HealthcarePapers

HealthcarePapers 8(Sp) September 2007 : 12-26.doi:10.12927/hcpap.2007.19216
Equity

Equity of What in Healthcare? Why the Traditional Answers Don't Help Policy - and What to Do in the Future

Anthony J. Culyer

Abstract

There are many deep philosophical issues regarding equity that I will slide over in order to address some practicalities of equity policy (see, for deeper material, Olsen 1997; Wikler and Murray forthcoming). However, I do want to try to link theory and policy rather than keep them in their usual silos. This is a dangerous plan. My amateur ethics will strike serious philosophers as gravely deficient, while my amateur policy strategizing will strike decision-makers as distantly up in the clouds. However, in the spirit of "nothing ventured …" I am going to try to link the two more directly than is usual. One reason for doing this is that, if we cannot discuss ethics explicitly as a foundation of policies for equity in health and healthcare policy, then I doubt we can do it anywhere else. A second reason is that I think there is a chance, if we can be more explicit about our ethics, that we might manage to translate them into policy action in reasonable and doable ways. Another reason is that I am fairly confident that the reasonable and doable ways will be different from the current ways. A fourth is that leaving the ethics largely implicit means that the huge differences between us that might otherwise remain submerged could become underwater reefs with the potential to rip the bottoms out of well-meaning policies for equity in practice - as soon as it becomes clear that one person's notion of equity is not also another's. We need to be clearer about what we mean and where we might differ in what we mean. So I shall say what I mean and what I think we all ought to agree to mean and then, from that, say what follows for policy.

In this paper I assume that we are discussing equity at the highest level of policy - distributive fairness in healthcare, its financing and the terms of access to it. I am also assuming that we care about distributive fairness in terms both of the outcomes of our decision-making processes and of the processes themselves.

I chose my title in order to provoke rethinking of some conventional ethical platitudes. I am not against platitudes in general - just these ones - and, indeed, I'll be offering some of my own shortly. The "answers" I am describing as "traditional," at least among health service researchers, are these:

Healthcare ought to be allocated in proportion to a person's need

together with its group or regional geographic companion:

Geographical allocation of healthcare resources (generally, purchasing budgets) ought to be allocated in proportion to the population's need in each area

and this:

Access/utilization of healthcare ought to be equal for all members of society

and this:

Equity and efficiency in health and healthcare usually conflict and, when they do, equity trumps efficiency.

These slogans do not help policy for four main reasons:

  • They are not good ethics.
  •  
  • Even if they were better ethics, they would still be confused and confusing.
  •  
  • Following these precepts can easily generate situations that we would all agree are more inequitable than what we have now.
  •  
  • The principles are not practical - it is unclear what policy steps follow for those who wish to embody them in practical actions such as measuring the size of a problem, the outcomes of doing something about it or managing a process intended to deliver a solution.

I have already argued against the bogus claim that there is conflict between equity and efficiency (Culyer 2006) and therefore I do not propose to discuss that topic here. However, there does exist a big conflict - or rather a whole suite of conflicts - between rival notions of equity. These rival notions hardly ever receive explicit discussion in policy frameworks. The common presupposition that equity in general trumps efficiency is a considerable irritant and dealing with it distracts attention from the more important trade-offs. Of course, outrageous inequity might rightly dominate any concern we may have about mild inefficiency; however, I do not think that concern about mild inequity ought to dominate over outrageous inefficiency. To try to convince you that I am right about the status of the other slogans, let us go back to first principles.

Equity as Fairness

It seems attractive to treat equity as a matter of fairness.1 It pervades all aspects of health and healthcare. It is significant at a high level of resource allocation (what is a fair distribution of money to Local Health Integration Networks [LHINs]?) and at the individual level (is it fair that Canadians Phyllis Thomas and Gladys Lawless have each lived with rheumatoid arthritis for more than 20 years and reside only two kilometres apart, divided by a provincial border, yet Gladys receives etanercept [Enbrel] via medicare virtually free while Phyllis would have had to find around $20,000 a year from her own pocket and so has gone without? [Abraham 2004]). Is it fair that Dr. Putter has closed his office this afternoon to play golf (it was all right for those who could easily get there in mornings but no good for me who could get there only in the afternoon)? It applies not only to individuals like Phyllis and Gladys but also to groups of individuals (is it fair that the infant mortality rate in Nunavut in 2004 was 16.1 per 1,000 live births compared with 4.3 in the two best Canadian provinces: New Brunswick and Prince Edward Island? [Statistics Canada 2007]). It applies to the outcomes of processes and also to the processes themselves (is it fair that the well-to-do on average have longer general practitioner [GP] interviews under medicare than poorer, less-well-educated people?). It applies to healthcare financing as well as healthcare delivery (is it fair that many employed Canadians get subsidized insurance for drug bills but other citizens, similar in all other respects, do not?). It applies in not-so-obvious policy choices (should the benefit accruing to people with chronic, disabling and painful conditions be valued the same as a similar benefit accruing to someone without those disadvantages when making formulary decisions or, more generally, deciding what services shall be available?). It applies at the margin of what is available as well as at the totality (should people who are willing to pay be able to purchase drugs that are judged to be insufficiently effective to be made available in public programs?). It raises questions about the similarities and differences between people having different ethnic, religious and linguistic characteristics or living at different levels of prosperity and in different locations (which similarities and differences matter and which do not?). Health equity is also everyone's business, not just that of the Ministry of Health and Long-Term Care (MOHLTC). After all, many of the key determinants of health lie well outside the MOHLTC's remit.

