Some of these controversies do have implications for the construction—not just the application—of summary measures. One approach to addressing certain controversies about distributional choices is to survey the population to elicit the moral weights people place on resource allocation trade-offs. That is, people are not asked to assess how much different impairments would affect their own quality of life but rather how they would weigh the use of resources to prevent or relieve different impairments for different groups. For example, how would they weigh the prevention of one death against some number of cases of paraplegia prevented? Some propose that these weights could be used to construct an ethically more sensitive summary measure (Nord, 1994).

An alternative view expressed during the December workshop was that summary measures should not attempt to incorporate “equity weights.” Rather, such weights, if they were to be developed and accepted, should be applied in a separate, explicit analytic step that makes the distributive analysis and judgments more explicit and visible. In any case, when analysts use an individual’s assessments of what risks (e.g., earlier death) he or she would be willing to trade for what benefits (e.g., better quality of life) to measure preferences, they may be inappropriately treating these as statements about acceptable social (rather than personal) trade-offs involving risks and benefits to others.

Yet another perspective is that quality weights should not be used at all in measures and analyses employed in making decisions about different priorities for saving lives (Kamm, 1993). That is, in making decisions, for example, about whether to invest more in treating one disease rather than another, differences in the number of lives saved should be considered but not differences in the quality of life the or level of disability experienced by those whose lives would be saved. This proposal, which has met with mixed reactions, is one way of recognizing the concern in the disability rights movement that policy analyses should not make distinctions among lives that are still worth living.

In his background paper, Daniels suggests that these and other distributive questions form a family of unsolved rationing problems and that persistent controversy involving them may be traced to systematic moral views that differ in their distributive implications. In the face of such divergent views, Daniels claims, it is not possible to elicit meaningful weights from survey techniques, and more attention has to be paid to designing publicly accountable, fair deliberative procedures to resolve disputes about these distributive questions.

In sum, the ethical issues raised by summary measures of population health include both questions about measurement strategies and questions about the principles for rationing or distributing resources. Progress in resolving the former questions does not imply there will be agreement about the latter.

RECOMMENDATIONS

The committee’s recommendations involve three basic areas: (1) strengthening understanding of the ethical and policy implications of different measurement strategies and applications; (2) developing agreement on standards for summary measures; and (3) educating policymakers, public health officials, and clinicians about the uses and limitations of summary measures of population health. Although it may seem unusual for work on ethical implications to precede consensus on standards for measures, the committee believes that better ethical under-



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 17
Some of these controversies do have implications for the construction—not just the application—of summary measures. One approach to addressing certain controversies about distributional choices is to survey the population to elicit the moral weights people place on resource allocation trade-offs. That is, people are not asked to assess how much different impairments would affect their own quality of life but rather how they would weigh the use of resources to prevent or relieve different impairments for different groups. For example, how would they weigh the prevention of one death against some number of cases of paraplegia prevented? Some propose that these weights could be used to construct an ethically more sensitive summary measure (Nord, 1994). An alternative view expressed during the December workshop was that summary measures should not attempt to incorporate “equity weights.” Rather, such weights, if they were to be developed and accepted, should be applied in a separate, explicit analytic step that makes the distributive analysis and judgments more explicit and visible. In any case, when analysts use an individual’s assessments of what risks (e.g., earlier death) he or she would be willing to trade for what benefits (e.g., better quality of life) to measure preferences, they may be inappropriately treating these as statements about acceptable social (rather than personal) trade-offs involving risks and benefits to others. Yet another perspective is that quality weights should not be used at all in measures and analyses employed in making decisions about different priorities for saving lives (Kamm, 1993). That is, in making decisions, for example, about whether to invest more in treating one disease rather than another, differences in the number of lives saved should be considered but not differences in the quality of life the or level of disability experienced by those whose lives would be saved. This proposal, which has met with mixed reactions, is one way of recognizing the concern in the disability rights movement that policy analyses should not make distinctions among lives that are still worth living. In his background paper, Daniels suggests that these and other distributive questions form a family of unsolved rationing problems and that persistent controversy involving them may be traced to systematic moral views that differ in their distributive implications. In the face of such divergent views, Daniels claims, it is not possible to elicit meaningful weights from survey techniques, and more attention has to be paid to designing publicly accountable, fair deliberative procedures to resolve disputes about these distributive questions. In sum, the ethical issues raised by summary measures of population health include both questions about measurement strategies and questions about the principles for rationing or distributing resources. Progress in resolving the former questions does not imply there will be agreement about the latter. RECOMMENDATIONS The committee’s recommendations involve three basic areas: (1) strengthening understanding of the ethical and policy implications of different measurement strategies and applications; (2) developing agreement on standards for summary measures; and (3) educating policymakers, public health officials, and clinicians about the uses and limitations of summary measures of population health. Although it may seem unusual for work on ethical implications to precede consensus on standards for measures, the committee believes that better ethical under-

