented in these experiences; and identify key elements of ethical analysis which can promote meaningful normative frameworks for health priority-setting at the macro-level.

Methods

The actual and proposed goals of macro level priority setting in a set of international experiences (published in the English literature) are described and analysed through a normative ethical lens. The experiences are analysed in order to illustrate how normative ethical concepts were identified and utilized in the experiences and whether they seem to have played a role in the ultimate decisions regarding public coverage of health care.

International Priority-Setting Experiences

Norway was one of the first countries to attempt explicit prioritisation. Its goal was to develop a planning tool that would guide resource allocation for the expansion of medical care. The Lønning I Commission (1985–1987) proposed a five level priority system grounded in Rawlsian principles (i.e., priority to and solidarity with the weakest) and the concept of need defined as burden of disease [6, 49]. Prioritisation criteria included severity of the disease (main criterion), treatment effectiveness, and cost effectiveness. However, this framework proved difficult to implement, and in 1997, the Lønning II Commission was appointed to revise the methodology for establishing priorities. This ‘bottom-up’ process was based on four prioritisation criteria: severity of the condition, expected benefits, reasonable cost-effectiveness and quality of the evidence. Expert groups categorized treatments specific to their specialty into four categories: basic services––indispensable services; additional health services––not indispensable but deliver significant benefit; low priority services––limited benefit; and services that should not be publicly funded, using the above prioritisation criteria.

In the Netherlands, the main goal of prioritisation was the definition of a publicly funded basic insurance package that would produce “the greatest medical benefit from the available resources” [43]. In 1987, the Dekker Committee initiated prioritisation by proposing that the universal, mandated basic care package be limited to 85% of the services provided at the time. The Committee suggested that only services that involved ‘financially insupportable’ risks or could not be easily substituted be included in the basic package. This proposal was never implemented due to considerable public opposition.

In 1990, the Dunning Committee attempted again to design a basic insurance package available to all Dutch citizens. The Committee argued that services included in the basic package should satisfy four criteria: they must be necessary from the community perspective; their effectiveness must be demonstrated; they must be efficient; and they cannot be left to the individual responsibility (i.e., are too expensive to be incurred by individuals). An abridged version of the Dunning Committee Report was sent to all doctors, hospitals and other heath professionals for feedback. Public involvement was via a massive public information campaign about the report involving over sixty organizations such as senior citizens’ clubs, women’s organizations, trade unions and academic institutions were consulted by the Committee. Two surveys, one to a randomly selected group of 471 nurses and 2196 physicians and another to 1700 randomly selected citizens, explored health professionals’ perceptions of prioritisation [57].

In Sweden, the Priorities Commission was implemented in 1992 “...to define a basic measure of security––or a minimum level of health and medical care––which is to be offered to all its citizens” [56]. The Commission developed a need-based prioritisation framework where need was informed by disease severity, suffering, functional impairment, and medical prognosis. It also proposed four criteria for the allocation of resources: human dignity (i.e., all people are treated equally with particular attention to the weakest), need and solidarity (i.e., resources committed to those with the greatest needs), and cost-efficiency (i.e., prioritisation decisions should aim for a “reasonable relation between cost and effect, measured in terms of improved health and enhanced quality of life” [56].

The Commission’s goal was also to stimulate public debate about prioritisation. It conducted five regional conferences and four surveys: public attitudes towards prioritisation through a random nationwide sample of 1,500 persons; physicians’ and nurses’ attitudes towards prioritisation through a random nation wide sample of 300 physicians and 300 nurses; attitudes towards international experiences of resource allocations in health care from a random sample of 571 persons, 168 politicians, 144 administrators, and 259 physicians; and a survey about new forms of management and changes in medical practice from a random sample of 671 physicians and nurses in the five Stockholm hospitals [55].