Health or Healthcare?

Underlying all issues regarding equity in health are a distinction and a concern. The distinction is between health and healthcare: they are not the same and, in general, the latter is there to improve the former. Healthcare is not an end in itself; health is. In particular, there is no reason to expect that equality in healthcare will generate equality in health. The concern arises from the fact that wealth and health are inversely related. As we all know, there is a social-class gradient: in the case of almost every disease, the higher the socio-economic group to which you belong, the longer your life expectancy and the better your health state at each stage of life.2 This generally means that those who are most in need of healthcare are also those who are worst placed to buy it in the marketplace either directly or through insurance. To all the other concerns about equity, therefore, we need to add a concern that the financing of healthcare is fair; unfair financing both enhances any existing unfairness in the distribution of health and compounds it by making the poor multiply deprived. However, it is not just a question of rich vs poor. The gradient implies that at every socio-economic level those further down the ladder die sooner and suffer more ill health than those immediately above them. Although this argument suggests strongly that the distribution of the costs of healthcare financing interacts with the distribution of health, I shall set aside issues of financial equity here (a good empirical discussion of the subject can be found in van Doorslaer et al. 1999).

Efficiency and Equity

Why don't we come clean about the reasons why equity in health - and healthcare - is of deep ethical concern? At root, I suggest that there are two principal aspects that demand our attention. They are sometimes incorrectly seen as being in conflict and they both have a common grounding in ethical importance. One is the principle that says, "more health is a good thing, ethically speaking." The other is the principle that says, "fairly distributed health is a good thing." Ethically speaking, the first underlies the rationale for evidence-informed practice. It is an efficiency argument: we should get the most we can out of our limited healthcare resources. The second underlies, ethically speaking, most issues in decisions about the allocation of resources to defined groups: classically regions, although territorial distributive fairness is far from being the only dimension that challenges us here. Suppose that a given expenditure could generate a gain in health for downtown dropouts equal to that generated for prosperous dwellers in leafy suburbs. Who ought to get it? Many (though not all) would say the downtown street people. And many would go further and say that the issue of justice trumps the issue of fairness so that the street people still ought to get the resource even if it would generate more health gain for the suburbanites. I suggest that this is, as a general presumption, wrong.

My argument is necessarily abstract. Consider the proposition that "more health is a good thing, ethically speaking." The primitive ethical proposition is the Aristotelian one that the ultimate human goal for which our society might aim is to be a society of flourishing individuals.3 I will not define "flourishing"; however, I mean it to imply something more than the enjoyment of mere goods and services or the economic-cum-utilitarian satisfaction of "preferences" often termed "welfarism" (e.g., Sen 1977; Boadway and Bruce 1984). I also do not wish flourishing to be restricted to the Aristotelian notion of an active life ruled by reason. This flourishing postulate is plainly a social value judgement, and if you and I differ fundamentally on it we are unlikely to agree on what is to follow. The next proposition is not a value judgement; it is factual. There is a range of concepts of "flourishing," all of which have in common that (a) they are ethically compelling and (b) they require - or usually require - good health for their full realization. Note the two factual and empirically rebuttable statements here: an assertion about there being a range of persuasive meanings for flourishing and an assertion that good health is a necessary condition for having a flourishing life. Now add a third factual assertion: healthcare is one of the means through which health is promoted.

So the extended syllogism goes like this: Flourishing lives are the ultimate good (a social value judgement). Good health is necessary for one to have a flourishing life (an empirically rebuttable statement, given an acceptable concept of flourishing). This factual proposition is true for a range of concepts of flourishing (another empirically rebuttable statement, because we can ask people what they have in mind by flourishing). Healthcare is often a necessary condition for health (yet another empirically rebuttable statement). From this combination of ethical and factual propositions comes the deep ethical significance of arrangements for the finance and delivery of healthcare (see, e.g., Culyer 1997, 2001). In general, if it is ethically good to flourish, it becomes good to have the things, such as health, that contribute to flourishing. And if it is ethically good to have health, it is good to have healthcare. And if, moreover, so fundamental a characteristic as health ought in principle to be equally experienced, then ought not healthcare to be distributed so as to bring this about? Enter the "fairly distributed health is a good thing" proposition.

