Some Key Ethical Problems in Using CEA for Health Care Coverage Decisions

1. Some Key Ethical Problems in Using CEA for Health Care Coverage Decisions Dan W. Brock Harvard University

2. Introduction • The problem: given limited resources for health care, how should they be allocated. • Macro level 1: how much to health care vs. other goods? • Macro level 2: how much to different health care needs? • Includes different diseases. • Includes different classes of patients, e.g. young vs. old. • Micro level: selecting among different patients with similar health care needs when not all can be treated. • E.g. selecting patients for organ transplantation.

3. Cont. • CEA has principally been used for Macro level 2 choices. • This is where coverage decisions are made. • Two kinds of issues at each of these levels: • What are the substantive principles by which the allocation decisions should be made? • Given that there will indeterminacies in, and disagreement about, the principles, what procedures should be used to make the choices? • Both principles and procedures are important. • I will focus on principles and the substantive issues.

4. Cont. • There are two broad goals for resource allocation—maximize the benefits from limited resources, and distribute the benefits equitably. • These goals can and do conflict. • Cost-effectiveness analysis (CEA) is the analytic method to determine what will maximize benefits. • But CEA ignores issues of equity or fairness. • So the problem is what are the equity issues. • I will not try to be systematic, but instead focus on a few central issues.

5. Cont. • Distinguish ethical issues that arise for how do a CEA, from issues that arise when use the results of the CEA for prioritization decisions. • Will discuss one of each kind of issue in some detail, and mention briefly some other examples of each kind of issue. • CEA requires a measurement of the benefits of alternative health interventions or programs. • Health benefits are gains in quality and length of life, typically measured in quality-adjusted life years (QALYs). • And CEA requires a measure of the costs. • Typically in dollars.

6. Ethical Issues in Doing a CEA

7. Mention: How should states of health and disability be evaluated? • They are typically evaluated on a zero to one scale, with zero being death and one being full health. • Values of different health states are determined using people’s preferences for the different states. • Whose preferences should be used for evaluation of health and disability states? • Normal functioning persons evaluate disability states as worse than those who suffer those states. • From false beliefs, prejudices and stereotypes about disability states.

8. Cont. • From accommodation, coping, and adjustment. • Using ‘disabled’ preferences undervalues prevention and rehabilitation. • Using ‘normal’ preferences undervalues life saving for the disabled. • Perspectives problem: neither evaluative perspective is mistaken.

9. Mention: Should QALYs be Age-Weighted? • CEAs generally reject age weighting QALYs. • Age weighting of DALYs has an ethically problematic rationale. • The social, economic and psychological dependence of the very young and old on persons in their productive years. • This values the health of persons according to their instrumental value to others.

10. Cont. • There is a fairness/fair innings rationale (Alan Williams) for a different age weighting, at least for life years. • A year of life extension has greater moral importance, the younger its recipient. • Give to those who, if not helped, will have had less of the good our resource can provide. • This implies a higher moral priority to reaching the normal lifespan than to living beyond it.

11. Mention: Should discount rates be applied to health benefits? • This is an ethical, not just an economic, issue. • No disagreement that monetary costs and benefits should be discounted—the issue is health benefits. • Why should the same health gain for an individual have less value merely because it occurs in the future? • Same size health gain for the same or for different future individuals. • Rationality requires equal concern for all periods of one’s life.

12. Cont. • Equity requires equal concern for all persons, independent of which generation they belong to. • Importance of this issue for preventive and public health programs.

13. Detail: What Costs and Benefits Should Count in the CEA? • First issue—should benefits be restricted to health benefits or include as well non health, e.g. economic, benefits? • Second issue—should only direct, or also indirect, benefits and costs be considered? • Practical importance of the issue—indirect, non health benefits can swamp the direct health benefits of health programs. • Advocates for particular health needs often appeal to indirect, non health benefits of meeting them—e.g. substance abuse.

14. Distinguishing the two issues: • Surgeon case (Kamm)—can save A or B. B is a surgeon who then will save 5 other lives. • The additional lives saved are an indirect, but health benefit. • Wisdom case—can treat A or B and cure their disease. • If treat B, will also impart great wisdom to her, that A would not get. • The wisdom seems a direct, but non health, benefit. • Practically, most non health benefits will be indirect, and vice versa. • Won’t analyze direct/indirect distinction here.

