Note by Secretariat

This room document presents a review of several multi-attribute utility MAU) instruments. These instruments provide a health status descriptive system as well as a valuation function to derive a single index score of health status. This document is a final version of an internal background paper, written by Erik Nord of the National Institute of Public Health in Oslo, Norway in May 1997. Based on the responses of a questionnaire from the developers of these MAU instruments, the paper was compiled and distributed to a network of experts in this field across OECD Member countries. Comments and feedback were collected and incorporated into this room document. In the room document, Nord addresses the validity of these instruments -- particularly focusing on the valuation of health states -- and offers one recommended approach in developing a composite health indicator for both population monitoring and cost- effectiveness. Please note the original internal background paper is included as an Annex to this room document.

The background paper offers more detailed information on the health status descriptive systems of the instrument.

Summary

Single index valuations of health states on a scale from zero (equivalent to being dead) to unity (healthy) may be used to calculate composite measures of health status such as, Health- Adjusted Life Expectancy, Disability-Adjusted Life Years, and Quality-Adjusted Life Years. Health measurements in terms of multidimensional profiles (e.g. the SF-36) cannot serve this purpose. The paper discusses the validity of different techniques for assigning single index values to health states, with emphasis on Multi-Attribute Utility Instruments. A case is made for seeing health state values as measures of societal value rather than individual utility in order to obtain values that are valid and meaningful over the whole range of possible states of illness.

A pragmatic procedure is outlined for estimating Health-Adjusted Life Expectancy in a standardised way in the OECD countries. First, health adjustment of life years would be based on data on three dimensions of health that form a common core in existing health measurement instruments: physical suffering, anxiety/depression, and limitations in daily activities. Second, each dimension would have a five-point response scale used in a quality of life questionnaire developed by the WHO (the WHOQOL). A sophisticated procedure was adopted to ensure semantic equivalence in different language versions of this instrument. Third, single index values for three-dimensional health states would be obtained by means of a valuation technique -- the person trade-off method -- used by WHO in estimating Global Burden of Disease.

Background

This paper is a follow up of a report on the multi-attribute utility (MAU) instruments (Nord, 1997). The review covered the Health Utilities Index (HUI, Feeny et al., 1995), the EuroQol Instrument (EQ 5-D, Brooks et al., 1996), 15-D (Sintonen and Pekurinen,1993), the Quality of Life and Health Questionnaire (QLHQ, Hadorn, 1995) and the Australian Quality of Life (AQOL, Hawthorne and Richardson, 1996). The instruments provide numerical values for health states to be used in population health statistics, cost-effectiveness analyses and resource allocation decisions. The present paper focuses on the validity of these values and offers a concrete, pragmatic proposal for a standard measure of health-adjusted life expectancy in the OECD countries.

Issues in Health Status Measurement

There are two main ways of describing individuals' health in numerical terms. One is in terms of scores on a number of different dimensions of health, like mobility, pain, hearing and seeing. Together such unidimensional scores form multidimensional health profiles. Examples are the Sickness Impact Profile, the Nottingham Health Profile, the SF-36 and the WHOQOL. The other way is to assign a score for overall health on a single scale from zero (equivalent to being dead) to unity (healthy). Such a single index score -- hereafter referred to as a health state value - is usually interpreted as a measure of health related quality of life.

Multi-attribute utility instruments provide both profiles and health state values. Clearly, health profiles convey more information than health state values. However, profiles have the weakness that they do not always allow judgements of which of two health states is better, since one state may have higher scores on some dimensions, and the other state higher scores on other dimensions, and there is no way of judging which of the differences is more important. Health state values provide a way to resolve this problem by concentrating all profile information in one single number, according to which different complex health states may be ranked in terms of their overall value to the individuals concerned.

Health state values have an important additional property. Not only do they allow an ordinal ranking of health states. They also purport to express trade-offs between quality of life and quantity of life. This is potentially useful in three contexts.