OCR for page 17
standings should inform standard setting and education. The recommendations for future work related to summary measures of population health focus on broad directions rather than specific technical issues. The objective is to promote the design of summary measures that are understandable and helpful and to encourage the use of these measures in ways that are accountable and credible. The recommendations are directed primarily at the Department of Health and Human Services, the sponsor of this study. The Department of Health and Human Services should initiate a process of analysis and public discussion to clarify the ethical assumptions and value judgments embedded in the summary measures and to assess the critical ethical and policy implications of differing designs, implementation approaches, and uses of these measures. Because controversies about the ethical bases and implications of different measures may so greatly affect the acceptability and application of summary measures in policymaking, those encouraging the use of summary measures and those who would actually use them should have a better understanding of the ethical underpinnings and implications of different measures. Potential users of these measures would then be better informed about the measures’ limitations and less anxious that controversial value judgments may lie buried in obscure, complicated statistical methodologies. As a consequence, they would be better prepared to make choices about whether and how to use summary measures. To support better understanding of summary measures, the committee suggests several lines of research. One would continue deployment and testing of several different measures, at least for monitoring purposes. Here the goal would be to develop a body of empirical work on the distributive implications of different measures. Such research would highlight the extent to which methodological and value choices embedded in the measure shape the picture presented to decisionmakers. A second line of research would aim to develop more sensitive and sophisticated techniques for examining public attitudes and reasoning processes related to resource allocations and valued health states. Here it would be crucial to dig beneath the initial preferences people might express to uncover the principles and rationales that guide their moral reasoning about these issues. A related area of investigation would involve a combination of empirical and normative work aimed at clarifying issues of public accountability for specific applications of summary measures in decisionmaking. A third—and more sweeping—line of investigation would involve further philosophical work on distributive issues. It would seek ways to increase public agreement about distributive policies. Given the divisive potential of such policies and the diversity of philosophic and political positions in the United States, the committee recognizes that such agreement may continue to be elusive. The Department of Health and Human Services should create a process to establish standards for population health metrics and assess the feasibility and practicality of a compatible set of health status measures that could be used for different descriptive and decisionmaking purposes.

OCR for page 17
Several agencies and units within DHHS would have to cooperate in establishing a standard-setting process and in integrating it with broader international efforts to secure agreement on standard summary measures to use for specific applications. The goals for the standard-setting process should include providing a coherent framework for the testing and application of summary measures of population health, including clear terminology and criteria for assessing when measures may not be suitable for specific uses; clarifying the values and assumptions underlying different measures and their application for different purposes; identifying similarities and difference between existing measures of individual and population health and proposed standard measures; coordinating with the development, application, and standard-setting efforts of other national and international bodies; and providing a long-term strategy to encourage compatibility between individual health outcome measures and summary measures of population health. It was beyond the scope of this project to propose specific strategies for setting standards and developing national and international consensus on the suitability of different summary measures for different purposes. Certainly, the National Center for Health Statistics, the Center for Chronic Disease Prevention and Health Promotion, the National Committee on Vital and Health Statistics, the Agency for Health Care Policy and Research, and other parts of DHHS have led or otherwise participated in similar types of activities to develop shared standards for collecting and reporting many kinds of health information. It is the consensus of the committee that linking public health and medical care measurement strategies would be of fundamental importance in assessing the performance of each system individually and in relation to one another. Accordingly, measurement strategies that afford this flexibility should have higher priority for further investigation and refinement. Other priority issues for exploration concern which domains of health are most important to include within measures designed for multiple uses; what ways of valuing health states are most valid and understandable; to what extent variations in measurement techniques result in markedly different estimates of health status or disease burden; and which measures best permit the identification of links between risk factors and health status. The committee does not underestimate the difficulties to be faced in reaching agreement on a credible, useful set of health metrics. Some of the differences are political, for example, resistance from those already committed to particular measures. Others are philosophical, for example, disputes over the ethical implications of different measurement strategies and assumptions. Other difficulties are practical or methodological, for example, adopting new or different data collection procedures that impose new administrative costs, complicate comparisons over time, or both. Similarly, efforts to gain agreement on a common summary measure to span all international settings should consider the availability—or nonavailability—of the required data in both more and less developed nations. Given the cost and complexity of establishing reliable, valid data collection systems, it is desirable to make a given data source serve as many uses as possible.

OCR for page 17
The committee also recognizes that efforts to encourage the adoption of a single summary measure of population health have both pluses and minuses. The pluses are comparability in different analyses and greater simplicity in educating users in the interpretation of the measures. The minuses are the risk of adopting a single measure that is ill-suited for some purposes and, possibly, the premature adoption of a measure without a solid understanding of its practical and ethical implications compared to other measures whose development may be truncated. Thus, the committee agrees with the workshop participants that “side-by-side” comparisons of different measures are essential. for movement toward adoption of a standard, comparable set of measures. Such comparisons should identify when and why measures produce different depictions of health when applied simultaneously to the same populations using broadly collected data. The Department of Health and Human Services should cooperate with other public and private organizations to educate and train public health, medical, and other relevant professionals to promote their understanding of the interpretation and appropriate use of summary measures of population health. This recommendation is directed primarily at public health professionals and clinicians but also be extends to policymakers, health plan managers, and others. Public health professionals have traditionally focused on population health, but clinicians are increasingly participating in managed care plans and large medical groups that are being held responsible for the health of populations and not just for taking care of individual patients. To reach clinicians—and benefit from their special understanding of patients and their health problems—educational efforts must involve the professional organizations to which practitioners look for guidance. Such efforts can build on existing activities such as the medicine and public health initiative in which the American Medical Association and the American Public Health Association are collaborating. To conclude, the committee believes that summary measures of health status can make important contributions to decisions about improving population health and well-being. To fully and successfully realize these contributions, continued work is needed to refine these measures and to compare the pictures of health status that they generate. It is likewise important to examine—empirically and normatively—how the use of summary measures and the differences among measures can shape, improve, or distort policy decisions. No measure will be perfect or complete for all purposes, so policymakers have to be aware of the limitations of different measures for different uses. They will continue to need many types and levels of information about individual and population health. Moving toward better summary measures will involve difficult work along many fronts including ethical analysis, measurement of individual and social preferences, and data collection. It will also require funding from national governments and international organizations and cooperation among agencies and investigators who are sometimes competitors. A variety of initiatives, including activities described in this report, provide encouraging evidence of such support and collaboration to improve summary measures of population health.