Research
The Ethics of Priority Setting for Health
Summary: This project aims to promote fair rationing of interventions to improve health. It has involved a three-pronged approach of conceptual analysis, survey research, and the use of decision tools to engage the public in priority setting.
| Section: | Ethics and Health Policy | ||||||||||
| Principal Investigator: | Marion Danis, M.D. | ||||||||||
| Collaborators: |
|
Background: One of the most ubiquitous and intractable dilemmas in bioethics relates to the question of how to apportion resources fairly when there are insufficient funds to guarantee the health of everyone. Various theories of distributive justice offer competing solutions to this problem. Application of these theories in practice has barely been examined. The scholarship of the department regarding the ethics of priority setting aims to make progress on the development of fair and practical solutions to priority setting.
Several assumptions underlie our research. First, we assume that it is not possible to provide every worthwhile intervention to everyone and that priority setting is therefore necessary. Second, we assume that the complexity of health care delivery necessitates examination of priority setting at multiple levels including the national, organizational, and provider levels along with examination of interaction among these levels. Third, we assume that the growing understanding of the socio-economic determinants of health demands that we pay attention, not only to priority setting of health care, but also to other interventions that might influence the socioeconomic factors that in turn can affect health. Finally, we assume that given the importance of health to the public, the public should play a large role in the priority setting process.
Departmental Research Initiative:
Theoretical Analyses
- Accommodation of universal need and diverse views of the good.
In one of the first books published by the faculty in the Department, Ethical Dimensions of Health Policy, both Danis and Emanuel offered arguments for priority setting in a manner that attends to the reality that the need for medical care is universal at the same time that people have diverse views of the good and how medical care should contribute to achieving the good. We suggest various strategies for arranging and funding health care that accommodates such diverse views.
- Bedside Rationing
No matter what theoretical framework is adopted for priority setting, the participation of the bedside clinician will be required to ultimately make decisions regarding the use of medical interventions for individual patients. No theory can be so fully spelled out in a way that can anticipate every necessary decision. To this end Truog, Brock, Danis and colleagues have developed a taxonomy of rationing at the bedside intended as a framework for ethical analysis. This taxonomy divides rationing decisions into three categories. First are those rationing decisions that may be justified by external constraints (such as not prescribing a potentially beneficial medication because it is not available on the hospital formulary). Second are those that may be justified by reference to clinical guidelines (as, for example, not prescribing a potentially beneficial medication because a valid guideline recommends treatment with a less expensive alternative). Third are those that are justified by individual clinical judgment (such as choosing which of two patients should be admitted into the last ICU bed, in the absence of any evidence-based guidance). Decisions made on the basis of clinical judgment deserve particular scrutiny, since they may mask unethical prejudices or bias.
- Critique of Accountability for Reasonableness
In the absence of clearly determinative substantive theories of rationing, Norman Daniels has argued for procedural justice as a fair and feasible approach to rationing. He and James Sabin have suggested that accountability for reasonableness requires four criteria: 1) rationales must be publicly accessible; 2) rationale will be reasonable if it appeals to evidence, reasons, and principles that are accepted as relevant by fair-minded people; 3) there must be mechanisms for challenge and dispute resolution; and 4) there is either voluntary or public regulation of the process to ensure that conditions 1 through 3 are met. While Alex Friedman has served as a fellow in the department he has written a critique of the ‘reasonableness’ criterion, arguing that this use of such a baseline criterion serves to prematurely foreclose debate and leave those with minority views no opportunity to make the case for their point of view. In addition, while Sabik was a fellow, she and Lie developed an argument in a paper currently under revision for the Hastings Center Report, that fulfilling the criteria of Daniels and Sabin does not confer legitimacy on prioritization processes, nor does it lead to an acceptance of the outcome of such processes.
- Priority Setting in Unique Settings Such as Rural Communities
The allocation of resources for the health of rural populations is a particularly challenging instance of distributive justice. Danis analyzes rural priority setting in several steps: characterization of rural health and health care; consideration of the ethical implications of this characterization; application of ethical perspectives to priority setting for rural health care; and consideration of the policy implications for rural health. The analysis suggests that allocation of resources for rural communities would not necessarily be best guided by aiming for equal access to services with urban communities because their unique geography precludes the possibility that similar services will yield similar results. Rather, allocation of resources for rural communities would be best guided by aiming for services that yield parity of health status. Efforts to guarantee the health status of rural communities should take advantage of the innovations in communication and technology that are available and have been tested in various rural communities around the world.