From here it takes but two further steps to get closer to the policy issues that motivate this discussion. First, if it is good to encourage flourishing it is also good to be efficient at it. For example, if we were using more healthcare resources than were necessary to achieve a given health gain, that would be inefficient. Resources devoted to morally compelling causes ought to be used so as to have maximum impact on the cause served - in this case, health (Culyer 1992). Second, the burden of proof lies with those who would depart from equality regarding opportunities to flourish. Because health is necessary for flourishing, the burden of proof lies with those who wish to depart from a presumption that so necessary a human characteristic ought to be equally distributed.

A powerful implication of this line of thought on the efficiency side is that healthcare that does not contribute to health has no place in the system. It also implies that cost-ineffective healthcare has no place in such a system - even if it is effective - because providing cost-ineffective care would imply that resources that could be put to achieving better health for at least one person were in fact being put to no apparent use at all. In short, the ethical reasons for caring about the distribution of health are also reasons for caring about the efficient production of health. It is insufficiently recognized that the case for cost-effectiveness is, at root, an ethical case.

Another implication, this time on the distributional side, is that inequalities in health ought not to be manufactured without compelling reasons and ought not to be allowed to continue if they can be removed using reasonable means. By "compelling reasons" I mean countervailing ethical arguments that carry moral weight. By "reasonable means" I mean actions and policies that do not have costs or undesirable downstream consequences that might outweigh their equitable gain. Pursuing greater equality of health does not always imply, however, that we must pursue greater equality of healthcare or access to it, or that we should match it to need.

Let me give an illustration. Figure 1 shows the quality-adjusted life years (QALYs) to be had from spending a given sum on healthcare for the poor and the rich in a given community. The light-shaded bars indicate the existing expectation of QALYs for an average person, aged 50, in each of these equal-sized groups. For the rich, QALYs are twice that of the poor. The best estimates indicate that, if the given sum were spent entirely on the poor, their expected health gain (in QALYs) would be 3 while the gain to the rich, if the sum were spent on them, would be 2. Supposing one were forced to choose which is the better way of spending the sum? You might say funds would be better spent on the poor because the gain would be 3 compared to 2, and this approach is therefore more efficient. However, that conclusion would be wrong because it entails an assumption that a QALY gain for the poor counts the same as a QALY gain for the rich. In fact, both dark bars indicate efficiency because we assume that maximum health gain is to be had for each group from spending the sum on them. Both are efficient.

To answer the fair distribution question one needs to make explicit interpersonal comparisons. There are three obvious ways of doing this in the above example. One could say that a health gain is of equal value to whomever gets it (in effect, we do not care if the recipient is either poor or rich). In that case, 3 outweighs 2 and the resource goes to the poor. This is straightforward QALY maximization coupled with the distributional value judgement that all QALYs are equal. Another possibility would be to say that QALYs received by people who have low expectations of future QALYs (for whatever reason, including that they are poor) are to be more highly valued than those going to others. In that case, the argument for spending the sum on the poor is even stronger. Third, one could say that the fair distribution is a more equal one. In that case, spending the money on the poor generates a 13:20 distribution, which is plainly more equal than 10:20 (let alone 10:22). In this example, all three distributional arguments go in favour of the poor. But both of the possible new distributions (13:20 or 10:22) are efficient and the test of fairness is not the relative sizes of potential health gain or the initial distribution of health, but the final (expected) distribution of health.

Equity vs Equality

Equity is not the same as equality, although they are often carelessly taken to be the same. However, they are connected. Equity often involves the equality of something. The critical question is "equality of what?" But sometimes equity also deals with just inequalities, and the question then is "what is the criterion for deciding which inequalities are fair or unfair?" Equity means treating likes alike and unalikes appropriately differently. Equity requires not only that relevantly similar cases be treated in similar ways but also that relevantly different cases be treated in different ways. These two concepts are as old as Aristotle and are known as horizontal and vertical equity:

  • Horizontal equity: The equal treatment of people who are equal in a relevant respect
  •  
  • Vertical equity: The unequal treatment of people who are unequal in a relevant respect

By "treatment" I shall follow the convention of talking about healthcare resources generally denominated in terms of dollars. But what might be the "relevant" respects? There are seven commonly adopted ones. I shall state each respect and the principle to which it seems to be connected, and then make some comments on it.

