It seems unfeasible to design a priority setting framework intended to curb unsustainable growth of health care expenditure without including some measure of cost-effectiveness. Both nations have de facto incorporated a
cost-
effectiveness principle as part of the backbone of priority setting guidelines. With regard to the measurement of effectiveness, over the last two decades the
quality-
adjusted life year (QALY) has become the canonical outcome measure for health technology assessment. The QALY is a preference based outcome measure integrating morbidity and mortality [
118]. However, there seems to be a broad consensus among the public, politicians, health professionals and bureaucrats that cost-effectiveness alone does not capture all the salient attributes of illness and disease relevant to public priority setting, and that there are more to be said before fair and morally acceptable priority setting can be carried out. Particularly, the idea that an intervention or a treatment targeting patients with particularly severe conditions ought to be prioritised above treatment targeting patients with less severe health conditions, despite being less cost-efficient, has broad support in policy and among health care personnel and the citizens [
29,
105]. Both Norway and Sweden have therefore included a
severity criterion. How the severity criteria are being interpreted and implemented, however, are rather different.
Severity in Norway
The first parliamentary committee concerned explicitly with priority setting in Norway, was the first Lønning commission in 1987 (Lønning I). As public health expenditure expanded rapidly from 1960 to 1980, previous commissions had already considered various aspects of rationing [
61,
100], but Lønning I was the first comprehensive public inquiry into when and how rationing of health care could be carried out. In the resulting white paper (NOU)
2 [
59], a very thorough discussion of priority setting in health care is given, and several criteria are discussed. The commission’s mandate included analysing the de facto practices for priority setting in the Norwegian health care system, and—if found to be acceptable—to build upon these. To this end, Lønning I identified
degree of severity,
egalitarianism,
waiting time, and
cost-
effectiveness as “relevant dimensions” for priority setting.
It might seem odd to identify
waiting time as a criterion of priority setting; systematically making patients wait for indicated treatment is probably better understood as a mechanism for addressing more urgent needs first, for securing equality of access, and to some extent to ensure that
expectant management is attempted before more costly treatment options are employed. That is, waiting lists are best understood as a priority setting tool. The discussion by Lønning I of waiting time in the health services highlights that patient’s wait for different reasons, and that while delayed treatment might be unacceptable for some, it is a reasonable rationing mechanism for others [
59, pp. 77–78]. Lønning I also finds that egalitarianism, phrased as universal and equitable access to sufficient health care for the attainment of one’s “health potential” should remain an important principle of Norwegian health services. Nevertheless, it is severity that the commission considers the paramount parameter for priority setting, and severity is consistently linked with the concepts of
need,
urgency,
(risk of) death, and
suffering. Cost-effectiveness is also considered important in Lønning I, but, it seems, subordinate to the severity criterion.
Lønning II appeared in 1997 [
60], and was the second parliamentary committee on general priority setting criteria
3 for the Norwegian health services. The mandate for Lønning II explicitly called for an operationalisation of the principles from Lønning I, including the aim of making them applicable also to individual patients.
It is clear that Lønning II considers that integrating cost-effectiveness and the severity criterion is inherently threatened by the conflict between consequentialist and deontological ethics. The resolution of this dilemma is not carried out, and Lønning II recommends rather vaguely that human dignity must be the loadstar of priority setting, recommending that the different medical specialities must deliberate and decide how exactly severity is applied to their patients. Equity in health was considered to be subsumed by these principles [
88].
In 2014 the Norheim commission completed the third NOU on priority setting—
Transparent and fair [
74]. This commission’s work upheld the focus on cost-effectiveness, but suggested a radical re-interpretation of severity and health equity. The
health loss criterion—building on lifetime-prioritarianism—was proposed as sufficient for dealing with the severity issue [
70,
72,
81,
82]. The ensuing debate was dismissive [
7,
35,
52,
69,
71,
75,
80], and a second commission, led by Jon Magnussen, was appointed to reassess the severity criterion. The Magnussen report was submitted in 2015 [
62], in which
absolute shortfall was proposed as a better alternative to Norheim’s health loss criterion. Under the health loss criterion, a life-time prioritarian account is employed [
82]. A condition’s severity is quantified by the number of life-time QALYs—considering both the past and the future—that an individual stand to lose, relative to a norm of 80 QALYs. This results in a priority setting criterion with a strong redistributive component. The absolute shortfall criterion differs in two important ways: it does not consider the past, but only considers the expected future QALY loss. Secondly, it does not assume a norm of 80 QALYs, but instead computes the gap to the age-specific expected number of remaining QALYs for an individual in good health. The absolute shortfall thus retains a redistributive aspect, but somewhat less pronounced, and no longer retrospectively. Both have in common that the QALY is the sole numeraire for severity.