- New Conceptual Framework for Prioritization
Lie has begun to develop a new conceptual framework for health care prioritization. Currently there are two dominant approaches, one emphasizing a balancing of various principles of prioritization, such as health maximization and a concern for the worse off, the other emphasizing reaching agreement through a process of public deliberation. Sabik and Lie in a paper under revision for Journal of Health Politics, Policy and Law argue that both of these approaches have failed to provide frameworks that can address real policy challenges in the case of eight countries whose priority setting experiences have been examined. Jayasinghe et al have suggested that the use of DALYs to determine health priorities is inappropriate in a developing country setting, and Lie has argued that using an international human rights framework will also not resolve the prioritization problem. Lie along with his advisee, Solli, a Ph.D. student at the University of Oslo, argue that it is necessary to use explicit criteria of justice when providing sickness certification or disability status. Certifying a person as diseased or disabled is not only a matter of identifying an objective, biomedical condition in that person, but controversially also of taking into account how well this person is able to cope and function in his or her environment.
- Rationing under Dire Circumstances
Under extreme circumstances such as a massive influenza epidemic, the distribution of scarce anti-flu interventions including vaccines and flu medications must be carried out rapidly and supply is likely to be exceedingly limited. In anticipation of these circumstances, in November 2005, the National Vaccine Advisory Committee(NVAC) and ACIP made unanimous recommendations to DHHS about what groups should receive priority for influenza vaccines.
Emanuel and Wertheimer have challenged these priority rankings arguing that the ethical justification for them seemed incorrect. NVAC recommended that after vaccine production workers and front-line health care providers, the young and the very old influenza should receive priority for vaccination. This was justified by the principle of saving the most lives, and reducing the most morbidity. In contrast, Emanuel and Wertheimer argued that a life cycle principle which places priority on people being able to pass through each stage of life, from childhood to old age was more defensible. In addition they weigh how much the individual and others have invested in the person's life. This investment refinement of the life cycle principle provides a different priority ranking for distributing vaccines — those 13 to 40 years of age should get priority after vaccine production workers and front line health care workers. In addition another principle would be relevant to distributing vaccine — a public order principle. So that within each “band” those people who protect public safety — policy, fire fighters, telecommuncations workers, etc. — should receive priority.
Empirical Research
- Interaction between Bedside and Systemwide Rationing
Along with theoretical analysis of the role of clinicians in bedside rationing, the department has pursued a set of studies to understand the interaction between rationing at the level of the provider and at the level of the healthcare system. In a national survey of internists in the United States, Hurst, Hull, DuVal and Danis have found that physicians rarely think about fairness as they make decisions that limit care for their patients. This triggered our subsequent interest in studying the reasoning used by clinicians as they ration and the development of strategies to promote greater attention to justice in clinician rationing at the bedside.
Hurst and Danis have conducted a survey with European colleagues, to examine the prevalence and determinants of rationing among general internists in England, Switzerland, Italy and Norway. We have found that bedside rationing is prevalent in all surveyed European countries and varies as a function of physician attitudes and resource availability. Furthermore we examined clinicians perceptions of rationing in their countries on the assumption that physicians witness the effects of national priorities and thus provide useful insights to assess these strategies and priorities. Although all surveyed countries offer universal coverage, 45.6 percent of respondents reported instances of underinsurance. Scarcity-related ethical difficulties were reported in all four systems. While 92.8 percent of respondents thought everyone in their country should have equal access to needed medical services, 58.8 percent thought access was not equal.
Design and Use of Priority Setting Tools
The department has taken the unique approach of designing decision tools to make the difficult task of priority setting more feasible. In particular, we have developed strategies for involving the public in the prioritization of health care services as well as other interventions aimed at improving health.
- Design and Use of the CHAT Exercise
The CHAT (Choosing Health Plans All Together) exercise was created by Marion Danis and Susan Goold, of the University of Michigan, to allow for the design of health insurance benefits by the public. It is a tool that aims to make complex choices understandable and tough decisions feasible through public deliberation. It has been tested by hundreds of groups and found to be easy to understand and use.