First, health state values may be used in the calculation of health adjusted life time. For instance, if an individual lives 70 years as healthy and then 10 years in a state that scores 0.8, then his/her health adjusted life time is 70 x 1 + 10 x 0.8 = 78. Averaging health adjusted life time in a population yields Health Adjusted Life Expectancy (HALE). HALE is calculated on the basis of life expectancy tables and survey data on health status in different age groups. HALE reflects both mortality and morbidity and is assumed to be a useful statistic in comparing overall health across countries and over time and in gauging the outcomes of health related public policies (Roberge et al., 1997).

Second, the World Health Organisation is organising a large international collaborative enterprise called the Global Burden of Disease Project (Murray and Lopez, 1996). The idea behind the project is to aid priority setting in health care at the global level by collecting statistics on the degree to which different diseases represent a burden to mankind in terms of the number of people affected, life years lost and losses in quality of life. Burden of Disease is estimated by assigning disability weights to different kinds of illness. The weights use the same 0-1 scale as health state values in HALE, except that the scale is turned around, so that zero represents "no burden" and unity "maximum burden" (equivalent to "as bad as being dead"). The weights are used in combination with age weights to translate individual life scenarios into a number of Disability Adjusted Life Years (DALYs).

Third, health state values may be used in cost-effectiveness analysis to evaluate health care outcomes in terms of Quality Adjusted Life Years (QALYs). For instance, if an intervention provides one person with 10 more life years in a state A that scores 0.5, the value of the intervention would be 10 x 0.5 = 5 QALYs. The same value would be attached to an intervention that provided one person with 5 more years in full health (5 x 1. 0 = 5 QALYs), or restored one person from state A to full health for 5 years (5 x (1. 0 -- 0.5) = 5 QALYs). By comparison, an intervention that provided 10 more life years in full health would have a value of 10 x 1. 0 = 10 QALYs. In the logic of cost-effectiveness analysis of health care, the latter intervention would justify twice as high a cost as the three former ones.

It is important to recognize that each of these three applications -- HALE, DALYs and QALYs -- presuppose health measurements in terms of single index scores. Measurements in terms of profiles will not do the trick. On the other hand, the validity and meaningfulness of compressing complex health profiles into global health state values is debatable. The purpose of the following is to explain how health state values are obtained today and to suggest how they might be obtained meaningfully and efficiently in the OECD context.

Techniques for Eliciting Health State Values

Consider a complex health state, labeled A for brevity. A person's valuation of state A in terms of a single number on the 0-1 scale can be elicited in a number of ways, the most common of which are the following:

The rating scale (RS). The person is asked to locate state A on a straight line with end points representing the worst imaginable and the best imaginable health state respectively. State A then receives a value (on the 0-1 scale) which is directly proportional to its closeness to the upper end point of the straight line.

The time trade-off (TTO). The person is asked to indicate a number X of years in full health that he would consider equally desirable as for instance 10 years in state A. The value of state A (to the person in question) is then defined as the ratio X/10.

The standard gamble (SG). The person is asked to compare two scenarios. One is to live in state A with certainty. The other is to take a treatment that offers a probability p of living in full health and a probability 1-p of dying immediately. The person is asked to indicate the value of p that would make the two options equally desirable. The value of state A (to that person) is then set equal to p.

The person trade-off (PTO). The person is asked to compare two programs. One will save the life of 1 person and restore him/her to full health. The other will restore N people in state A to full health. The person is asked to indicate a value of N that would make the two programs equally worthy of funding. Say the response is N = 10. The person is then saying that the value of a cure of one person in state A is one tenth of the value of a saved life. The value of state A is then 0.9 (1. 0 -- 0.9 = 1. 0 / 10).

Historically, health state values used in QALY calculations have been derived from studies using the rating scale, the time trade-off and the standard gamble. Values used in calculations of HALE in Canada are based on a combination of rating scale and standard gamble measurements (Roberge et al., 1997). In the Global Burden of Disease Project, disability weights were initially based on rating scale questions, but are now based on person trade-off questions (Murray and Lopez, 1996).

