Priority setting in health care: Lessons from the experiences of eight countries

Norway

In 1987, in the context of increased demand for health care resources and the question of how to prioritize their use, the Norwegian government convened the Lønning Commission, the first body to set forth principles for prioritization and discuss their implementation[23, 24]. The commission decided to use severity of condition as the exclusive basis for prioritization, outlining five priority groups

▪ Emergency care for life-threatening diseases

▪ Treatment which prevents catastrophic or very serious long-term consequences: example, cancer

▪ Treatment which prevents less serious long term consequences: example, hypertension

▪ Treatment with some beneficial effects: example, common cold

▪ Treatment with no documented effects

They believed this division could guide funding decisions for various treatments [24]. Ten years after the first, Norway convened a second commission. This commission acknowledged the need to take into account potential effect and cost-effectiveness as secondary principles to be balanced with severity, and introduced four priority groups: core or fundamental services, supplementary services, low priority services, and services with no priority. The commission attempted to define what services should be included in the first category by providing three criteria:

▪ Risk of dying from disease within five years is more than 5–10% (severity)

▪ Increase in probability of five year survival is greater than 10% (expected treatment outcome)

▪ Costs reasonable in relation to benefits

This second commission focused more on the process of setting priorities, than on the principles used to set them [23, 25]. One of their suggestions was that the process of priority setting should be started by establishing clinician groups that would set priorities within their specialties.

Netherlands

In 1990, the Netherlands established the Committee on Choices in Health Care, the so-called Dunning Committee, to discuss methods and principles for setting priorities. That year, the Dutch Cabinet decided to include approximately 95% of available health services in the publicly funded package [26, 27]. The committee felt that to ensure all necessary services could be readily provided, non-essential services should be eliminated from the package. It delineated four priority principles: necessity, effectiveness, efficiency, and individual responsibility. They described these principles as forming a sieve for sifting out services that should not be publicly funded ("Dunning's funnel") [27]. The idea was that one should apply these successively beginning with the principle of necessity, and then limiting the number of services provided. The principle of necessity was defined very broadly, basically meaning any intervention that could provide some medical benefit. With regard to the principle of effectiveness, the committee distinguished between interventions where there was evidence for an effect, where there was limited evidence, and where there was no evidence. The services would be further narrowed down by those that gave value for money, by only funding efficient services, and finally, services that were best dealt with by the individuals themselves were excluded. This latter criterion was not meant to exclude lifestyle choices, such as illness because of bad eating habits. Rather, it was meant to exclude services that could easily be paid for by the individuals themselves. One example the committee gave was routine adult dental care. Although such dental care is necessary, effective and efficient, adults can easily pay for such services out of pocket, and it is therefore best left to an individual's responsibility. There was also a strong focus on solidarity and emphasis on approaching macro-level decisions from the community point of view. While Dutch attitudes seemed to be shifting to accept more reliance on personal responsibility, the Dunning Committee did not want changes based on this shift to overlook the needs of vulnerable populations [26].

Sweden

In 1992, Sweden convened the Parliamentary Priorities Commission. Much work on substantive issues was left to local health authorities; however the central government commission outlined three platform principles: human dignity, need and solidarity, and cost-efficiency. The commission defined five priority groups based on the type of disease or treatment in question [17]. While cost-efficiency was listed as a platform principle, the committee was clear that cost should only be considered in comparing treatments for the same condition, and that measures of effectiveness that attempt to quantify quality of life, such as quality adjusted life years (QALYs), should not be used. As in the Netherlands, there was an emphasis on solidarity and needs of vulnerable populations. In 1994, Sweden convened a second committee, which focused more on eliciting public opinions, a factor that had virtually no role in the work of the first commission [21].

Denmark

In 1996 the Danish Council of Ethics laid out values that should form the basis of the health service: equality, solidarity, security, and autonomy. From these values, the council derived a general goal for the health service, framed in terms of "opportunity for self-expression...irrespective of...social background and economic ability" [28]. In attempting to fulfill this general goal, there were a number of "partial goals" under consideration, including equity, quality, cost-effectiveness, and democracy. The committee was explicit that these secondary considerations must be "balanced against each other," and that even once goals are agreed upon it can "become difficult when these goals and partial goals are to be translated into decisions with consequences for everyday life in the health service" [28]. However, it did not specify methods for choosing between them.

