J7ournal ofMedical Ethics 1996; 22: 267-272

Preferences, needs and QALYsJoshua Cohen University ofAmsterdam, the Netherlands

AbstractQuality Adjusted Life Years (QALYs) have become ahousehold word among health economists. Their use as ameans ofcomparing the value of health programmesand medical interventions has stirred up controversy inthe medical profession and the academic community. Inthis paper, I argue that QALY analysis does notadequately take into account the differentiated nature ofthe health state values it measures. Specifically, it doesnot distinguish between needs and preferences withrespect to its valuation of health states. I defend the viewthat needs and preferences are clearly distinguishable,and that the concept of needs cannot be dispensed with,as many health economists suggest. It is argued that thescale along which health states are measured in QALYanalysis is not a continuous interval scale, but onewhich concerns two distinctly different value dimensions.Measuring the values of health state intervals mayreveal the weighting attached to the different valuedimensions.

A traditional feature of 19th century classicaleconomics was that it categorised commoditiesinto two distinct groups: necessaries and luxuries.Necessaries are considered those commoditiesrequired for sustaining life, while luxuries are thosecommodities that do not serve the purpose ofsubsistence, but rather, serve to satisfy higher orderwants. This distinction between necessaries andluxuries was criticised by Alfred Marshall, whobased his criticism on the difficulty of establishingwhether a commodity belonged to the category nec-essaries or luxuries.'The so-called marginalists before Marshall,

starting with Jevons and Menger in the 1 870s,argued that a hierarchical ordering ofwants could beestablished, but that whether the term needs orwants was used was immaterial. Despite this,implicit distinctions between needs and wants interms of their differing nature were made by themarginalists. Jevons, for instance, suggested that

Key wordsQuality Adjusted Life Years; needs; preferences; quality oflife.

"the necessaries [needs] of life are so few and simplethat a man is soon satisfied in regard to these, anddesires to extend his range of enjoyment".2 Thesuggestion here is that the fulfilment of needs,however few and of marginal importance toeconomic analysis, is a prerequisite to having desiressatisfied. In addition, there is the implicit suggestionthat needs and wants are distinguishable on accountof their nature, with one being indispensable to life,the other not.

Further on in his book, Jevons explained the "lawof variation of utility", utility being somethingmeasured by and defined as "the addition made to aperson's happiness".3 He used the example of theconsumption of quantities of food, where the utilityof the first increment of food is infinite since the firstincrement is "indispensable to life".4 The utilities offurther increments of food become definable anddeterminate at a certain point, and diminish withevery additional increment. Clearly, Jevons's initialconflation of needs and wants, is unravelled in hisexplanation of diminishing (marginal) utility. Asimilar thing happens in Menger's Principles ofEconomics.5 In this book, Menger, like Jevons,initially suggested that there are "needs of differentkinds" that can be satisfied and that each need can besatisfied "more or less [completely] ".6 The terms,needs and wants, are used interchangeably in thetext, indicating that Menger did not see a reason todistinguish sharply the two concepts. Nevertheless,throughout his analysis of how goods get value thereare allusions to the fact that man has certain basicneeds which differ in nature from wants, and that theformer are to be analysed separately from the latter.Modem mainstream economists' charges against

the concept of needs include its alleged ambiguity,its indeterminacy and its subjectivity or value-laden-ness.7 Putting needs into practice, as opposed topreferences, as part of social policy, is considered tobe an arbitrary, paternalistic business, not in linewith the economist's espoused professionalobjectivity. Contemporary economists assert thateverything that is done with needs can be done withpreferences, using conventional indifference curveanalysis. If, indeed, someone feels that he or sheneeds something desperately, this will show up as a

268 Preferences, needs and QALYs

very intense preference. One possible way to showthis is to define degrees of intensity of preferences interms of elasticity of demand. Those goods for whichthere is inelastic demand are intensely preferredgoods. Other goods have more elastic demand andpreferences for them will tend to be less intense. Buthere, as above in the case of the marginalists, we areleft with a sense that although the differences aresupposedly only ones of degree, at a deeper levelthey appear to be founded on differences in kind.Perfectly inelastic goods, for instance, would appearto have a different (ontological) character fromgoods that are elastic. Goods that are perfectlyinelastic would seem to fit the description of neces-sities or needs, whereas elastic goods would appearto fit the label of preferred goods or items desiredbut not required.