Rival Relevant Respects

 

  • Need: Populations with equal needs should receive equal treatment and populations with greater needs should receive more favourable treatment.
    A disadvantage of this principle is that it is far from clear what "need" means. It might mean one or more of the "respects" that follow.
  •  
  • Ill health: Populations that are equally ill ought to be treated the same; those that are sickest ought to get more.
    A disadvantage of this principle is that it seems to assume what might not be the case - that the conditions in question are effectively treatable by healthcare and that all conditions are equally costly to treat. Unfortunately, the effectiveness of healthcare can vary widely (in cases of iatrogenesis, for example, it is negative4). It surely cannot make much sense to require a population to have the same amount regardless of their morbidity characteristics, the effectiveness of relevant preventive and restorative medical care and the cost, whether high or low, of delivering that care.
  •  
  • Desert: Populations of equal desert ought to be treated the same and those of greater desert ought to receive more.
    Common elements that advocates of this view have in mind are lifestyle choices (e.g., smoking, drug abuse, poor diet, dangerous sports, careless and promiscuous sex) that increase the chances that someone will need healthcare and, moreover, that might reduce the chances that the care will be effective. These are mutually reinforcing grounds for giving such individuals and groups a low priority. This view suffers from the problem that it is virtually impossible empirically to distinguish lifestyle effects from other effects, that it assumes that lifestyle differences are avoidable, not socially conditioned and, if deleterious to health, that the patients in question are culpable. Another, more positive, argument holds that groups with higher productivity (e.g., people with higher earnings or more dependent children or who do more public service work) deserve a higher priority. This argument suffers from the problem that the claim of desert rests heavily on a claimed contribution to the welfare of other people, which is hard to measure without arbitrariness and is, at best, a partial measure of deservingness.
  •  
  • Resources themselves: This is usually presented as a purely horizontal equity argument - since all people are fundamentally to be regarded as equal, each ought to have equally available resources; the per capita distribution ought to be everywhere the same in a jurisdiction.
    A disadvantage of this principle is that, like the previous one, it ignores the productivity of resources. It is difficult to see why, for example, there should be any concern for the equitable distribution of ineffective care or why people whose needs are different ought to have the same care.
  •  
  • Capacity to benefit5: People with equal ability to benefit from healthcare ought to be treated the same and those with high capacities to benefit ought to receive more.
    This principle addresses the productivity issue. However, if it turns out that populations with the greatest ability to benefit are normally also initially relatively healthy then the application of the principle will lead to greater health inequalities and, because such people are also likely to be relatively wealthy, they will be made still healthier as well.
  •  
  • Health: This principle aims at greater equality of health not, usually, through reducing anyone's health but by giving priority to those with relatively low health or who are furthest from the average.
    A disadvantage of this principle is that it might imply the use of enormous amounts of resources for the very sick (but for whom medical care is not at all effective), resources that would generate much greater health gains if others were to receive them.
  •  
  • Equality of access: This principle is perhaps the most frequently encountered type of equity in healthcare.
    A disadvantage of this principle is that it can be satisfied at very high levels of cost of accessing - just so long as they are equal (e.g., an equal $1,000 co-payment each per GP visit).

Things a priori and Algorithmic

I prefer "health" as the distribuendum compared to any of the other candidates for being a relevant respect. One - and only one - of seven candidates really addresses the heart of the ethical problem. None of the others is even a reliable tracker of health and it is easy to conceive of occasions when there might be a considerable divergence between them. For example, to use current ill health as a driver (inverse, of course) for healthcare resource allocation will, in situations where healthcare is of no avail, cause an unambiguous waste of resources - resources that could have been used to improve the health of those with poorest health whose condition can be improved through healthcare. The principle leads to both inefficiency and increased inequity. Despite this, current morbidity and mortality are two of the most frequently met arguments of resource allocation formulae in all jurisdictions. For much the same reason, the so-called "burden of disease" is a poor indicator of the likely productivity of research (Mooney and Wiseman 2000).

It is health inequality that is inequitable, not inequality of healthcare. We therefore need quantitative and qualitative measures of health outcomes to determine the fair distribution of purchasing budgets and the extent to which the current distribution falls short of the ideal, just as we need them to make comparisons among interventions in health technology assessment (HTA). In HTA, a generic outcome measure is needed so that one can make comparisons across technologies of different types (e.g., drugs, imaging and other diagnostic aids, devices, surgical procedures). In equity policy, one likewise needs to make systematic comparisons - in this case, not between technologies but across population groups.

t must be recognized that the selection of the dimensions of any generic measure entails social value judgements, as do their scaling and combining. It is natural to reach for a formula, and there are lots of candidates. Formulaic or algorithmic approaches to health outcome measurement have many advantages, provided the variables embodied in them have sufficient construct validity and provided they are applied in an appropriate context. These advantages include their transparency, the fact that once their construction has been completed the business of using them is relatively straightforward and low cost and the fact that, used in appropriate contexts, they usually deliver precise solutions (e.g., a specific incremental [or average] cost-effectiveness ratio or a given budget allocation to a given LHIN). Until one of these constructs, or a satisfactory substitute, is chosen, "health" ought always to be surrounded by scare quotes.6 However, once selected and despite their virtues, these algorithmic approaches are not enough.

