Framing the problem

The major intellectual framing of Making Fair Choices on the Path to Universal Health Coverage is implicit in its title. The report concerns fairness and efficiency ‘on the path’ to universal coverage. In other words, the approach is to think about securing universal coverage as a journey in which the task is to guide and steer the process in some directions rather than others, avoiding ethically unacceptable routes. The ultimate goal of a comprehensive, high quality system of universal health care without financial barriers to access is seen as central, but there is a sober realisation that compromises in practice will need to be made during the progressive realisation of that ideal. The report addresses some of the complex issues involved in those compromises, and one of its major contributions is to define various ‘unacceptable trade-offs’ to guide policy makers and policy choices. Unacceptable trade-offs are ways of balancing competing objectives in a way that threatens fundamental values. For example, the report argues, it is an unacceptable trade-off to expand coverage for low and medium priority services, that is to say services, like dialysis, that deliver relatively small health gains for their cost, before establishing near universal access for high priority services, like vaccines against tuberculosis. Similarly, the report argues that it is unacceptable to expand coverage for well-off groups before doing so for worse-off groups when the costs and benefits are not vastly different, including expanding coverage for those who are already highly covered before groups with lower coverage (pp. 38–40).

As these examples show, a major value conflict underlying the thinking of the report is that between cost-effectiveness, understood as securing the greatest aggregate health gain for a given expenditure, and fairness, understood as being attentive to the distributive dimensions of health priority-setting, with priority given to relieving high burdens of disease. Although never made fully explicit, the report also seems underpinned by a strategy of seeking to secure what could be called a ‘reverse trickle-down’ effect. Rather than secure access for those currently excluded through an extension of existing coverage, the strategy seeks to give greater priority to those currently deprived of coverage as societies move towards universal coverage. If, on successive stages on the path to universal coverage, the poor and ill are advantaged little by little, the cumulative effect of this strategy, the hope seems to be, will secure universal coverage of a kind that is ethically justifiable.

At a conceptual level, the report can be seen as the confluence between two major intellectual frameworks in thinking about the justice of health care. The first is health economics, with its origins ultimately in welfare economics and public sector investment appraisal. This broadly utilitarian tradition counsels that resources are put to best use when they maximise the gains that can be achieved by any given expenditure. The second framework is contemporary, post-Rawlsian theory, in which justice is conceived in a holistic way (Scheffler, Reference Scheffler2003: 81–82). Holistic theories make the justice of individual allocations depend upon the profile of an overall distribution, for example ensuring that no one falls below a specified minimum of care relative to the average. The most influential strand of this tradition counsels that distributions are just when they secure for the least advantaged the best possible outcome. Reasoning from these two traditions leads to the familiar efficiency/equity trade-off. Despite their obvious points of disagreements, both traditions also have two important assumptions in common. First, they make the justice of individual advantages dependent on how those individual advantages fit into a broader pattern. Second, they are forward-looking theories. Justice is a matter of looking ahead; the past is nowhere.

If you define the problem of determining the fair path to universal coverage in terms of these frameworks, then Making Fair Choices does as good a job as possible. The question is whether this is how the issue should be conceived. Should we take for granted that making coverage choices is a matter of determining priorities within some overall fixed budget? And is it right to think that for public policy the path to the future starts from nowhere?

Priorities, thresholds and bounded rationality

In one meaning of the term ‘priority-setting’, it is obvious that the problem of what to include in the reimbursement package of universal coverage is a matter of priorities. In some general sense – the sense in which a categorical distinction is made between conditions that will be covered and conditions that will not be covered – we properly speak about the process as one of priority-setting. But, in practice, reimbursement decisions are threshold decisions, made at the margins of existing activity. They do not rest on a rank-order, even a partial rank-order, of all available alternatives, but on an in/out judgement as to cost-effectiveness. As such they neither presuppose nor require the determination of overall priorities. Indeed, if we take the existence of bounded rationality seriously, then no decision-maker ever could construct a list of relevant interventions rank-ordered by their degree of cost-effectiveness; by the time the list had been constructed, health technology would have moved on.

It may be argued that a complete list of pair-wise comparisons is always presupposed when making judgements about what should be covered and what should not be covered. However, this is only true if we assume that there is a unitary decision maker, or benevolent social planner, who could have control of a global health budget. However, such a supposition is fanciful. For example, different ministries will control different budgets but all with health-related elements. The health ministry may determine medicines and medical facilities, but transport or justice ministries will control budgets related to interventions to prevent road deaths or drunken driving, finance ministries the taxation of tobacco, environment ministries resources devoted to the control of air pollution and so on. This is quite apart from any decentralisation of government power that may exist in federal or highly devolved political systems. Even in mature systems of health technology appraisal, like NICE in the United Kingdom, the judgement about the coverage of any particular intervention is always made piecemeal by reference to incremental comparisons. The organisational processes of government mean that decision makers ‘satisfice’ (Simon, Reference Simon1976, Reference Simon1983) rather than optimise.

Is there an alternative to thinking of decisions about coverage in terms of comprehensive priority-setting? Consider, as a possible alternative, the WHO list of ‘essential medicines’, now in its 18th edition (World Health Organisation, 2013). For various medical conditions that list itemises the medicines that are the ‘minimum medicines needs for a basic health care system, listing the most efficacious, safe and cost-effective medicines for priority conditions’. In this way of thinking, there is no need, for example, to argue the relative merits of antituberculosis medicines over high cost treatments for certain types of cancer as does Making Fair Choices (p. 19). Those antituberculosis medicines are on the essential list and from this point of view, the expensive anti-cancer interventions are an irrelevant alternative. In making a decision about antituberculosis medicines, it does not matter whether or not the anti-cancer intervention exists or not.