Basic needsAs was mentioned above, one of the chargeslevelled against needs by economists concerns itscultural and context-dependent value-ladenness. Atone point in time, a good might be considered aluxury, something preferred but not needed. Atsome later point in time that same good might cometo be considered a need. Goods like refrigerators andtelevisions come to mind. Surely, this charge hasmerit in certain instances. However, if we restrictour definition and application of the concept ofneeds to basic needs (as will be done below) requiredfor sustaining life, then, I believe, this charge is leftrather harmless.

Those who suggest that needs are distinguishablefrom preferences or wants assert that needs areirreducible, and that they therefore may not beconflated with preferences. It does seem that certainbasic needs fit this description, basic needs beingthose needs requisite for sustaining life. One suchneed which appears to be a legitimate candidate ismedical care, particularly in its relation to prolong-ing life. Here, a patient's basic needs are consideredto be directly related to required courses of actionwithout which a patient ceases to be functional.Being functional would, of course, include survival,though not necessarily every form of survival. On thecontrary, a patient's preferences may encompassexpressions of personal preference unrelated toconsiderations of survival or subsistence.8Needs also seem to have a more "objective"

character than preferences. Although they are notvalue-neutral as Willard9 rightly argues, one can tellin a reasonably objective way what they are and howthey can be satisfied. Certainly in the case of anunconscious patient whose preferences we cannotdetermine, one can, on medical grounds, determinerather objectively which basic needs the patient has.

Arguably, a societal consensus on basic needs iseasier to achieve than one on preferences. It wouldseem reasonable to suggest as Braybrooke does,'0

that drawing up lists of satisfiable basic needs whichdeserve society's first priority is a comparativelyeasier task than drawing up lists of satisfiableindividual preferences. Evidently, in contrast topreferences, "we already know a fair amount aboutthe needs common and permanent among economicagents. They are specified in life science's findingsabout minimal levels of sustenance, the limits onlife-sustaining climate and weather, and the othervariables on which persistence and evolutionaryfitness of homo sapiens depend". Furthermore, "inmany cases we know at least in principle how and inwhat units to measure [needs] ".

It can also be said that needs, drawn up by societyon a consensus basis, seem to make a morecompelling claim on us than preferences. Despitethe fact that economists tend to oppose the norma-tive use of needs, it seems that in outlining andcarrying out social welfare policy, use of needs isunavoidable. How else do we explain proposals putforward by sensible policy-makers (some of whomare economists) with regard to the establishment ofminimum wages, basic welfare, food stamps, etc,other than by reference to needs.

This is not to say that needs must alwayssupersede the preferences of individuals. Surelythere are cases in which preferences exist which dis-regard needs. As the early institutionalist economistVeblen astutely observed: "No class of society, noteven the most abjectly poor, foregoes all customaryconspicuous consumption".... "[even] the last itemsof this category of consumption are not given upexcept under stress of direst necessity"."12

Indeed, there is no contradiction between aperson having a set of needs and him preferring notto have any of them met. Consequently, chargesagainst policies based on needs concerning theiralleged heavy-handedness and restriction ofpersonalfreedom are not valid so long as allowances are madefor the disregarding of needs by those individualswho prefer to do so.

Infinite regressOne opponent of the concept of needs, Willard,9states that needs are not valuable in themselves, buta means to "something else considered to bevaluable".'3 To establish a need is to establish ameans to something else considered to be valuable.Hence, to say that one needs food is to say that foodis needed in order to survive. The problem with thisinstrumentalist reasoning is that it runs into aninescapable infinite regress. Accepting for themoment that the need for food is not valuable initself, but only a means to something else, namely,survival, one is left without an explanation for whymost living creatures express an apparent need tosurvive. I believe that it is useful to distinguish in thisregard between instrumental needs such as medicalcare and food, and non-instrumental needs such as

J7oshua Cohen 269

survival. The former are the needs Willard is talkingabout, while the latter are a special category of needswhich are valuable not as a means to something else,but valuable in themselves. In the case of a patient indanger of losing his or her life, the medical attentiongiven as an instrumental need serves the purpose ofsatisfying a non-instrumental need, namely survival.Perhaps Willard would counter the suggestion thatsurvival is a non-instrumental need by saying that itis an instrumental means to procreate. But, thequestion would then loom, why does there appear tobe an evident primarily non-instrumental need toprocreate? Surely, needs to survive and procreate arenot primarily means to something else. It seemsreasonable to suggest that survival and procreationas needs constitute things to be valued for their ownsake.