In Denmark, the aims of the Danish Ethics Council were to make priority-setting more visible and “to stimulate a process which may ultimately imply that the priorities set are better thought out and that they are set on the basis of more relevant considerations” [14]. The Council’s guidelines for prioritisation included: social and geographical equality; need, defined by the gravity and prognosis of the disease, urgency and capacity to benefit; and optimum efficiency [14]. There was no explicit public participation in developing the framework. However, the Council developed a dilemma game with examples of specific choices facing politicians and doctors, and in 1996, a public debate day was held in co-operation with the Association of County Councils asking “Do rational criteria for priority-setting in the health service exist?” and “Who is to set priorities and how?” A total of 380 participants from the political sphere, health services, patient and other organizations, lay persons, and the press attended. The feedback from these events provided inspiration for the Council’s guidelines. In 1997, the Council issued a debate package on priority-setting to employees in the health service and to counties’ hospital committees. Its purpose was to make people involved at different levels of decision-making consider the values they use to set priorities during everyday work.

In New Zealand, the Minister of Health initiated reform with the goal of explicitly defining publicly funded core health services that everyone should have access to “on affordable terms and without unreasonable waiting times” [42]. In 1992, the government appointed a Core Services Committee to increase public awareness about prioritisation and develop consensus about acceptable ways to define publicly funded services. The Committee conducted several waves of public consultation: town hall meetings were used to define broad health care priorities (1992); the public was asked for feedback on the proposed prioritisation criteria, i.e., benefit, value for money, fairness in access and use of resources, and consistency with communities’ values (1993); experts and public representatives were invited to provide input about effective services via consensus panels and conferences (1992–1993); and special efforts were made to consult disadvantaged groups (e.g., Pacific islanders, elderly, low SES) via focus groups about ‘fair’ resource allocation (1994) [17]. Ultimately, the Committee decided that a ‘simple list’ approach to core services was inappropriate because of the difficulty in determining treatment effectiveness and argued that all existing publicly funded services should be included in the core [10].

In the United States, the first large scale explicit attempt to prioritize health care services was conducted in Oregon [7, 20, 54]. The goal of the Oregon Health Services Commission was to develop a detailed list of primary and emergency care services that would be accessible to those qualifying for Medicaid. The approach was based explicitly on utilitarian principles, achieving the maximum health gain within the Medicaid budget. The first list of prioritized services proved sufficiently unpalatable to the population to require a series of changes to the methods used, from cost utility to net benefit analyses, to patients’ probability of surviving for five years, cost information and to a final ‘hand’ adjustment by the Commissioners. Public preferences about the values that ought to inform priorities were obtained via forty-seven community meetings with over a thousand Oregonians. From these meetings, 13 values emerged: prevention, quality of life, ability to function, cost-effectiveness, community compassion, mental health and chemical dependency, benefits many, impact on society, equity, effectiveness of treatment, personal responsibility, and personal choice. Additionally, 1,001 randomly selected Oregonians were asked to complete a survey (response rate: 23%) designed to collect numeric values representing the respondents’ evaluation of 26 disabling health states, including physical and emotional symptoms and limitations or impairments in mobility, physical activity, and social activity. Seven public hearings allowed special pleadings from individuals and interests groups.

In summary, these experiences of prioritisation at the macro government level varied significantly in their goals, criteria and consultation processes. Despite these variations, the results have been similarly disappointing. These prioritisation experiences resulted in marginal, if any, changes with regard to the stated goals of setting health priorities, allocating resources based on priorities, or the development of lists/packages of publicly funded core services.

Within seven months of its inception, the New Zealand Core Services Committee decided that a ‘simple list’ approach to the core was inappropriate because of the difficulty in determining treatment effectiveness and efficiency. The Committee shifted its focus from the macro question of which services should be in or out of the core to the micro question of how individuals’ access to appropriate services could be facilitated, and promoted the development and use of evidence-based guidelines and clinical criteria to determine this, with an emphasis on access to elective surgery [10].

In the Netherlands, the Dunning framework met with a great deal of public opposition. Recommendations to remove from the basket the contraceptive pill, long-term psychotherapy, speech therapy and home-help after a normal birth, initially failed. It took 3 years to remove adult dental care from the core services, a decision that was reversed in 1997, and reintroduced in 2004. As well, it took 12 years to remove the first round of IVF, despite the report recommendations [25, 58]. This restriction has since been widely contested because it has led to a significant drop of IVF treatments, particularly among less affluent people. Ultimately, the Government of the time recanted on its effort to develop a basket of essential services based on Dunning’s explicit prioritisation criteria (i.e., necessary care, effectiveness, efficiency, and individual responsibility [30].