15. Restrict CEA to Health Benefits: The Separate Spheres’ View • The sphere of an activity is determined by its purpose. • For example, criminal punishment, democratic elections, social gatherings, health care. • The purpose determines the basis for distributing goods and bads. • The purpose is determined by actual purposes of participants and the social meanings of the activity. • But the purpose is plausibly constrained by the causal consequences of the activity—the purpose of health care could not plausibly be to produce great literature. • And the purposes of social activities cannot be changed at will by individual participants.

16. Could purpose of health care be health and economic benefits? Yes. • German health system in the 19th Century sought health and a productive workforce. • Would be no conceptual mistake. • Will need a moral justification for limiting the purpose to health. • We could give the activity a new name reflecting these dual purposes. • In fact, the purpose is now accepted as not health, but rather pts overall well being. • Health can be sacrificed for overall well-being.

17. Moral Significance of the Distinctions • Moral argument for considering all costs and benefits. • Indirect, non health benefits and costs are real benefits and costs. • Ignoring them has opportunity costs and will result in failing to identify the most cost effective alternative resource allocations. • We need a moral reason for ignoring them. • We often legitimately use indirect means to our ends and have multiple aims in our activities. • Why not also in the health care system?

18. Important that this is not just a problem for Consequentialists. • For example, Prioritarians give special weight to benefiting the worse off. • Are the worse off the sickest, or those with worse overall well-being? • Will come back to this later.

19. Fairness Objection. • It is unfair to favor some patients or health care needs over others merely because doing so produces indirect, non health benefits for others. • If health care needs are equal, then people have equal moral claims to have them met. • Treating working age substance abusers (group A) also benefits their employers and the economy, treating retired substance abusers (group B) does not.

20. But both groups have equal claims to have their health needs met. • It would be unfair to give preference to the working age patients on this ground. • Broome—distinguish moral claims (“a duty owed to the candidate herself for a commodity that she should have it”) vs. other moral reasons why she should get it. • Fairness is about mediating claims of individuals. • A has no claim to a resource merely because his getting it would benefit C.

21. Likewise, Surgeon has no greater claim to needed care because she would save others if treated. • Would be no unfairness to A if she did not get preference for the resource for this reason. • Need an account of what grounds claims to health care. • Common view is claims depend on the urgency of individuals health needs. • Might all things considered give preference to employed substance abusers or to the Surgeon. • But this would be because the additional benefits outweigh, but do not remove, the unfairness.

22. Kantian argument • Giving priority to group A over group B would violate the Kantian injunction against treating people solely as means. • Does not treat group A solely as means--they need treatment as much as B. • They are not disadvantaged for the sake of others without their consent. • Treats group B solely as means--gives them lower priority solely because they are not a means to economic benefits to others.

23. Pragmatic Arguments for Ignoring Non Health Benefits • Health planners and physicians are trained to evaluate health benefits and costs, not indirect non health benefits or costs. • These other benefits and costs generally are difficult, uncertain, and costly to calculate. • This increases the potential for bias, prejudices, stereotypes, and self- or group interest to affect the assessment of benefits or costs. • For example, work in the home, traditionally done usually by women, will be undervalued.

24. If we are confident of the indirect, non health effects only in some cases, it would be inconsistent and in turn unfair to selectively use them only in those cases. • These non health assessments will also be controversial and could undermine confidence in the fairness of the allocation process. • In many cases the added effort, time, and expense of gaining data on indirect, non health effects may not justify doing so.

25. Calculation of Costs in CEA • Future costs from life-saving—e.g. future medical and Social Security costs that will be incurred for those saved.. • PH panel—optional to count these. • Doing so would, for example, reduce the benefits of smoking prevention and could make it not CE. • Should saving lives have less social value if doing so incurs these economic costs?

26. Importance of Social Context and Role • It matters for three reasons. • Will affect the alternatives from which choices must be made. • Legislators allocate between health and other aims. • Health Ministers or administrators allocate between different health needs. • Physicians must choose between different patients.

27. Will affect the nature of what is to be allocated. • Legislators and Health Ministers or administrators allocate money to health care and particular health care programs. • Physicians allocate treatments to individual patients. • Will affect the professional roles and responsibilities of the allocators. • Legislators are responsible to the electorate. • Health plan administrators are responsible to plan members. • Physicians are responsible to their individual patients.