Oregon

The experience of priority setting in Oregon's Medicaid program starting in the early 1990's represents the most explicit, as well as one of the most controversial, examples of priority setting in health care in the US. The goal of the program was to extend coverage to all Oregonians below 100% of the federal poverty level (as opposed to 58% FPL, as it was previously) by limiting coverage to a basic bundle of services decided by the Medicaid budget and a cost-effectiveness ranking of available medical services [29–32]. The Oregon process was concerned with incorporating public values from the beginning, and the Oregon Health Services Commission sponsored eleven public hearings, forty-seven town meetings and fielded a telephone survey before the initial rankings were decided [29]. The information gathered was employed in developing the Quality-of Well-Being Scale used to determine the cost-effectiveness ranking of condition treatment pairs and subsequently the services that would be covered by Medicaid [33]. This initial method was, however, abandoned because of public outcry over the resulting ranking of services. In response to this public reaction as well as criticism of the methodology, the Commission basically set out to rank health care services based on more broadly defined criteria, where expert knowledge and intuitive judgments by the Commissioners about appropriateness played a much larger role. While the program of rationing Medicaid services in Oregon was successful in covering a larger population and reducing the number of uninsured, it was extremely controversial and faced a number of political and practical roadblocks along the way [32]. The Oregon Health Services Commission continues to this day and continues to update the methodology used to prioritize and the resulting list of prioritized services. There are several community representatives on this Commission, although its ongoing work is not accompanied by as much public discussion as was the original implementation of the program. Oregon presents an example of both the potential and difficulty of implementing explicit priority setting in the US context.

New Zealand

In 1993, New Zealand established its first National Advisory Committee on Core Health and Disability Support Services to evaluate which services should be included in the publicly funded health package. While recognizing that existing practices represented an ad hoc list of priorities [34], the Core Services Committee started with this list and worked to identify: 1) discrepancies in provision and management of services (between Maori and non-Maori, men and women, urban and non-urban populations, populations in different geographic regions, etc.), 2) areas where efficiency could be improved, and 3) preferences of communities regarding health care.

While the committee in New Zealand discussed principles in much the same way as other commissions, the discussion occurred in the context of making recommendations about covered services. The committee rejected the idea of having an "Oregon-like" list of services constitute the basic package, but did define eligibility criteria for coverage of specific services. In order to make appropriate recommendations, the Committee looked at unit cost and volume of treatment data for common conditions and identified areas where efficiency could be improved [34, 35]. They also used information on public values and opinions, gathered through public meetings, to inform their recommendations. Subsequently, the Committee, renamed the National Health Committee (NHC) in 1996, met yearly to reevaluate and recommend changes to publicly funded health services based on new information or developments.

Israel

In 1995 Israel passed a National Health Insurance (NHI) Law, guaranteeing health insurance coverage to all citizens. The insurance would be provided by competing private sick funds, with the government acting as the single payer. Under the NHI law, all of the sick funds are required to provide the same basic basket of services to enrollees, with services not in the basic basket available at additional cost to the individual. When the law was first adopted it was decided that the basket comprising the extensive list of services offered by the largest existing sick fund at that time would also be the basic basket covered under the law. Thus, as in New Zealand, no explicit process for deciding on the basic basket was undertaken at the time, though it was recognized that a process was needed for updating the basic basket of services [36, 37].

The Ministry of Health created a process to undertake technology assessment and make recommendations regarding the updating of the basic basket of services. Under this system new technologies are screened on a regular basis in order to identify those with no existing alternative or with significant clinical efficacy compared with existing technologies. The assessment of these technologies includes the "evaluation of evidence-based clinical, epidemiological and economic aspects" [[38], p 172]. After several ad-hoc teams assess these aspects of the identified set of new technologies, their evaluations are passed on to the Medical Technology Forum, chaired by the Director of Medical Technology and including senior officials, managers, and researchers in technology assessment. Using a set of guiding criteria (including considerations such as the potential of the technology to prevent mortality or morbidity, the number of patients to benefit, the financial burden on society and the individual patient, and whether the net gain to society is higher than the cost) the Forum grades each technology on a scale of 1–10 and categorizes each as high, intermediate, or low priority. The rankings are then passed on to a National Advisory Committee, which, as of 1999, includes public representatives in addition to officials from the Ministries of Health and Finance and the sick funds. The Committee takes into account the assessments and recommends whether new technologies should be included in the basic basket, as well as limitations on their use [38].

United Kingdom

From its early days, the idea of rationing was a contentious political issue in the British National Health Service (NHS) [15, 39]. When other countries began convening commissions on priority setting there was a call by some for the UK to follow [19, 40–43]. Instead, in 1999 the National Institute for Clinical Excellence (now the National Institute for Health and Clinical Excellence) (NICE) was established to: 1) appraise new health technologies, 2) develop clinical guidelines, and 3) assess interventional procedures [44, 45]. In conducting these activities, NICE addresses questions including what constitutes necessary and appropriate care, how to incorporate cost-effectiveness considerations, and what interventions should be publicly funded. NICE also includes various pathways for public input.