Things Deliberative

Judging the impact that a changing resource pattern has on health requires multiple skills and the exercise of judgement. A scientific clinical epidemiological knowledge base will sometimes be available. If it is, this is what Lomas et al. (2005) have called "context-free scientific evidence." But its use requires interpretation and judgement - practical clinical voices capable of expressing professional opinions about the applicability of the scientific knowledge in the social and professional contexts to hand - together with any available context-sensitive scientific research. One also needs economic estimates of what health outcomes might be achievable from different levels of resourcing, as well as the evidence and experience of social scientists that relate to the possibly distinctive cultural and ethnic circumstances that might affect the productivity of various ways of deploying resources. The thoughtful integration of this knowledge also requires social value judgements to be made and, to give the process credibility in the public imagination, probably some lay participation too.

The decision-making process will almost certainly also involve what Lomas et al. (2005) have called "colloquial" evidence: evidence that is not scientific at all, but professional recollections, experience, case studies and other knowledge that, although scientifically weak, might be all there is on a particular aspect of a problem. Sifting this evidence cannot be done using only an algorithmic approach. Here the essence of the problem is that the knowledge needed to determine equitable distribution is incomplete and fragmented across disciplines, medical specialties and professions. It is also (probably) controversial and it cannot escape being intimately interwoven with values and the making of interpersonal comparisons of benefit and cost. It therefore requires synthesis, quality assessment, discussion of its relevance and applicability in the context of proposed application, the calling and interrogation of experts, the explicit confronting of possible trade-offs, the possibility of decision- makers changing their minds during the course of the deliberations as new knowledge is acquired and the making of an overall judgement informed, but not determined solely, by the evidence.

This is what I call a deliberative process. It is founded on the propositions that the facts do not speak for themselves, that decisions can never be solely evidence-based and the (un-evidenced) belief that evidence-informed decisions - using whatever is available - are better decisions.7

Health

It is not possible to have a practical policy about equity in health without a measure of it. It is ludicrous that we are celebrating 50+ years of equitable medicare in Canada and most of the developed world and we still have no proper measure. Mortality data will not do: they tell us about the numbers of dead people but nothing of the quality of life of the living. In the United Kingdom, National Health Service (NHS) hospital data used to include a throughput measure called "deaths and discharges" - as though the difference did not matter. Choosing an appropriate measure of health is not, however, a matter requiring us to strain at gnats while swallowing camels. The literature on health measurement has attention-riveting properties for health service researchers delving into the minutiae of measurement methodology. These experts have developed a welter of candidates for the role of health-as-an-outcomes measure (e.g., Assessment Quality of Life; DALYs; DASH; EuroQol [EQ-5D], Health Utilities Index; Healthy Year Equivalents; QALYs; short-form health surveys such as SF-6D, SF-8, SF-12, SF-36).

The need in empirical equity policy is for a practical, low-cost instrument that has reasonable construct validity - i.e., one that takes account of the most important dimensions of population-level health - and is as sensitive as it needs to be (and no more). My own inclination is to select the EQ-5D, using Canadian weights, on grounds of simplicity, ease of use and its having well-understood virtues and vices - so that, should the latter prove to be important in any particular context, the need for a considered judgement that goes beyond the QALY becomes clear. The EQ-5D essentially interprets health in terms of five dimensions: mobility, ability to self-care, ability to perform usual activities of daily living, level of pain/discomfort and level of anxiety/depression. These are scored and combined using weights derived from the populations whose health is being measured. So that is an algorithm we need - or at least something like it.

Combining an Algorithm and Other Elements in Deliberative Processes

The act of using an algorithm requires two important further steps that are not themselves well suited to algorithmic solution. The first addresses the way in which algorithmic measures are to be combined, not across attributes of health as discussed before but across groups of people: young or old, male or female, different ethnicities, different geographical locations, different histories of chronic or congenital disease and disability and so on. How health is combined across people amounts to determining the weights attaching to those with disadvantages or other vertical equity claims for favourable treatment compared with others. Only if there are no ethically relevant differences between people can we assume that the issue is one of horizontal equity and that a QALY=QALY=QALY, whoever gets it.

Making interpersonal comparisons also crops up in less conspicuous ways. For example, the seemingly technical field of HTA is loaded with interpersonal value judgements, as are all decision processes that involve the measurement of individuals' health and their adding up across individuals. So is the prioritization of people's claims (e.g., on waiting lists, for treatment in a treatment room, for research into new treatments). Typical equity-related questions raised by outcome measures in HTA include the following:

  • Ought the fact that older people have shorter life expectancies than the young, and hence on average a shorter period of time in which to enjoy any benefits of healthcare, be reflected in benefit calculations? If so, how?
  •  
  • Ought the fact that some people have lived extremely painful and restricted lives for many years or have had multiple handicaps than others affect the social valuation of their respective future health benefits? If so, how?
  •  
  • Should the fact that some people might stand to make major gains in health benefits while others might gain only some reduction in the speed of their health deterioration affect the relative valuation of any additional future health benefit? If so, how?

The same issues crop up in making judgements about distributions of healthcare resources across social groups and between Ontario regions. In England and Wales, questions such as these have been put to a Citizens' Council (University of Toronto Priority Setting in Health Care Research Group 2006).