28. Administrators’ Allocations Within the Health Sector • For example, within a public or private health plan, or a hospital. • Money will be what is typically allocated, which again is a fungible good without a specific purpose. • But it could be argued that this money is still intended only for the goal of health. • The responsibilities of allocators will typically be for the health of a population, not for economic development or other ends.

29. As a pragmatic matter, governments divide up their responsibilities to different agencies. • The Health Minister’s responsibility is to promote health. • Spillover effects, positive or negative, are viewed as not his department.

30. Rough Generalization about Social Context and Role • The higher the level the macro prioritization decision, the more defensible it is to give weight to indirect non health benefits and costs. • The closer the decision is to micro level choices between individual patients, the stronger the case is for ignoring them on grounds of fairness. • An alternative--give them some, but lesser, weight than direct health benefits.

31. Ethical Issues in Using CEA for Resource Prioritization

32. Mention: the aggregation problem. • CEA puts no limits on the aggregation of different size benefits to different persons--only the aggregate benefits of different alternatives matter. • Oregon’s problem--capping teeth versus appendectomies. • Ordinary people’s priorities are typically based on one-to-one comparisons, but this ignores cost differences.

33. Cont • But aggregation of benefits is not always rejected--the case of Coby Howard. • The ethical problem is when, and for what reason, is aggregation of benefits ethically acceptable?

34. Mention: Does CEA unjustly discriminate against the disabled? • Or more generally, those with lower QL or life expectancies? • In life saving from lower life expectancies or QL in the disabled. • In QL treatment when disabilities make treatment less effective and/or more costly. • But, in each of the above the disabled have the same health need and it is because of their disability that they get lower priority.

35. Cont. • An alternative position--so long as the disabled person’s QL is acceptable to her, it should not count against her for life saving. • And, lower life expectancy from disability should not count against one for ‘significant’ life extension.

36. Detail: Priority to the Worse Off --The Three Main Issues • What are the moral reasons for giving priority to the worse off? • Who are the worse off for purposes of health care resource prioritization? • How much priority should the worse off receive?

37. Why give Priority to the Worse Off? • Concern for equality in outcomes-but prioritarianism is different from outcome egalitarianism. A B C 1 10 15 15 2 11 17 25 1 is more equal, but 2 maximizes the outcome for the worst off. • Outcome egalitarianism should be rejected—the leveling down objection. • Deontic egalitarianism—eliminate inequalities brought about by unjust actions. • But many health inequalities are not the result of injustice.

38. Commitment to equality of opportunity. • The maximizing v. the prioritarian interpretation • Priority view: ”Benefiting people matters more the worse off people are.”(Parfit) • The worse off one is, the greater relative improvement a given size health benefit will produce. • The strength of claims for health improvements are greater, the greater the undeserved health deprivation. • Contractualist reasoning—minimize individuals’ moral complaints; treat most urgent health needs first.

39. Who are the Worse off for Health Care Resource Prioritization? • The sickest or those with worse overall well-being? • Global view--units for distribution are whole human lives (Nagel). • Focuses on the poor whose overall well being tends to be worse. • Counterintuitive implications—treat the less sick poor before the sicker rich.

40. The “Separate Spheres” view—priority for health care should depend only on health needs. • A Kantian moral justification. • This gives equal weight to equal health needs. • A pragmatic policy justification. • Physicians and health policy analysts can reliably judge health needs, but not people’s overall well-being.

41. Interpreting the Separate Spheres View • Are the worse off those with overall worse health or the sickest now? • Nord’s studies on the importance of severity. • People prefer to sacrifice substantial aggregate health benefits in order to insure that the sickest are treated. • People give more weight to how bad patients’ initial health is than to the size or duration of benefit. • Severity and urgency. • Does past health count?

42. How Much Priority Should the Worse Off Receive? • Absolute priority leads to the “bottomless pit” problem. • Balancing priority to the worse off with other considerations such as degree of benefit. • The search for a principled basis for balancing. • Person trade off studies of people’s actual tradeoffs. • Fair procedures for making tradeoffs.

43. Conclusion • Illustrated (mention/detail) some ethical problems in doing or using CEA for resource prioritization. • Fair procedures may be necessary for legitimate political resolutions. • But fair procedures are not enough. • Those procedures should be informed by our best analyses of ethical issues like those I have discussed.