Solicitation of public input and promotion of public discussion and education

To fully evaluate the extent to which the priority setting bodies have promoted public discussion and education, we would have to distinguish between what the commissions did to encourage discussion during their tenure, and any effects the commissions have had on subsequent public involvement. Here we will confine ourselves to the involvement during the existence of the commissions, and the structures set up by the standing bodies established. All the priority setting bodies recognized the importance of transparency in setting priorities, by emphasizing the importance of promoting public discussion in order to make the need for priority setting clear. The Swedish commission, for example, emphasized that public discussions of priorities help clarify the reasons and methods on which priority setting decisions are based [21]. Similarly, the Dutch Committee emphasized the importance of introducing the topic of priorities into public discussion, not only to make people aware of the need for prioritization, but to encourage individuals to make their own choices regarding health care options [27]. The Commissions, however, varied in the way the public was involved.

While Norway involved public representatives on its commissions, the Norwegian Commissions only discussed the importance of public education, whereas the Danish Council actually held public meetings and distributed materials on priority setting [28]. The Commissions in the Netherlands and Sweden actively incorporated feedback from public meetings and surveys into their deliberations [21, 27]. The second Swedish commission referred to four public surveys, two of which it funded, and held five regional conferences[21, 63]. The Dutch committee set out a plan for a long-term program to solicit feedback from various consumer groups, including women, the elderly, the "low-involved," and different patient groups by beginning a discussion primarily through the media, then soliciting feedback through various meetings [27].

One of the stated goals of the New Zealand's Core Services Committee was to ensure that core health services "reflect the diverse needs and values of the population being served" [34]. The first step in this effort is to inform the public and engage them in a discussion of health services. The New Zealand Committee held periodic "Best of Health" public meetings based around discussion documents. The views expressed at these meetings were part of the committee's considerations in making recommendations regarding core services [64]. Further, they recommended the continuation of consultation by the Health Authorities in charge of funding services with communities, including Maori communities, in different regions [34, 65, 66]. The NICE model of public involvement allows for public input at different levels on both broad principles and specific guideline development. Input is solicited by including lay members with relevant experience on guideline committees, posting draft guidelines on NICE's website to solicit public feedback before issuing final guidelines, and convening Citizens Councils, composed of 30 individuals representing the public, to discuss questions such as how to define and evaluate clinical needs for treatment or what role age should play in making clinical decisions [44].

In Israel, explicit public involvement was not taken into account in establishing the terms of the NHI Law in 1995, though the importance of involving members of the public in the subsequent decision-making process about the basic basket was quickly realized. Thus, beginning in 1999, a quarter of the membership of the National Advisory Committee that makes the final recommendation to the Minister of Health regarding which new technologies should be added to the basic basket were public representatives with no medical background [38].

Establishment of principles

Except the UK, all of the countries considered here started the discussion by establishing some set of principles on which to base priorities [14, 66]. In the UK, this discussion has gone on through other avenues, but has not been centralized and systematic [39]. The principles cited by the commissions include a range of medical, philosophical, and economic factors [17, 21, 27, 28, 34].

It seems unreasonable to base prioritization on a single principle. Indeed, when the Norwegian Commission attempted to formulate a system based exclusively on severity, it found that important considerations, such as effectiveness of interventions and cost, were excluded and saw the need to add additional principles. With multiple principles, the challenge is determining how to balance them when they conflict. As the Danish committee pointed out, balancing "become [s] difficult when [principles] are to be translated into decisions with consequences for everyday life in the health service" [[28], p 57]. While commissions acknowledged the challenge of balancing, none solved the problems they identified.

In general there is hesitancy to place much weight on CEA due to discomfort and uncertainty in dealing with cost. In Norway, the first commission avoided cost [24], and only after negative responses did it add the secondary principles of potential effect and cost-effectiveness. The Danish committee discussed problems with using cost related analyses, including uncertainty about what measure of utility to use and the need for more information. Rather than addressing these problems, the committee was hesitant to endorse the use of any economic analyses [28]. The initial Swedish report stated that cost should only be a deciding factor when "all else is equal"[21]. In both Denmark and Sweden, the commissions specified that cost should only be considered when comparing treatments for the same illness, such as comparing a titanium hip prosthesis to a less expensive but less durable steel prosthesis [21]. While Israel's process does not include a requirement that cost-effectiveness specifically be taken into account in making decisions, it does include an economic evaluation that considers the overall cost of including a new technology in the basic basket by comparing it with currently available treatments. In a small number of cases, explicit cost-effectiveness or cost-benefit analysis is conducted as well ([38] The Dutch Committee in principle allowed the lack of cost-effectiveness to determine that an intervention should not be covered.