Applying the CHAT exercise, Danis, Goold, and colleagues have ascertained the priorities of diverse groups of people in publicly funded insurance programs including the Medicare Program, the Medicaid Program in California, and privately funded employer-sponsored insurance. The exercise has also been particularly useful in identifying the insurance preferences of the uninsured and defining a minimum basic benefit package that would be less costly than currently available insurance. The assumption is that it is crucial to design affordable benefit packages in order to make expansion of coverage possible.
This research has shown that the public is willing, under the right circumstances, to accept the need to set limits on covered benefits. They are willing to sacrifice some degree of choice and to assume some financial risk to insure a broad array of services. They are willing to accept restrictions such as tight provider networks, limited formularies, and limitations of interventions that go beyond repairing basic function as they design affordable health insurance. The process of deliberation yields greater understanding among group participants of the need to pool risk and design benefit packages that accommodate various medical needs.
Currently we are collaborating with researchers and policy makers outside of the United States to facilitate the design of affordable health insurance in other countries. We are participating in projects in a highly developed country — New Zealand, and in a developing country through projects in several states of India.
Less than 10 percent of the population in India has health insurance. At the same time, publicly funded health care facilities in India are not providing adequate or free care. In the absence of public and private payers in rural India, micro health insurance units, created and operated at grassroots level, have been proposed as a solution but their benefit package must be affordable by the poor. In a project funded by the European Union to strengthen the development of micro health insurance schemes, the CHAT exercise was used to identify public priorities for covered benefits within the constraints of an insurance package that is affordable to the poor. The exercise was tailored to use actuarial costs and health events based on information gathered in a household survey in India, producing choices at an annual premium per household of Rupees 500 ($11.25). We administered the exercise to residents of 17 poor communities in India, identified their benefit preferences, and judged the effectiveness of their coverage choices based on three criteria: Reimbursement criterion — mean reimbursement of services reported from the household survey; Fairness criterion — whether more expensive services reported in the household survey were more highly reimbursed; and Catastrophic coverage criterion — whether the most expensive illness events (top 10 percent) were reimbursed. Looking at the most frequently chosen packages we have found that they scored reasonably well using the selected criteria; participants select benefit packages that are likely to reduce their financial risk during illness. A second wave of participation has involved over 1,500 participants with the aim of ascertaining willingness to pay for the packages they have designed.
-
Design and Use of the REACH Exercise
The REACH (Reaching Economic Alternatives that Contribute to Health) exercise is a variation of the CHAT exercise. Creation of this exercise was prompted by evidence that health status is correlated with socioeconomic factors such as income and educational level. This reality has led to interest in designing interventions to reduce disparities between low and high income populations. The REACH exercise is intended to facilitate the prioritization of interventions in both the public and private sector.
Danis and several fellows, Lovett, Sabik and Adikes, have recently completed a project, based on the REACH exercise, which allowed low income residents of the Washington, DC–Baltimore metropolitan area to design employment benefits aimed at improving their health. From an array of possible employment benefits, participants ranked benefits in the following order of preference: healthcare, retirement, vacation, disability, training, job flexibility, family time, dependent care, monetary advise, anxiety assistance, wellness programs, housing assistance, and nutrition programs. Younger employees were the most interested in employment benefits whereas older employees were more interested in receiving more take home pay. African Americans are more interested in housing subsidies, given their residence in less safe neighborhoods. These preference among young and minority employees are prudent ones that might well allow them to use employment-benefits to improve life prospects.
Impact of Research: The book Ethical Dimensions of Health Policy has sold over 5,000 copies in the hard back edition and in November of 2005 was published in paperback edition.
The manuscript ‘Rationing in the ICU’, which was recently published in Critical Care Medicine, was used by the journal as Continuing Medical Education material so it is likely to be carefully read by many critical care practitioners.
Emanuel and Wertheimer's analysis of priority setting during pandemics seems to have led DHHS planners for influenza pandemic to re-assess the NVAC and ACIP proposal for the distribution of vaccine. The White House preparedness plan, released just days prior to publication of the paper, indicated that additional reflection was needed on the priority groupings for distributing influenza vaccine. Subsequently, the Federal Government found the issue too complex and relegated decisions about priority groups to the states. In addition, many public health officials have contacted us to discuss alternative ethical frameworks for thinking about public health interventions other than the traditional utilitarian “save the most lives” principle.