The second step is even less well suited to an algorithmic approach. This is the determination of the kinds of health-affecting interventions that would promote greater equality in the distribution of health. It would be highly desirable for the set of interventions to be taken as broader than those under the control of the MOHLTC and, in any event, to include public health interventions. Experience with attempts to develop evidence-informed formulary decisions teaches that the formal scientific knowledge base is commonly fairly unsatisfactory - the research might simply not have been done, the technologies investigated might have had policy-irrelevant comparators, the published work might be of poor general quality, it might be of high quality but unknown generalizability, it might be incomplete (e.g., with respect to long-term consequences or economic consequences of any kind) and it might be scientifically controversial. These elements are likely to be even more prominent in public health research and research on the impact of healthcare on the distribution of health. If similar resource allocations seem to produce different outcomes in different locations and between different cultural groups, then we had better understand the reasons why - and even involve those who understand the local or ethnic cultures in the decision-making process - at least as commentators or consultees8 and, possibly, as participants in the decisions themselves. Both of these issues (making appropriate interpersonal comparisons and judging the cost-effectiveness of interventions) involve the use of both algorithmic and deliberative methods. Decisions about equitable resource allocation seem to meet most of the conditions conjectured to characterize the appropriate use of deliberative methods and decisions that are "accountable for reasonableness" (Daniels 2000a, 2000b).

As reported in Culyer and Lomas (2006), a deliberative process is more likely

  • to generate guidance that is consistent with the context-free scientific evidence set in a relevant context;
  •  
  • to identify relevant clinical, social and political contexts for interpreting context-free scientific evidence;
  •  
  • to command wide credibility in professional circles and beyond;
  •  
  • to generate recommendations whose implementation will be speedy; and
  •  
  • to identify impediments to the implementation of guidance and to propose solutions.

One might also expect that the reasonableness of a process will depend upon the following:

  • The quality of chairperson
  •  
  • The clarity and openness of process The reasonableness of timelines for evidence submission and consideration
  •  
  • The use of colloquial evidence to challenge context-free evidence, set contexts and plug gaps in science (but not to supplant scientific evidence of either kind)
  •  
  • The possibility of interaction between decision-makers and non-participant stakeholders through consultation and commentary
  •  
  • The availability of time for study, discussion and reflection before, during and after meetings
  •  
  • The scope for decision-makers to request further information and take face-to-face oral evidence
  •  
  • The opportunity to appeal a decision not because an appellant disagrees with the decision but on the following grounds:

    • Decision-makers failed to act fairly and in accordance with their published procedures.
    •  
    • Their decision was perverse in the light of the evidence submitted.
    •  
    • They exceeded their powers.

In essence, I am recommending the creation of a new institution tasked with the blending of an algorithmic approach (the health measure) with a deliberative approach (determining the patterns of resource distribution to deliver changes in the outcome health indicator of choice and thereby to move the overall allocation of health in a more equal direction). The contribution of non-healthcare determinants of population health is also best considered in a deliberative process, one that requires the consent and collaboration of ministries other than the MOHLTC.

Need for New Mechanisms

The implementation of the policy for equity implied by the foregoing entails the following necessary key steps for the MOHLTC:

  • Taking a policy decision regarding the entity whose equitable distribution is the focus of concern (this distribuendum is conjectured here to be health)
  •  
  • Setting up a mechanism to select a pragmatic empirical measure (the algorithm) of health (suggested here to be EQ-5D with Canadian weights)
  •  
  • Setting up a province-wide deliberative process (e.g., a healthcare distribution commission) whose tasks would be as follows:

    • Annually to determine the allocation of non-tertiary and non-experimental personal healthcare and public healthcare resources (and, preferably, other resources affecting health) to the regional commissioners (LHINs) with as wide a range of commissioning power as possible
    •  
    • To give advice to LHINs on the intra-LHIN distribution of resources for equity between social groups
  • Setting quantitative and qualitative annual targets for the commission and the LHINs for greater equality in the distribution of health (not healthcare)

Slogans for Health Equity in Ontario

Having frowned on some common slogans purporting to be guides for policy, it is incumbent on me to suggest replacements:

All needed healthcare ought to be provided free. Healthcare that is not needed must be paid for privately.

Equity is a factor in determining resource allocation decisions only in respect of healthcare that is needed; i.e., of the healthcare that it would be technically possible to provide, only that which is (a) necessary for a person's timely health improvement and (b) cost effective may be said to be needed.

Access should be as cheap as is necessary to enable utilization of needed healthcare.

Equality of access is not specifically equitable but policies should seek to ensure that access is cheap by lowering barriers - whether financial, geographic, ethnic, cultural, linguistic or social - to service use. This is because diagnosis is a necessary condition for establishing whether there is a need for healthcare. The greater the barriers to the receipt of care, the more likely it is that genuine healthcare needs will go undetected and untreated, to the detriment of both efficiency and equity. Without cheap access, the community's need for healthcare goes unassessed. How cheap access ought to be will depend on the elasticity of demand for care and the impact of healthcare co-payments and other costs of access and use on a person's overall purchasing power (this is required if other forms of inequity are not to be generated by healthcare policy).