Only New Zealand lists cost-effectiveness as a primary consideration [34]. NICE explicitly integrates cost in every case of guideline development and technology assessment [44, 67]. Still, the use of CEA in NICE guidelines has been controversial in the past; its recommendation against using beta-interferon for treatment of multiple sclerosis based on cost-effectiveness grounds, for example, caused an outcry from MS groups, and its final recommendation was controversial among the medical and research communities as well [34, 65, 68–72].

Impact on policy and practice

Because of their abstractness, government commissions that outlined principles have had little direct impact on their countries' policies. For example, the Danish Council of Ethics explicitly noted that their partial goals would sometimes conflict, yet did not set out methods for implementing them in these cases [28]. Even when a commission outlined priority groups for use in practice, the guidelines were so broad as to be useless in resolving difficult prioritization questions [17, 24, 27]. In Norway, for example, the division of priority groups into fundamental, supplementary, low priority and no priority did little to help resolve questions of choice in individual circumstances. Further, the lack of political tension within the Swedish Commission, which included parliamentary representatives of all major parties, was seen by many as a sign that the group avoided controversial issues central to priority setting [23].

After the early rounds of priority setting discussions, principles set forth by the committees had little direct effect on health care planning and provision. According to Berg and van der Grinten, the criteria that make up Dunning's funnel were difficult to implement because of contention about the definition of "necessary" and the difficulty of making macro level judgments about the effectiveness and efficiency in specific cases or whether an intervention can be left up to individual responsibility. Even when the Dutch government took the approach of the Dunning committee seriously, "problematic substantive criteria..., financial considerations...and political pressures ...made it very difficult to remove services from the package" of publicly funded health services [[26], p 124]. In general, governments and planning groups continued to make decisions about coverage based on a host of other factors, including political concerns and media pressure [14, 61, 73].

After a decade of discussions and repeat performances by some commissions, little progress had been made. Changes in health services after recommendations were issued did not reflect the extensive discussions and proposals put forth by the commissions [73, 74]. One study showed evidence that less than half of new medical technologies actually implemented between 1993 and 1997 in Norway fit the Lønning definition of core services [74]. "By 2002, few of the recommendations [of the Lønning Commissions in Norway] had been implemented" [[24], p 2003]. In particular, specific priority setting committees were supposed to issue recommendations for different fields of medicine, but as of 2007 this had not been done. While the discussion of priority setting was successfully started, it was not clear that the government commissions had any significant impact on actual practices at the policy, planning, or clinical levels.

The approaches of Israel, the state of Oregon, New Zealand and the UK affected policy and practice most directly. For example, by making specific recommendations regarding covered services in New Zealand and establishing specific clinical guidelines in the UK, the groups in these countries have anchored the discussion of priorities in concrete policy determinations. Yet, within NICE, there is no systematic review of existing health technologies; thus there is a bias towards reviewing only new technologies. While the UK is a step ahead in affecting practice, there are still improvements to be made in the system.

All the countries that have set up bodies that decide on priorities, and the state of Oregon, have ongoing processes that are conducive to iterative reflection on past successes and failures. In Israel, new technologies are screened and assessed for inclusion in the basic basket of services on a regular basis. Similarly, the ongoing nature of NICE creates opportunities for review and evaluation of the process, though the review has been less systematic than in New Zealand. The yearly reevaluation of the health services and discussion of other health care issues by the NHC in New Zealand is the best example of effective review and evaluation [14, 41]. For example, in 1996 the Committee recommended against population screening for prostate cancer using the prostate specific antigen (PSA) test, but recommended the question be kept under review [75]. In 2001 the NHC began a review that culminated in a 2004 report echoing their earlier recommendation against population-wide PSA screening [53, 76, 77]. Rather than attempting to settle questions of prioritization with one discussion, the New Zealand Committee has established an iterative process that allows priority setting to evolve with medical and political changes.

Finally, one should note that the priority setting efforts had little effect on the political events that actually led to the establishment of the priority setting exercises in the first place. The Norwegian effort did not solve the problem of waiting lists, the Dutch politicians did not use the criteria of the Dunning committee to decide what should be included in the package of health services, and the Oregon effort did not lead to any substantial cost-savings by eliminating interventions from the services provided that could be used to increase the number of people covered by Medicaid [24, 26, 78].