The Norwegian experience demonstrated that clinicians are deeply influenced by a sense of duty to individual patients and reluctant to restrict patient services, regardless of cost or formal determinations of benefit. Progress to define essential and less essential services has been exceedingly slow. Additionally, studies indicated that the same physicians who were expected to apply the priority setting criteria failed to do so in a coordinated fashion and that physicians were willing to ‘game’ the system by re-interpreting the degree of severity of their patients’ illness to facilitate their access to care [49]. Norway focused on adopting a Patients’ Rights Act in 1999 to ensure that Norwegians have equal access to quality health care by granting patients a ‘right’ to necessary health care. Patients’ right to health care is based on the prioritisation criteria set out by Lønning II (disease severity, treatment effectiveness, ‘reasonable’ cost-effectiveness, and quality of evidence). The Act was strengthened in 2004 with the introduction of economic sanctions on Regional Health Authorities if patients do not receive a service within the specified time limit. An initial evaluation of this process as a whole has shown that there is still a wide geographical variation in the percentage of patients being assigned this right [48].

The Swedish prioritisation framework did not result in any definition of essential services. Even though the Swedish Health and Medical Services Act now requires by law that priority decisions be based on the three principles outlined in the Swedish prioritisation framework, there has been low activity within most County Councils and municipalities regarding concrete work with priority-setting [2]. Careful analysis has concluded that the principle of need dominates priority setting with very little attention given to cost-effectiveness [3], making efficient resource allocation almost impossible.

The Denmark outcomes are still unknown.

In Oregon, lobbyists forced the exclusion of many services from the ranking list (e.g., long-term care) and delayed other services from being added to the list (e.g., mental services). Services excluded from the Oregonian prioritisation process represented seventy percent of the budget, while prioritised services only represented thirty percent of the budget [5, 19].

Ultimately, these experiences either ‘grandfathered’ all existing services and applied prioritisation criteria only to consideration of new services or relaxed the prioritisation criteria and only rejected services for which there was evidence of ineffectiveness, while continuing to fund publicly those for which there was no evidence [34].

Ethical Analysis of the Prioritisation Exercises

The failure of explicit prioritisation to define core services has been attributed to varied causes: lack of political will, since explicit prioritisation creates winners and losers and, as such, is unlikely to be is a popular element of political reforms; implementation difficulties arising from a lack of high quality information, inadequate resources [46]; and competing ethical perspectives [37]. We argue that at least some of the failure can be attributed to the lack of clarity and rigor regarding the normative/ethical issues involved in explicit prioritisation [29, 59]. We examined the above prioritisation experiences through an ethical lens with specific attention to three inter-related types of values and principles: the ends or goals of prioritisation experiences, the process utilized, with a focus on the role of public participation, and the criteria developed for prioritisation decisions.

The most frequently identified goal of these experiences was the determination of core health care services for public funding. Except for Lønning I, they were more about the objectives of controlling publicly funded health care expenditures or rationing services, than establishing priorities for public funding (e.g., health vs education; health vs health care; acute health care services vs mental health, chronic illness). From the ethical perspective, these “goals” look more like practical policy objectives rather than values and principles directing the ultimate good desired. Rarely was priority setting understood as a process of identifying what mattered i.e., the values and social goals related to justice, equity and solidarity in the setting of practical priorities, for such a personally value-laden area as health care. Yet, as has been elegantly argued by Liss, unless this ultimate good is described clearly, it is virtually impossible to have a good policy result [38]. For example, the goal of improving access to care is not the same as the goal of reducing health inequities or that of balancing the budget. Neither the goals of medicine nor the goals of health care are simple and self-evident [23, 45]. Establishing priorities pertaining to what is most important to a nation, state or province requires articulation of the ultimate good(s) the policies ought to facilitate.

Most countries identified priority-setting as an exercise in distributive justice concerned with sharing limited resources. However, different conceptions of justice were evident [27]. While these differences are to be expected, the failure to identify them in public discourse resulted in no clear direction.