Economics imperialismWillard's strictly instrumentalist reasoning has noroom for non-instrumental needs. Needs do noteven figure as real entities in his account. Hemaintains that only those terms which can beconnected to observable facts are referential. Since,according to Willard, needs are not facts, it does notmake sense to think of them as referential: "Anattempt to discover human needs is as funda-mentally misguided as would be an attempt todiscover human rights".'4 However, one could offera realist interpretation of needs to counter Willard'sargument. People may by nature be endowed with ahierarchy of needs, some of which, the basic needs,demand attention only after the others have beenmet. 5

As an economist, I am interested in examining thevarious areas of application of economic analysisoutside the realm of economics. One particular areawhere so-called "economics imperialism" hassuccessfully made inroads is medical care. Healtheconomists have been involved in health caredecision-making for well over three decades. One ofthe most controversial items to come out of thehealth economist's toolkit is the Quality AdjustedLife Year (QALY): a measure of health benefits topatients in terms of life expectancy following medicaltreatment adjusted for quality of life, usuallymeasured in terms of degrees of disability and levelsof distress.'6-1 The QALY's origins can be tracedback to three decades of interdisciplinary researchamong operations researchers, clinicians, psycholo-gists and health economists. Upon its inception, theQALY took on a role as a potentially useful tool incost-effectiveness studies. It was this latter role thatdistinguished QALYs from other health-relatedquality of life measures (HRQLs). HRQLs had beenaround for quite some time (starting perhaps withthe 1948 Kamofsky index) before the advent ofQALYs.20 Unlike QALYs, which are almost invari-ably used as a cost-effectiveness instrument, HRQLs

are used as evaluative instruments for judging theeffectiveness of medical treatment on patients'quality of life independent of cost considerations.A QALY can be divided into two parts; a

life expectancy part (which will not be discussed inthis paper), and a quality of life component to whichthis paper is directed. The latter component is calledthe quality of life adjustment factor. This factor con-stitutes a health state preference measure. Onecommon method used to determine the quality oflife adjustment factor is the category-rating method.Respondents to a questionnaire designed by QALYanalysts are asked to rate various health states on aso-called category-rating scale. Respondents can bepatients, doctors, nurses, or even non-patients fromthe general public who volunteer to participate in aQALY questionnaire. Health states are described torespondents in terms of a number of levels anddegrees of physical, mental and social functioning.So, for instance, being on a kidney dialysis machineis described as entailing certain levels and degrees ofphysical, mental and social functioning which therespondent evaluates.The quality of life scores on the category-rating

scale are usually transformed for each health stateonto a scale from zero (state of death) to one (stateof normal health). The arbitrary setting of the scalevalues zero and one, death and normal healthrespectively, is commonplace. However, occasion-ally, the states of death and normal health are alsoranked and measured on the scale. The "index forthe quality of life can have a value between one forthe best health state [as perceived by the respondent]and zero, or even a negative value, for the worsthealth state".2' Empirical studies suggest seriousdifficulties pertaining to the valuation of extremelybad health states, those health states which might bevalued worse than death, death itself, and states justabove death. Similarly, respondents (particularlysick patients) seem to have trouble pinpointing oreven imagining the state of normal health on thescale.22 23

Bare minimumSeparate from this issue, there is reason to besomewhat sceptical about the significance of thequality of life component in valuing life. Evidently,the importance of quality of life improvementscompared to extending survival duration may besomewhat exaggerated by the researchers.24Moreover, researchers working with quality of lifemeasurements have found that people appear towant to prolong their lives whatever the quality oflife as long as there is a bare minimum quality oflife. An extensive study of patients with previousintensive care unit experience25 demonstrated thatthese patients tend to exhibit "extreme willingness toundergo intensive care regardless of their age, func-tional status, perceived quality of life, hypothetical

270 Preferences, needs and QALYs

life expectancy, or the nature of their previous inten-sive care unit experience . .. [i]t appears that regard-less of health-related disability or perceived qualityof life"26 patients choose survival, because theyattach greater relative value to survival than qualityof life.