The rating scale, time trade-off, standard gamble and person trade-off are primary techniques for valuing health states. In population health surveys and in clinical outcome studies, multi-attribute utility instruments allow analysts to assign values to individuals' health without in each case having to employ costly primary valuation procedures. The instruments have questionnaires that establish the health profiles of the individuals included in the studies. Additionally they have tables or mathematical formulas that yield an estimate of the single index value associated with any given health profile. The tables or formulas are based on prior statistical analyses of population preference data that show how highly different dimensions of health are valued relative to, and in interaction with, each other. These underlying preference data have all been collected by means of the rating scale, the time trade-off or the standard gamble, or combinations of these.

The Validity of State Values

Direct techniques for valuing health states (RS, TTO, SG and PTO) yield different results (Nord, 1992). Similarly, different MAU instruments yield different values for the same health states (Nord, 1997). These findings raise the question of the validity of these single index values. To answer this question, it is useful to distinguish between two different ways in which health state values may be understood. One is in terms of utility, the other in terms of societal value.

Utility is basically an emotional category: How good is a health state or a health outcome felt by the individuals concerned? Total utility for a group of people, for instance for a nation as a whole, is simply the sum of all individual utilities. Societal value, on the other hand, is a broader, ethical concept. While it is partially a function of total utility, it is also determined by concerns for fairness and hence by the distribution of utility across individuals.

In the following, I discuss health state values provided by MAU instruments both as measures of utility and as measures of societal value. For simplicity, I focus on measuring population health, which I believe is the most relevant application of health state values from the OECD's point of view. However, the conclusions I draw regarding measuring population health apply also to cost-effectiveness analysis.

Health State Values in Calculations of HALE

The use of Health Adjusted Life Expectancy as an indicator of population health assumes the existence of three trade-offs. First, there is a trade-off between quality of life and length of life. For instance, 75 years with considerable discomfort and disability may be regarded as equivalent to 70 years in full health. We may call this a time-quality trade-off.

Second, there is a trade-off between quality of life and the number of persons involved. Consider for instance a situation in which 100 people all live 75 years in a state that scores 0.8. Their average health adjusted life years is 60. Consider two possible changes in this situation. One consists in 20 people getting to live as healthy, while the other 80 people remain at level 0.8. The other possible change consists in all 100 people getting to live in a state that scores 0.84. In both cases the resulting average health adjusted life years is 63, i.e. an increase of 3 years. Effectively this means that, according to the HALE, taking 20 people from 0.8 to healthy is equivalent to taking 100 people from 0.8 to 0.84. We may call this a person-quality trade off.

Third, there is a trade-off between length of life and the number of persons involved. For brevity, I do not discuss this here, since it is independent of the values that are assigned to health states (see Nord (1997)). In a HALE, time-quality trade-offs and person-quality trade-offs follow mathematically from the values that are assigned to health states. One way to test the validity of these values is therefore to examine whether the trade-offs that are implied by the values, fit with the trade-offs that people in society make if they are asked directly. This is the so-called test of reflective equilibrium. If they do not fit, the values are unsuitable in the construction of a HALE. To perform this validity test, we need to do one more thing. That is to specify the substance, or the object, of the trade-offs. There are two possibilities: utility and societal value.

Utility

Let us assume that the trade-offs that one wishes to build into a HALE are in terms of utility. The proposition that 75 years of life in a scenario S is equivalent to 70 years in full health (time-quality trade-off) then means that the utility that individuals derive from these two scenarios is the same. Similarly, the proposition that taking 20 people from 0.8 to healthy is equivalent to taking 100 people from 0.8 to 0.84 (person-quality trade-off) means that the utility gained by the former change equals the utility gained by the latter change. A policy maker may want to know how these interpretations can be verified. This implies checking that the utility assigned to the various health states involved is correct. How can this be done?

One issue here is whom to ask to value health states. Partly for convenience reasons, utilities have in the past mostly been measured by asking members of the public to imagine themselves in various states of illness and to value these states. Some argue that since utilities are supposed to aid resource allocation decisions, in which most people have an interest in the long term, the general public is not only a convenient, but also the theoretically correct source of utility information. However, this position is due to a conflation of two issues. One is the measurement of quality of life associated with different health problems. The other is the measurement of distributive preferences in resource allocation. The concept of utility relates to the former and not to the latter. Arguably, the quality of life associated with a health state can only be truly known by people who have personal experience with it. It is hard then to avoid the conclusion that patients must be better sources of utilities than healthy people (Bellagio Team, 1997).