The conflation of priority setting and cost-containment appears to be an important factor in the lack of success, especially with lack of public acceptability. For the public, priority setting meant limits and denials of coverage. Moreover, when multiple goals were identified, there was little guidance on how goals were to be balanced against each other or utilized as criteria for making practical policy decisions.

Enhancing social capital through the stimulation of public debate and increased public awareness of the importance of prioritisation of public resources were identified goals for Sweden, Denmark, and New Zealand. However, these goals were not clearly assessed in any of the projects. Since these priority setting projects were about public resources and public goods, there was a notable lack of attention to the public policy element of these experiences and to intentional coherence between the goals and objectives of these health priority setting exercises and other public policies such as education or social welfare.

With the inevitable tensions and conflicts regarding goals of public policy in pluralist societies, procedural ethics are crucial. Denmark explicitly identified a process that should be open (the rationale for the prioritisation decisions is made clear), democratic (health care users are ensured influence on the prioritisation decisions), and inclusive of all parties concerned [7]. Some of the projects identified “accountability for reasonableness” [11] as ethical criteria for the process.

Today, meaningful public participation in public policy is understood to be a procedural facet of justice and fairness [12]. All of the prioritisation experiences, except Lønning II, involved some form of public participation. However, these experiences suffered from a number of shortcomings: the role of the public was unclear (e.g., consultative or decision making; setting the values base for decision-making or addressing specific decisions); ‘the public’ was often understood as having a single voice rather than the increasingly pluralist reality; there were significant problems achieving meaningful representation; there was inconsistent provision of information to participants; and, with the possible exception of New Zealand, general failure to promote strategies that fostered deliberation. Public involvement via town hall meetings (or equivalents) were either poorly attended or attended by those who had vested interests in specific outcomes. In Oregon approximately 2/3 of those who attended the community meetings were health care employees. Mass approaches such as large-scale surveys tended to reflect the ‘raw’ views of the public were of varying reliability and validity and did not always allow for generalizations. More deliberative strategies, such as focus groups and consensus conferences, involved small numbers of citizens with varying amounts and forms of information, and as such, were unlikely to represent adequately the range of views at the local, regional, or national level. The complexity of providing appropriate information to inform public involvement [16, 46, 53] was not clearly identified. While these experiences emphasized the importance of meaningful public participation in major public policy debates and decisions, they also highlight considerable difficulty in achieving it [39, 44].

Only a few countries (e.g., New Zealand) provided special opportunities for disenfranchised groups to participate in priority setting. There were no clear plans for how public input was to be utilized and no clear accounting to the public for their input in any of these experiences.

There was a clear privileging of professional expertise, particularly medical expertise, in most prioritisation experiences. Since physicians have an ethical obligation to act in the interests of individual patients and virtually no sense of duty to systems of funding and care, this dominance may be a crucial factor in the failure of these projects.

Ethical analysis of policy has focused on the substantive values, expressed as criteria or principles, upon which decisions are based. Not surprisingly, these elements were the most visible ethical aspects of these projects. They are summarized in Table 1.

Table 1 Prioritisation principles and criteria
Full size table

The variety of concepts on the lists and the lack of clear relationship between the normative concepts of principles and criteria are striking. Some clearly normative, ethical concepts such as universality, solidarity, equity, dignity, and care were identified. However, each of these concepts has many philosophical and political meanings. There was little definitional clarity provided. So often, for example, competing conceptions of justice were utilized in the same country-clearly utilitarian and libertarian, equity-based, in the same list of criteria. Important new thinking about issues fundamental to shared public resources such as the crucial values of solidarity [27, 31, 32] did not appear to be reflected in the reporting of these experiences.

Other principles and criteria were a mix of technical, political and ethical concepts such as acceptability, rejection of preference for status or discounting for age and lifestyle related conditions. Many were technical issues with a deep values base such as need, severity of disease, medically necessary, benefit, effectiveness, efficiency, and ability to benefit. The normative judgment was often “hidden in technical criteria” [26]. The ordering of the identified principles and criteria, in the face of inevitable conflicts, was unclear, except for Sweden where human dignity took precedence over the principles of need and solidarity, which took precedence over cost-efficiency. In general, clear and consistent definitions of crucial, and often contested, concepts such as need, health, medical necessity, burden of disease, ability to benefit, and clinical effectiveness were missing. The vagueness of the language did not allow for projects to provide clear direction, and so the actual decisions often became implicit, and reverted to providers on the ground.