Findings of the European survey of bedside rationing have been presented at several professional meetings (Society of General Internal Medicine and the International Society for Priorities in Health Care). It was presented at Newcastle to the UK Clinical Ethics Network 5th Annual Conference, where the panel also included representatives from the National Institute for Clinical Excellence.
Our book chapter, “The reality of physician bedside rationing in four European countries: results from an empirical study,” in Zimmermann-Acklin M., Rationierung im Gesundheitswesen, will be published in German and will be available to contribute to the discussion in Germany. Stella Reiter-Theil is planning a collaboration in Germany to gather comparative data using the same survey tool that we used in our European survey. We have recently been asked by German medical ethicists to mentor a fellow in a project to design explicit guidelines for rationing selected interventions in the coronary and intensive care units in Germany.
The Swiss Academy of Medical Sciences (SAMS) working group on rationing in health care is reviewing our survey data as it writes ethical guidelines regarding rationing in health care.
Our study served as a basis for our current “Justice at the Bedside” project, which is funded by a grant from the Swiss National Science Foundation to study the concept of equity in clinical practice.
The CHAT exercise has been recognized as a valuable tool for educating and engaging the public in priority setting; the exercise and related publications have won awards including the Foundation for Accountability Paul Ellwood Award and the Ehrenreich prize in Healthcare Ethics. The exercise has become widely known and has been licensed to over 100 groups. It is being used as a teaching tool in medical schools, It is being used in health policy forums and among health insurers such as several state Blue Cross Blue Shield programs to promote discussion of new ways to design insurance benefits.
The California Health Care Foundation has funded three projects using the exercise: one to engage employees in both the public and private sector in health insurance coverage decisions; one to study the coverage priorities of disabled Medi-Cal enrollees as the California legislature considers funding cuts to its Medi-Cal; and one project to ask Californians to decide on a basic benefit package that would be the minimum acceptable package to cover the uninsured.
The CHAT exercise is being used by several communities in the United States to develop low cost strategies to extend insurance coverage to the uninsured including Galveston, Texas and Pueblo, Colorado. An insurance provider in Northern California is now designing a low cost benefit package based on the basic package selected in California.
The exercise is now in use in the international arena. New Zealand is beginning to use the exercise to make decisions in one of its district health boards. Micro insurance schemes in India, whose members have participated in the most recent wave of CHAT groups, will be revising current insurance coverage based on the results.
Future Research Initiatives: We anticipate new initiatives in each of the three major prongs of our work:
Theoretical Analyses
- Tackling the Problem of Priority Setting Across Societal Sectors
As we begin to consider interventions outside of traditional healthcare sector to improve the health status, the task of prioritizing interventions within and beyond the healthcare sector arises. For poor and low-income populations that receive marginal healthcare and also need interventions to improve education, housing and employment, what is the best combination of interventions? Interventions in other sectors such as education are valuable not only because they likely to enhance health but also because they achieve other valuable goals. Lovett and Danis will explore an approach to this problem.
- Exploring Approaches to Overcoming the Intransitivity Problem
One of the more difficult theoretical concerns for ethicists interested in priority setting is the concern that rankings must be internally consistent. Logic demands that if intervention A is more useful than intervention B, and B is more useful than C, then if must follow that A is more useful than C. Yet there are scenarios in which the latter is not the case. We will explore ways to address this conundrum.
- Comparative Analysis of Accountability for Reasonableness vs. Deliberative Approach to Priority Setting
While Daniels has argued for a procedural approach to priority setting, he has given insufficient attention to the role of the public in this process. We will examine normative and instrumental reasons for involving the public in priority setting.
Qualitative Research of Bedside Rationing
In ongoing collaboration with Dr. Samia Hurst and colleagues in Switzerland, we are exploring the reasoning of clinicians as they engage in bedside rationing. In a collaborative project with clinicians in Germany, a strategy will be developed for rationing interventions in the coronary and intensive care.
Further Application of Priority Setting Tools
- Further Translation of Results of CHAT Deliberation into Actual Insurance Design
We anticipate projects in which the results of CHAT exercises in India will be used to develop micro-health insurance packages.
- Use of the REACH exercise to design community-based public policies to reduce health disparities
In collaboration with the National Center for Minority Health and Health Disparities and the Mayor’s Office of Health Policy we will be ascertaining the priorities of residents of Washington, D.C. regarding public programs aimed at improving their health.