The following is based on this premise. Since utility is a matter of subjective feeling, and the strength of feelings is not directly observable, verification of the utility assignments is not straightforward. However, in theory individual utility may be assessed indirectly by looking at a behavioural correlate to the subjective feeling of utility. One such behavioural correlate is the quality-of-life-scores that subjects assign to themselves when asked to evaluate their own health on rating scales. However, rating scale scores have been shown not to have interval scale properties. That is, equally large intervals on different parts of such scales (for instance a movement from 0.4 to 0.6 versus a movement from 0.7 to 0.9 on a scale from zero to unity) do not carry the same significance to the individuals concerned (Allison and Durand, 1989; Nord, 1991; Richardson, 1994). Hence, rating scale measurements do not allow a meaningful summation of utility across individuals. I therefore do not recommend the use of utilities operationalised as self ratings on rating scales in the calculation of health adjusted life expectancy.

Based on economic theory and decision analysis, a more widely accepted behavioural correlate to perceived utility is the individual's willingness to sacrifice life expectancy to be cured of a state of illness or to obtain a given health improvement. People's willingness to undertake such sacrifices are elicited by means of standard gamble or time trade-off questions.

Societal Value

The trade-offs that one wishes to build into a HALE may alternatively be understood in terms of societal value. To say that 75 years of life with a given amount of health problems is equivalent to 70 years in full health (time-quality trade-off) then means that society values these two scenarios equally much. Similarly, the proposition that taking 20 people from 0.8 to healthy is equivalent to taking 100 people from 0.8 to 0.84 (person-quality trade-off) means that society values these two improvements equally much. Again, there is a need to make sure that these trade-offs that become built into the HALE through the health state valuations that are used, fit with actual societal preferences.

There are very few data on the time-quality trade-offs that society wishes to make in valuing alternative life scenarios for other people. Almost all studies focus on individuals' personal time trade-off preferences. An exception is a small study by Richardson and Nord (1996), who did not find any significant difference between preferences expressed in these two perspectives.

Possible Strategies to Improve Validity

One response to the above state of affairs would be to adjust the valuation algorithms of MAU instruments in such a way as to produce the necessary upper end compression of values. However, here we need to distinguish between the utility perspective and the societal-value perspective. With respect to the utility perspective, the preceding sections are based on the common assumption in health economics and decision analysis that the disutility of a state of illness may be measured as the willingness to sacrifice life expectancy to be relieved of the illness. Given this particular definition, I am led to draw the conclusion that values from synthetic indicators generally lack upper end compression, since people with illness express great reluctance to make such sacrifices when asked directly in preference studies.

If this true preference structure were built into the values assigned to health states, a serious sensitivity problem would arise: A great number of mild and moderate states would be assigned the value of 1. Real health improvements for people with such conditions would then not be captured by the MAU instruments, since there would be no differences between the values for such conditions and the value for full health. This would be a problem both in measuring population health and in cost-effectiveness analysis.

Possibly, the problem can be resolved by looking at the utility of health states from a different angle. The reluctance to sacrifice life expectancy to be relieved of a given illness may derive from a general preference for maintaining status quo, or from a general aversion to accepting any kind of loss in health, including life expectancy. In long term health planning, such preferences are arguably a source of bias, which should be avoided in the valuation of health states. In stead of asking people how much they would be willing to sacrifice to be relieved of an illness they already have, one could therefore ask them which of different life scenarios they would prefer if they were at the start of life and therefore not yet attached to any particular scenario. It is conceivable that with such a perspective, sacrificing life expectancy to gain quality of life would be an easier choice, in which case the trade-offs suggested by existing synthetic indicators would be closer to the truth.

