• allocation of health care resources
• distributive justice
• ethics
• health economics
• public health ethics

In a recent extended essay—The significance of ‘severity’—philosopher Daniel Hausman critically scrutinises the notion of severity as it is often employed in the literature on priority setting in healthcare.1 Hausman argues that severity is an undertheorised concept that lacks persuasive moral justification, and he thinks that we should give up operationalising severity as a criterion for priority setting in healthcare. We do agree that severity is an undertheorized concept. Indeed, we have pointed to the problems raised by Hausman, and an array of further ones, in our recent article Severity as a priority setting criterion.2 However, we do not agree with Hausman’s proposition that the notion of severity is without moral significance in the priority setting context. We will not address Hausman’s argument in full detail here, but instead elaborate on two essential points that have to do with balancing values and the definition of severity. Contrary to Hausman, we conclude that one should not abandon severity at this crossroad.

One generic problem that Hausman points to is that there is no, to all parties, fully satisfactory suggestion on how to balance different values against each other, as when prioritarians, for example, struggle to strike the right balance between maximising health outcomes with a concern for the worse off. The latter concern is an example where a notion of severity would likely be considered relevant for healthcare priorities. The general act of weighing values is neither novel nor unique to notions of severity. It is rather true of most, if not all, value conflicts we struggle with in healthcare ethics. Still, theoretical developments seem to bring us forward in concrete decision situations.

More importantly, Hausman seems concerned with a relatively narrow scope of conceptualisations of severity: his arguments seem directed at that specific phylum of ‘severity’ which is derived from the Health-Related Quality of Life/Quality-Adjusted Life Year-paradigm (HRQoL/QALYparadigm). After having rejected an age-dependent severity concept, illustrated by absolute and proportional shortfall QALY, with reference to intuition and balancing problems, Hausman focuses on scrutinising and arguing against the presentist HRQoL-based notion of severity that is advocated particularly fiercely by the health economist Erik Nord.3 4 Within this presentist paradigm, Hausman focuses on the commonly used relevance criterion of severity, implying that only relevant parts of a patient’s situation, as opposed to the lifetime total of the patient’s situation, should be taken into account when assessing severity. While the QALY type of severity Hausman discusses might be ‘the most popular definition of severity’,1 it is certainly not the only one.2

We, the authors of this response, do not necessarily agree on how to prioritise scarce healthcare resources. What we do agree on, however, is that cost effectiveness alone appears insufficient as the sole guide to priority setting in healthcare, all things considered. Hence, instead of abandoning severity—as Hausman suggests—we allude to a different path for overcoming the relevant critique raised in Hausman’s article. By pointing out unresolved questions about severity, we hope that our article— Severity as a priority setting criterion—will inspire research and discussions among scholars and policy-makers with the aim of reaching a better established view on how severity can be understood and ultimately incorporated in priority setting processes.

In Severity as a priority setting criterion, we distinguish between two types of questions regarding severity. One type of questions address the underlying reason(s) for why, or perhaps how, severity matters in the first place. These questions concern, for example, the way in which the concept of severity is best understood in relation to the more general discussion about distributive justice and other closely related concepts such as need and urgency. A second type of questions about severity seek to analyse what severity is. These questions seek to clarify the values which are imperilled when severity is invoked. More specifically, these questions concern how the notion of severity is related to, for example, an individual’s desires, (subjective) well-being and social contexts. A further subset of questions concerns severity in relation to a temporal perspective and patients suffering from comorbidity.

We do find reasons to explore further the temporally extended QALY-based versions that integrate the (perceived shortfall of) QALYs aggregated over the patient’s expected future (eg, absolute shortfall and relative shortfall), or the expected gap in lifetime QALYs to some externally set standard, as in Ole Frithjof Norheim et al’s health loss criterion.5 The QALY-based temporally extended versions are strongly motivated by—and capture several important intuitions of—lifetime egalitarian and prioritarian accounts of distributive justice. As such, these latter QALY-based accounts of severity can also be seen as identifying severity with a relative shortfall of health. Some of us have argued that principles of distributive justice (interpreted as mid-level principles) should be applied in the discussion on severity.6 7 This also solves the problem of moral justification: for prioritarians the worse offs should simply get priority because it is bad in itself that some people are badly off. Hence, Hausman’s rejection of this alternative seems premature.

A point we make in Severity as a priority setting criterion, and which the above discussion illustrates, is that severity is likely an essentially contested concept.8 But, contrary to conclude that severity ought to be discarded, we believe that it may be a diamond in the rough. Underneath the unpolished surface of its colloquial and more formal uses, we sense that there are many brilliant moral facets to be discovered and call for intensified theoretical and empirical research aimed at mining morally relevant severity from severity.

We see several possible paths such a research programme could pursue. One important route is exploring in more detail the different possible rationales for severity, taking into account recent theoretical developments, establishing different theoretically grounded and more well-developed accounts—plural—of severity. By probing what severity actually means to people, in order to find out which defensible ethical theories might support these popular views, we believe that the concept can be refined in several different directions. Some of these may demand labels other than severity, once further developed. It seems unlikely, however, that all of the different intuitions people draw on when invoking the term severity will prove ethically indefensible. Furthermore, to ensure a sustainable democratic legitimacy of priority setting criteria—paramount for their successful implementation—it is crucial that a notion as familiar and quotidian as severity does not become a purely technical term above the public’s grasp. In Norway, the demand for some severity criterion has been an imperative from legislators since the eighties. Presumably, politicians and administrators sense that the priority setting guidelines should become unpalatable to patients and healthcare professionals alike unless some observance of severity’s significance to people is incorporated. The notion of severity may very well consist of one, or several, well-established although non-codified ethical accounts. Instead of forcing all good intentions onto the single notion of severity, we think a more reasonable approach is to first search for the underlying moral accounts, extract the ethical content of those accounts and see what is still left uncovered within the notion of severity. This could lead to a more well-founded, well-argued and better understood severity criterion alongside other criteria that could contain those moral intuitions presently conflated with severity.

Severity may or may not be amenable to formalisation within the HRQoL/QALY paradigm. However, that severity’s moral relevance turns on its being expressible though the QALY framework seems quite unlikely: the underlying intuition for taking severity into account in priority setting is that death and suffering matter from an ethical point of view. Moreover, a crisper understanding of severity that necessitates non-QALY concerns would also buttress the conclusion that cost effectiveness is inadequate as a sole criterion for priority setting on the grounds that it does not account for all relevant ethical considerations.

In conclusion, the wholesale abandonment of severity as a relevant attribute of illness and injury for consideration in the context of priority setting in healthcare is too hasty. However, we wholeheartedly support Hausman’s call for intensified research into severity and consider his ethical critique as an important input in an ongoing and forward-looking discussion. This discussion will now in part be carried forward by us through the research programme Severity and priority setting in health care (SEVPRI) which was recently funded for 2020–2024 by the Research Council of Norway (project number 303724), which will also focus on non-QALY aspects of severity like need, social contexts and urgency. Severity needs analysis, and a definition anchored both in ethical theory and its vernacular usus. Only then can severity’s merit for priority setting be prudently appraised. Do not despair about severity—yet.