The main inequity is inequality of health.

Addressing other inequalities (e.g., of resources per head) is a distraction and can lead to greater health inequality.

Equity in health is impossible without an empirical measure of health.

The measure required does not have to be perfect nor suited for all decision contexts. It must, however, have construct validity and enable the making of politically acceptable comparisons between differing population groups.

Avoidable gross inequalities in health are intolerable moral outrages.

Good health is normally necessary for people to flourish as human beings. Gross inequalities in health imply gross inequalities in people's flourishing. Policy targets for reducing health inequalities should be set by the MOHLTC.

Let the largest differentials between persons and groups command the highest priority.

In seeking to promote the health of all Ontarians through cost-effective healthcare, policy should address the biggest disparities in people's lifetime experiences of health through selective resource allocation and specific policies aimed at having maximum impact on the health of the least healthy.

Unavoidable gross inequalities ought to be accompanied by generous palliative provisions and other compensating variations.

Avoidable gross inequities ought to be avoided. Although unavoidable inequalities may not be fully compensable through other policies, other policy opportunities for promoting more equal flourishing ought to be considered. This is but one policy element requiring inter-ministry collaboration.

Achieving equity in health requires a policy implementation process that is deliberative.

Achieving the equitable allocation of resources requires a combination of judgements about social values and judgements about the contribution that various interventions and types of care are likely to have on population health. Interventions ought ideally to include public health interventions and other non-healthcare determinants of population health. A deliberative process is more likely to deliver well-informed and politically acceptable decisions than other methods.

Equity in health is impossible without an information database.

A policy for the thoughtful distribution of health-affecting resources routinely requires the following:

  • Information about the current distribution of resources
  •  
  • Information about the current distribution of health across relevant social groups
  •  
  • Information about the technical potential of health and other services to improve health - incremental impact ratios of resources on health
  •  

From Talk to Action

For the past 75 years healthcare policy has rarely been discussed without reference to equity and it is undoubtedly equity that drives four of the principles of the Canada Health Act (comprehensiveness, universality, portability and accessibility). Despite this longstanding concern, Canada, both federally and provincially, along with most other jurisdictions, has failed to develop

  • clear definitions of terms such as equity and inequity;
  •  
  • routine databases for measuring inequity or inequality;
  •  
  • policy targets for achieving equity; and
  •  
  • mechanisms, beyond the health system's broad structural characteristics, for promoting greater equity.

In my contribution to this collection I have tried to present a coherent set of principles for equity in health and healthcare, together with some of the steps required to address what has hitherto been lacking. I believe these actions are in broad sympathy with the historic roots of Ontario's healthcare policy and are in tune with contemporary moves toward greater transparency and more participative policy decision-making.

About the Author(s)

Anthony J. Culyer
Visiting Professor, Department of Health Policy,
Management and Evaluation, University of Toronto
Professor Economics, University of York, UK

References

Abraham, C. 2004, September 18. "Health Talks Offer No Remedy for Disparities in Drug Policies." Globe and Mail A4.

Boadway, R. and N. Bruce. 1984. Welfare Economics. Oxford: Basil Blackwell.

Chalmers, I. 2003. "Trying to Do More Good in Policy and Practice: The Role of Rigorous, Transparent, Up-to-Date Evaluations." Annals of the American Academy of Political and Social Science 589: 22-40.

Chalmers, I. 2005. "If Evidence-Informed Policy Works in Practice, Does it Matter if it Doesn't Work in Theory?" Evidence & Policy 1: 227-42.

Culyer, A.J. 1992. "The Morality of Efficiency in Health Care - Some Uncomfortable Implications." Health Economics 1: 7-18. Rpt. in A. King, T. Hyclak, S. McMahon and R. Thornton, eds., North American Health Care Policy in the 1990s. Chichester: Wiley, 1993. 1-24.

Culyer, A.J. 1997. "The Principal Objective of the NHS Ought to Be to Maximise the Aggregate Improvement in the Health Status of the Whole Community." BMJ 314: 667-69. Rpt. in B. New, ed., Rationing: Talk and Action. London: King's Fund and BMJ, 1997. 95-100.

Culyer, A.J. 2001. "Equity - Some Theory and Its Policy Implications." Journal of Medical Ethics 27: 275-83.

Culyer, A.J. 2006. "The Bogus Conflict between Efficiency and Equity." Health Economics 14: 1155-58.

Culyer, A.J. and J. Lomas. 2006. "Deliberative Processes and Evidence-Informed Decision-Making in Health Care: Do They Work and How Might We Know?" Evidence and Policy 2: 357-71.