An other possible strategy is to regard health state values as numbers that are supposed to express trade-offs in terms of societal value rather than utility. The rationale for this is that the preference for the preservation of life itself may be somewhat less absolute when people are asked to prioritise between different health care programs in a budgeting context than when they are asked about their willingness to sacrifice own future life years or certainty of survival. In other words, I am suggesting that more health states will be assigned values below unity if person trade-off questions are posed to the general population than if time-trade off or standard gamble questions are posed to people with illness and disability.

Choosing the societal value interpretation may therefore be a way to ensure both validity and sensitivity in the practice of assigning numerical values to health states.

A Pragmatic Proposal for Calculating HALE in the OECD Area

The variation in, and weaknesses of, existing MAU instruments suggests that it will not be an easy task to achieve consensus across 29 OECD countries with different languages, cultures and values on how to measure and value health for the purpose of calculating Health Adjusted Life Expectancy. Two strategies may help to limit the difficulties. One is to aim at a relatively simple set of dimensions and levels of health that may constitute a common core in health statistics across countries. The other is to adopt elements and techniques of data collection that already enjoy wide international support.

The following concrete outline of a possible way to go is motivated by these pragmatic considerations. Three main issues are addressed:

1. Which dimensions of health should be included in the standard instrument for measuring individuals' health?
2. Which response scales should be used for each dimension?
3. How should preference data be collected to allow a meaningful transformation of multi-dimensional scores into single index values?

A. Dimensions
In most people's eyes, the most salient dimensions of the concept of health seem to be physical suffering, anxiety/ depression and limitations in daily activities. The significance of these factors is reflected in the fact that MAU-instruments with only two or three dimensions concentrate on these three (Rosser and Kind, 1978; Rosser et al., 1992; Hadorn, 1995). They are also shared by more complex MAU-instruments that otherwise differ with respect to the selection of dimensions (Kaplan and Anderson, 1988; Sintonen and Pekurinen, 1993; Feeny et al., 1995; Brooks et al., 1996). While they imply a much narrower definition of health than that adopted by the WHO (1946), they are in accordance with how the general public usually thinks of health (Mathers, 1997).

A simple common core of questions in health surveys in the OECD countries could therefore look roughly as follows:

1. Do you have health problems that cause you physical suffering (pain, nausea, itching etc)?
2. Do you suffer from anxiety or depression?
3. Are you limited in your daily activities (washing, dressing, eating, (house) work, leisure activities) due to health problems?

B. LEVELS
In the OECD context, it is important that response scales are equivalent across countries. That is, people in different countries with equal degrees of health problems should ideally assign themselves the same scores on the various dimensions. TheWHOQOL (WHO, 1995) is a quality of life questionnaire with 5-point semantic response scales developed simultaneously in some 25 countries. A complex procedure, followed by all participating countries, was adopted to ensure semantic equivalence between different language versions.

One of the scales -- applicable to the questions suggested above -- is as follows in four main OECD languages:

Considering the unique effort made in the WHOQOL project to secure semantic equivalence, it might be a good idea to use this response scale for the questions indicated above. This would yield a descriptive system covering 125 possible health states (125 possible combinations of three dimensions of health with five levels each).

C. VALUATIONS
There is disagreement among constructors of MAUs on how to establish utilities. Furthermore, it is not clear at present that any of the conventional ways of measuring utility are both meaningful at a cardinal level (a problem with the rating scale) and at the same time sensitive to moderate conditions of ill health and disability (a problem with standard gamble and time trade-off).

References

Brooks R. et al., EuroQol: The current state of play. Health Policy 1996,37,53-72.

Feeny D., W. Furlong, M. Boyle, G.W. Torrance, Multi-attribute health status classification systems. Pharmacoeconomics 1995,7,490-502.

Fowler F.J., P.D. Cleary and M.P. Massagli et al., The role of reluctance to give up life in the measurement of the value of health states. Medical Decision Making 1995,15,195-200.

Hadorn D., Large scale health outcomes evaluation: How should quality of life be measured? Journal of Clinical Epidemiology 1995,48,607-618.

Hawthorne G., and J. Richardson, An Australian multi-attribute utility: Rationale and preliminary results. Working paper 49. Melbourne: Centre for Health Program Evaluation, 1996.

Kaplan R.M., and J.P. Anderson, A general health model: Update and applications. Health Services Research, 1988,23,203-235.