If this approach is extended beyond medicines to other interventions, then Making Fair Choices itself contains evidence about what interventions might be on the ‘essential interventions’ list. They include food supplementation for children under 5 years, malaria prevention and skilled birth attendance (Ch. 3). Note also that considerations of political accountability enter at this point. Whereas it is relatively easy to know whether or not essential interventions are being supplied on a universal basis, it is hard to determine whether a particular intervention meets the condition that it ranks higher in the list of priorities than all the conceivable alternatives with which it might be compared. If political accountability is important, as Making Fair Choices (Ch. 7) argues, then requirements of comprehensibility and transparency enter into the picture, and in turn this requires avoiding the idea that behind any one decision about inclusion or exclusion from coverage is a priority ordering.

The shadows of the past

Health economics, drawing on the welfare economics from which it is constructed, is forward-looking. So too are holistic theories of justice. In both ways of thinking, legacies from the past, while they may impose constraints of feasibility, do not impose requirements of justice. Yet, no path comes from nowhere, and the claims implicit in the status quo need to be considered in any theory of justice in transition.

The issue can be seen most clearly by considering what the principles of the report implies for the treatment of existing interventions that are at odds with the priorities defined by an unacceptable trade-off. Suppose a middle income country already covers renal replacement therapy. If we took the unacceptable trade-offs approach seriously, would it follow that the renal replacement programme should be abandoned, and alternative programmes, yielding greater value, should be implemented instead? This is not simply a matter of political feasibility, although it is that (Littlejohns and Chakildou, Reference Littlejohns and Chakildou2014). After all, existing patients could be said to have a claim in justice, grounded in legitimate expectations. The same principle can be applied to potential beneficiaries who had paid into voluntary or occupational schemes on the assumption that replacement would be provided if needed. Finally, there are the claims of health service personnel who will have trained in certain specialities and who cannot, in justice, be treated as though they were simply the human resource instruments of a centrally determined plan.

And yet, if the unacceptable trade-offs do not mandate immediate disinvestment, what do they imply? One way to retain the idea of unacceptable trade-offs is to interpret the principles on which they rest as tests of reasonable rejection, rather than as direct guides to right and wrong. On this interpretation, it would not be wrong for countries at a certain level of development to reject dialysis, although, equally, it would not be wrong if they chose dialysis. By the same token, although it might be right for a country to reject a certain technology if it did not already have it, this would not imply that it should disinvest if it does have it. Such an approach would allow more than one type of choice as reasonable, rather than determining an ordering.

The fact that these decisions are not made by unitary decision makers complicates matters. In countries without universal access, sections of the population, particularly those in the public services, will be covered by private or occupational health insurance. Private and occupational insurance is mentioned at some points in the report (for example at p. 35), but the subject is treated rather as an embarrassing uncle at a wedding is treated: no one can ignore his presence, but everyone wishes he was not there. But this is to wish away the starting point of any reforms.

Thus, the Thai Universal Health Coverage scheme established in 2002, and referred to in Making Fair Choices (p. 33), complemented two existing health plans, one for government employees and one for workers in the formal sector (Tantivess et al., Reference Tantivess, V., Román, Jomkwan, Adun and Teerawattananon2012: 332). Those existing schemes will have set up rights and expectations, and the question inevitably arises as to how consistent extensions of coverage should be with existing entitlements for those already covered. Since existing programmes may not have been influenced strongly by considerations of cost-effectiveness, it could be argued that supplemental schemes should adopt their own criteria and procedures of decision making in which cost-effectiveness is more to the fore. Conversely, it could be argued that if one feature of the health care system of a country is on an established path, it would make sense to assimilate extensions of coverage to that path, even if the path is not ‘optimal’ in some sense. Thus, in Thailand, renal replacement therapy has been included in Universal Health Coverage because it was already covered by the prior existing schemes (Tantivess et al., Reference Tantivess, W., C. and Teerawattananon2013).

One of the unacceptable trade-offs that the report urges policy-makers to avoid is “first to include in the universal coverage scheme only those with the ability to pay and not include informal workers and the poor, even if such an approach would be easier” (p. 40). (I assume that this means “easier not to include informal workers and the poor”, though the phrasing is ambiguous.) But there is a fine line between ‘easier’ and ‘not feasible’. Schemes for workers in the formal sector may be feasible taking into account administrative capacity, since those outside the formal economy are those for whom it may be hard to establish contribution records or eligibility requirements. More generally, whether a ‘pro-poor’ stance is best advanced by the reverse trickle-down strategy implicit in Making Fair Choices, as against a strategy of ‘blanketing in’ to existing arrangements those currently outside is an important empirical question. Much of the historical development of universal coverage in western Europe came through a blanketing in strategy, as household members not employed in the labour market, unemployed workers, the retired and so on were included in benefit packages. Of course, we should not suppose, along with Marx (Reference Marx1976: 91), that the “country that is more developed industrially only shows, to the less developed, the image of its own future”. Instead, the paths to universal coverage may differ depending on their starting points. But, if this is so, the origins as well as the destinations of paths matter, not least as low and middle income countries face a dual burden of disease, with communicable diseases vying with non-communicable diseases for priority.