Daniels, N. 2000a. "Accountability for Reasonableness in Private and Public Health Insurance." In A. Coulter and C. Ham, eds., The Global Challenge of Health Care Rationing. Buckingham: Open University Press. 89-106.

Daniels, N. 2000b. "Accountability for Reasonableness." BMJ 321: 1300-01.

Lomas, J., A.J. Culyer, C. McCutcheon, L. McAuley and S. Law. 2005. Conceptualizing and Combining Evidence for Health System Guidance. Ottawa: Canadian Health Services Research Foundation.

Mooney, G. and V. Wiseman. 2000. "Burden of Disease and Priority Setting." Health Economics 9: 369-72.

Olsen, J.A. 1997. "Theories of Justice and Their Implications for Priority Setting in Health Care." Journal of Health Economics 16: 625-39.

Rawls, J. 1971. Justice as Fairness. Cambridge: Harvard University Press.

Ross, N., N. Wolfson, G.A. Kaplan, J.R. Dunn, J. Lynch and C. Sanmartin. 2006. "Income Inequality as a Determinant of Health." In J. Heymann, C. Hertzman, M.L. Barer and R.G. Evans, eds., Healthier Societies: From Analysis to Action. Oxford: Oxford University Press. 202-36.

Ruta, D., C. Mitton, A. Bate and C. Donaldson. 2007. "Programme Budgeting and Marginal Analysis: Bridging the Divide between Doctors and Managers." BMJ 330: 1501-03.

Sen, A.K. 1977. "Social Choice Theory: A Re-examination." Econometrica 45: 53-89.

Sen, A.K. 1980. "Equality of What?" In S. McMurrin, ed., The Tanner Lectures on Human Values. Cambridge: Cambridge University Press. 197-220.

Statistics Canada. 2007. "Infant Mortality Rates by Province and Territory (Both Sexes)." Ottawa: Author. Retrieved April 3, 2007. < http://www40.statcan.ca/101/cst01/health21a.htm >

University of Toronto Priority Setting in Health Care Research Group. 2006. Review of Citizen's Councils and Recommendations for the Creation of a Citizens' Council in Ontario as Mandated under the Transparent Drug System for Patients Act. Toronto: University of Toronto Department of Health Policy, Management and Evaluation.

van Doorslaer, E., A. Wagstaff, H. van der Burg, T. Christiansen, G. Citoni, R. Di Biase, U.-G. Gerdtham, M. Gerfin, L. Gross, U. Häkinnen, J. John, P. Johnson, J. Klavus, C. Lachaud, J. Lauritsen, R. Leu, B. Nolan, J. Pereira, C. Propper, F. Puffer, L. Rochaix, M. Schellhorn, G. Sundberg and O. Winkelhake. 1999. "The Redistributive Effect of Health Care Finance in Twelve OECD Countries." Journal of Health Economics 18: 291-313.

Wikler, D. and C.J.L. Murray, eds. Forthcoming. "Goodness" and "Fairness": Ethical Issues in Health Resource Allocation. Geneva: World Health Organization.

Footnotes

1 An idea most recently introduced and developed in Rawls (1971).

2 For a modern review of the evidence, see Ross et al. (2006).

3 The Greek concept of eudaimonia is often translated as "happiness," which seems to me somewhat to trivialize it. I prefer "flourishing."

4 The famous Dr. Spock advised countless thousands of mothers thus: "There are two disadvantages to a baby's sleeping on his back. If he vomits, he's more likely to choke on the vomitus. Also he tends to keep his head turned towards the same side, this may flatten the side of his head … I think it is preferable to accustom a baby to sleeping on his stomach from the start" (cited in Chalmers 2003: 23). Millions of Spock's readers followed this apparently rational, theory-based and authoritative advice. "We now know from the dramatic effects of the 'Back to Sleep' campaigns in several countries that the practice promulgated by well-intentioned experts like Spock led to tens of thousands of avoidable sudden infant deaths" (Chalmers 2005: 229).

5 Capacity to benefit is similar to Sen's (1980) idea of capabilities.

6 For the sake of readability, however, I have resisted the temptation to pepper my article with such devices.

7 An example of the sort of process I have in mind is Program Budgeting and Marginal Analysis (PBMA); see Ruta et al. (2007).

8 The difference in the practice of the National Institute for Health and Clinical Excellence (NICE) in England and Wales is that commentators are not those invited to make explicit submissions in connection with a technology appraisal: manufacturers of comparator technologies, specific agencies such as the NHS Quality Improvement Scotland, the relevant NICE National Collaborating Centre, other related research groups and "other groups where appropriate." Consultees, by contrast, can participate in the consultation on the draft scope and the other documents used in the appraisals process. Consultee organizations representing patient/carers and healthcare professionals may nominate clinical specialists and patient experts to present their personal views to the appraisal committee. All consultees are given the opportunity to appeal against the NICE conclusions before they are published.

Comments

Be the first to comment on this!

Note: Please enter a display name. Your email address will not be publically displayed