Mathers C., Developments in the use of health expectancy indicators for monitoring and comparing the health of populations. Report prepared for the OECD. Canberra: Australian Institute of Health and Welfare, 1997.

Menzel P., M. Gold, E. Nord, J.L. Pinto, J. Richardson and P. Ubel, Numerical valuations of health programs as a function of individual utility gains and societal concerns for severity of illness. Working paper no 2 from the Bellagio Group on Value analysis of Health Care. Mimeo. Oslo: National Institute of Public Health, 1997.

Morris J., and A. Durand, Category rating methods: Numerical and verbal scales. Mimeo. University of York: Centre for Health Economics. 1989.

Murray C.L.J and A.D. Lopez, The Global Burden of Disease. Harvard University Press, 1996.

Nord E., The validity of a visual analog scale in determining social utility weights for health states. International Journal of Health Planning and Management 1991, 6, 234-42.

Nord E., Methods for quality adjustment of life years. Social Science & Medicine 1992, 34, 559-569.

Nord E., The person trade-off approach to valuing health care programs. Medical Decision Making 1995,15,201-208.

Nord E., Health status index models for use in resource allocation decisions. A critical review in the light of observed preferences for social choice. International Journal of Technology Assessment in Health Care 1996,12,31-44.

Nord E., Time trade-off scores in patients with chronic disease. Comparison with the York hypothetical TTO tariff. Paper for the EuroQol Plenary Meeting, Oslo, October 1996.

Nord E., A review of synthetic health indicators. Report for the OECD. Oslo: National Institute of Public Health, 1997.

O'Leary J.F., D.L. Fairclough, M.K. Jankowski et al., Comparison of time trade-off utilities and rating scale values of cancer patients and their relatives. Medical Decision Making 1995,15,132-137.

Patrick D., J.W. Buch and M.M. Chen Methods for measuring levels of well-being for a health status index. Health Services Research 1973,8,228-245.

Pinto J.L., Is the person trade-off a valid method for allocating health care resources? Health Economics, 1997.

Richardson J., Cost-utility analysis: What should be measured? Social Science & Medicine 1994,39,7-21.

Richardson J., and E. Nord, The importance of perspective in the measurement of quality adjusted life years. Medical Decision Making 1997,17,33-41.

Roberge R., J.M. Berthelot, M.C. Wolfson, Adjusting life expectancy to account for morbidity in a national population. Quality of Life Newsletter No 17. Lyon: Mapi Research Institute, 1997.

Rosser R., and P. Kind, A scale of valuations of states of illness: Is there a social consensus? International Journal of Epidemiology,1978,7,347-358.

Rosser R., M. Cottee, R. Rabin and C. Selai, Index of health-related quality of life. In: Hopkins A (ed). Measures of the quality of life, and the uses to which they may be put. London: Royal College of Physicians of London, 1992.

Sherbourne C., and R. Sturm, Utility measures from the screener. Mimeo. Paper presented at RAND, Santa Monica, February 1997.

Sintonen H., and M. Pekurinen, A fifteen-dimensional measure of health-related quality of life (15 D) and its applications. In S.R. Walker and R.M. Rosser (eds). Quality of life assessment. Key issues in the 1990s. Dordrecht: Kluwer Academic Publishers, 1993.

Stavem K., Valuing theoretical versus own condition: A comparison using time trade-off and the EuroQol Instrument in COPD patients. Paper for the EuroQol Plenary Meeting, Oslo, October 1996.

Ubel P.A., G. Loewenstein, D. Scanlon and M. Kamlet, Individual utilities are inconsistent with rationing choices. Medical Decision Making 1996, 16, 108-116.

WHOQOL Group. The World Health Organisation Quality of Life Assessment (WHOQOL): Position paper from the World Health Organisation. Social Science & Medicine 1995,41,1403-1409.

Table 1

Health State Values According to Different Multi-Attribute Utilities

Table 2

Disability Classes and Several Weights for the Global Burden of Disease Study Based on the PTO Protocal Used at the Geneva Meeting on Disability Weights in August 1995