An important feature of the Norheim-Magnussen debate is that, despite acknowledging that severity is ambiguous and multifaceted, the impact they had on official documents and priority setting debate was to highlight QALY-based measures of severity as the standard and to relegate other concerns to the periphery of the debate [
77,
90]. Very simplified, we could say that Norheim’s health loss integrated severity and inequity and that Magnussen’s absolute shortfall fix to this (perceived) problem, largely deposed non-QALY concerns from consideration.
In 2018, another commission—the Blankholm commission—submitted yet another NOU:
The most important first (
Det viktigste først) [
9]. This NOU discusses priority setting for municipal services, and largely concurs with Magnussen’s report on the issue of severity. However, an important deviation is implied by questioning the applicability of the QALY for non-specialist services—epitomised by the term
mastery as an independent goal for treatment. The Norwegian term employed—‘
mestring’—is perhaps better translated as
coping, although
coping here does not quite capture the full connotations of the Norwegian ‘mestring’, which suggests getting ‘on top of’ (e.g. capability to meet social roles, conduct activities of daily life, and find meaning and dignity.) rather than merely ‘getting by’. ‘Self-management’ is also a candidate term. Secondly, the Blankholm commission insists that a notion of
basic needs should be defined and take precedence over severity (and equity) concerns in a priority setting context. The report goes as far as stating that covering these—still to be defined—basic needs should
not be subject to cost-effectiveness constraints.
Severity in Sweden
In Sweden a parliamentary committee on priority setting in health-care was established in 1992, resulting in a white paper in 1995 and in legislation in 1997 proposing three principles for priority setting:
The Human Dignity Principle, the
Need-
Solidarity Principle, and the
Cost-
Effectiveness Principle [
28,
87,
104,
105].
The Human Dignity Principle is the overriding ethical principle and addresses personal characteristics and functions in society that should not impact on priority setting (and therefore should not impact the assessment of severity), e.g., talent, social position, responsibility, income, chronological age, and gender. The concept of ‘need’ in the Needs-Solidarity principle is defined in terms of providing more resources to patients with the more severe diseases, but also worse quality of life, seemingly making a distinction between severe disease and bad quality of life. It is claimed that the care of severe disease and substantial quality of life decline should be prioritised before mild conditions, even if the former care is substantially less cost effective. Later in the white paper it is argued that a patient’s need is likely to depend both on the severity of the disease but also the duration of disease—seemingly making a distinction between current severity and duration.
It is claimed that a disease can vary in severity over time even for patients with chronic ill health conditions. It is also claimed that severity can be judged according to the suffering of the patient, the medical prognosis, the disability, and the existential despair experienced. In a later operationalisation by several central actors in the Swedish health-care system of the needs-solidarity principle, all of these different considerations were in some way subsumed under the concept of severity of a condition [
85].
Moreover, severity has come to play an essential role in determining which treatments are subsidised by the Swedish pharmaceutical and benefits system (TLV), i.e. that which society pays for rather than the patient herself. More specifically, different thresholds of cost-effectiveness are considered acceptable depending on the degree of severity. For example, during the last few years, decisions indicate that a threshold of 1,000,000 SEK/QALY is accepted for conditions considered to have the highest degree of severity, 750,000 SEK/QALY for severe conditions, 500,000 SEK for conditions of moderate degree, and a similar decrement to the lowest degree. At present there are, however, different approaches to assessing severity in the Swedish health-care system, causing potential inconsistencies. Nonetheless, the policy of TLV is clearly different from the absolute shortfall policy of Norway: in Sweden severity is not measured using shortfall in terms of QALYs. At least not consistently. The Human Dignity Principle and its ban on taking chronological age into account has generally been interpreted as an obstacle to adopting an absolute shortfall approach to severity.