Many of the criteria identified concepts based on values in empirical evidence such as quality, effectiveness, cost-effectiveness, efficiency and value for money. However, lack of high quality data about crucial factors such as the health status of populations, distribution of diseases, disability and risk factors, the relative degrees of need in different populations, and the relative effectiveness and cost-effectiveness of various services further hampered prioritisation experiences. It is widely acknowledged that, at best, only fifty percent of health care services have been evaluated for their effectiveness and even fewer have been assessed for their efficiency [34]. Much of the failure of the first Oregonian list has been partly attributed to the cost data, which have been described as a “crude guesstimate” created in a “data free environment”.

Different economic-based criteria were often used inter-changeably (e.g., cost utility, cost effectiveness, efficiency). The technical and ethical issues related to the utilization of QALYs were not addressed explicitly [1, 51]. Debate persists about such technical but critical issues as to how benefit and utility should be defined and measured (e.g., different instruments have been shown to produce different utility values); whose utility should count (individual or societal); and even the selection of a discount rate (which influences the attractiveness of prevention programs whose impact would be felt over many years). While economic evaluations, such as cost-effectiveness analysis (CEA), bring important information to the priority table and are crucial components of responsible stewardship of public resources, there are many important social values, such as solidarity, compassion and maintenance of hope not respected by reliance on CEA as the sole or main criteria.

Conclusions

Allocating public resources for health care has become a crucial issue for developed nations. As public systems are pressured by increasingly expensive and powerful advances in medical science and technology, new market-oriented values such as consumer choice, competition and profit are being incorporated into policy options. Globalization threatens traditional values that have been the basis of public health care such as solidarity, equity and compassionate response to health need. At the same time, the values of effectiveness, efficiency, responsible stewardship and sustainability need to be incorporated into policy choices. Health priority setting exercises at the macro level offer the hope that countries can set public priorities in response to health need that promote and sustain new conceptions of solidarity, equity and citizenship. Meso and micro level priorities are dependent upon fundamental system values.

While others have addressed many of the ethical considerations in health priority setting, our analysis identifies the importance of explicit identification of three interdependent value sets necessary for practical success:

terminal values–the ultimate goals

procedural values–process and

substantive values–criteria and principles utilized in the actual decisions.

None of these projects identified clearly a vision of the kind of society that was being shaped by setting priorities (or by failing to set priorities). The lack of clarity regarding the higher social purposes of priority setting for shared public resources in facilitating renewed conceptions of citizenship may be crucial in their failures. The meaning of and limits to solidarity need to be explicated more fully. The nature of public goods and public policy needs to be clarified.

The lack of a clear and common vocabulary within these projects made practical application and cross-experience comparisons virtually impossible. There was little attention given to the inherent conflicts between underlying philosophical perspectives inherent in fundamental concepts with clearly normative power such as utlilitarianism (value for money) and communitarianism (solidarity). While such conflicts are to be expected, there was no direction as to how to reconcile such differences in the final policy choices.

The dominant role of health care elites with their obligations to individual patients; the complexity of truly representative public input; the failure to clarify the role of public participation (advice, decision-making or politically correct ‘window dressing’); and the lack of adequate accountability to the public for the process are problematic from the perspective of procedural justice. The role of public participation needs to be better identified and the type and scope of decisions requiring the investment of public time and resources needs to be justified.

If explicit health priority-setting exercises are to be meaningful, we need to translate substantive academic work on important conceptual issues, such as health need, benefit, effectiveness, cost-effectiveness and efficiency need clarification, more effectively to public and policy audiences. The implications of failing to set meaningful public priorities in such an important area need to be made clear. Finally, the role of political power and values in priority setting cannot be overemphasized. No matter how good a priority-setting process, it requires political will to put into effect. Political acceptance of explicit priority setting may be the most difficult task of all.