Publications:
Danis M. Health policy, vulnerability, and vulnerable populations. In Ethical Dimensions of Health Policy. Danis M, Clancy C, Churchill LR. Eds. Oxford University Press, New York, 2002.
Emanuel E. Patient v. population: Resolving the ethical dilemmas posed by treating patients as members of populations. In Ethical Dimensions of Health Policy. Danis M, Clancy C, Churchill LR. Eds. Oxford University Press, New York, 2002.
Danis M, Biddle A, Goold SD. Insurance benefit preferences of the low-income uninsured. J Gen Intern Med. 2002;17:125-133.
Jayasinghe S, Mendis N, Lie R. Use of disability adjusted life years in health planning – a plea for caution. Ceylon Medical Journal. 2002;47:61-63
Danis M, Biddle AK, Goold SD. Enrollees choose priorities for Medicare. The Gerontologist. 2004;44:58-67.
Goold SD, Biddle AK, Danis M. Will insured citizens give up benefit coverage to include the uninsured? J Gen Intern Med. 2004;19:868-874.
Lie RK. Health, human rights and mobilization of resources for health. International Health and Human Rights Journal. 2004; 4:4.
Hurst SA, Hull SC, DuVal G, Danis M. Physicians’ Responses to Resource Constraint. Archives of Intern Med. 2005 Mar 28;165(6):639-44.
Danis M, Benavides E, Nowak M, Goold SD. Development and Evaluation of a Computer Decision Exercise for Consumer Participation in Insurance Benefit Planning. Forum for Family and Consumer Issues http://www.ces.ncsu.edu/depts/fcs/pub/ltc.html. (2005, Vol 10, No 2)
Goold SD, Biddle SK, Klipp G, Hall CN, Danis M. Choosing Healthplans All Together: A deliberative exercise for allocating limited health care resources. Journal of Health Politics, Policy, and Law. 2005;30:563-601.
Emanuel EJ, Wertheimer A. Who should get influenza vaccine when not all can? Science. 2005;312:854-55.
Solli HM, da Silva AB, Lie, RK, Bruusgaard D. Biomedisinsk sykdomsmodell og materielle kriterier for rettferdig fordeling av uførepensjonssaker [Biomedical model of disease and criteria of distributive justice in disability pension cases]. Tidsskrift for den norske laegeforening [Journal of the Norwegian Medical Association] 2005;125:3293-6.
Danis M, Ginsburg M, Goold SD. The Coverage Priorities of Disabled Adult Medi-Cal Beneficiaries. Journal of Health Care for the Poor and Underserved. 2006;17:592-607.
Ginsburg M, Goold SD, Danis M. (De)constructing “Basic”: Consumers Define The Core Elements Of Coverage. Health Affairs. 2006 (In Press).
Hurst SA, Slowther AM, Forde R, Pegoraro R, Reiter-Theil S, Perrier A, Elizabeth Garrett-Mayer E, Danis M. Prevalence and determinants of physician bedside rationing: Data from Europe. J Gen Intern Med. 2006;21:1138-1142.
Kantner L, Goold SD, Nowak M, Monroe-Gatrell L, Danis M. Web tool for health insurance design by small groups: Usability study. Proceedings of the SIGCHI Conference on Human factors in Computing Systems, 2006. Available at www.acm.org/dl.
Truog R, Brock DW, Cook DJ, Danis M, Luce JM, Rubenfeld GD, Levy MM, Rationing in the ICU. Critical Care Medicine. 2006 Apr;34(4):958-63.
Dror D, Koren R, Ost A, Binnendijk E, Vellakkal S, Danis M. Health insurance benefit packages that low income clients in India prioritize: Three criteria to estimate effectiveness of choice. Social Science and Medicine. 2006 (In Press).
Danis M, Lovett F, Sabik L, Adikes K, Cheng G, Aomo T. Low-income employees choose employment benefits aimed at improving the socioeconomic determinants of health. American Journal of Public Health. 2007 (In Press).
Danis M. Ethics of Allocating Resources toward Rural Health and Health Care in Rural Health Care Ethics, Ed. Klugman C. Johns Hopkins University Press. (In Press).
Hurst S, Danis M. A framework for rationing by clinical judgment. Kennedy Institute of Ethics Journal. 2007 (In Press).