Most of the criticism by bioethicists of the use ofQALYs in health policy focuses on a utilitarianprinciple which would maximise QALYs as thecost benefit standard does, without regard to thedistribution of benefits to the various partiesaffected. Lockwood27 suggests that there are poten-tial dilemmas posed by a possible conflict arisingbetween what he sees as the "QALYs principle ofmaximisation of perceived benefits" and the principle"to each according to his needs". Although I thinkLockwood is right to point this out, I believe that thistype of criticism is often misplaced and based on anumber of misconceptions regarding QALYs.QALY analysis is not necessarily linked to a max-imisation principle such as the one Lockwood isdescribing. In fact, as Culyer28 points out, it is quitepossible to use QALYs in accordance with a widerange of distributional principles, includingutilitarianism and egalitarianism.29 This said, I dothink that the QALY approach's neglect of needs is aserious flaw which can, in certain instances, bepotentially detrimental to the interests of certaingroups of patients. It is my view that the issue ofwhat QALY analysis measures, in particular whatthe quality of life adjustment factor measures, is onethat should be examined before one can construc-tively discuss distributional principles. If QALYs areto be distributed, then there should be agreementthat the QALY outcome is the "right" outcome todistribute. If our conception of a health care systemis founded on fulfilling basic needs to everyone as faras this is possible, as a prerequisite to satisfyinghigher order needs and preferences, then we willrequire a measure of medical benefits expressed interms of those needs. Likewise, if our conception ofa health care system is based on satisfying prefer-ences without prioritising basic needs, then we willrequire an outcome expressed in terms of thosepreferences. Without taking a view on the matter ofwhich conception of a health care system is the"right" one, I want to investigate whether QALYanalysis is adequate in measuring (basic) needs,if the latter is what is desired by policy-makers,clinicians, etc.

Deliberate conflationAs was stated above, the QALYs approach measuresbenefits yielded by medical treatment. The measure-ment of the quality adjustment factor is based,however, solely on patients' preferences. It appearsnot to take into account (nor to measure) patients'basic needs. The reason for this seems to stemfrom a rather deliberate conflation of needs and

preferences by health economists. It is assumed bythe economist that the health state preferences auto-matically take needs into account since preferencesand needs belong to the same category. As a result, aclear specification of what a need is and what apreference is, is judged unnecessary. Although it isundeniable that preferences can and do take needsinto account under many circumstances, it seemsimprobable that all needs at all times are takenaccount of in a person's preference structure.

In my view, it would also seem that the relation-ship between preferences and benefits on the onehand, and needs and benefits on the other, is of adifferent nature. Conferring benefits through treat-ment, benefits based on health state preferences,presumably satisfies these preferences. However, therelationship between needs and benefits is substan-tially different. While fulfilling a need does confer a"potential benefit",30 the reverse proposition is notnecessarily the case; conferring a benefit does notnecessarily satisfy a need.

Hypothetical exampleI hope that the following hypothetical example willhelp to make this clear. Say we are comparing ahealth programme which saves lives with a pro-gramme which improves quality of life. Specifically,let us suppose that we are comparing, in terms ofQALYs, the benefits of a kidney dialysis programmeto the benefits of a cosmetic surgery programme.Suppose that groups of patients belonging to the twoprogrammes have the same average life expectancyfollowing appropriate treatment. With treatment theaverage quality adjustment factor for the kidneypatients is found to be 03, while the cosmeticsurgery patients' average is 07. Without treatmentthe kidney patients die; death's quality of life adjust-ment factor is zero. While without treatment, thecosmetic surgery patients have an average quality oflife adjustment factor of 04. A calculation of netQALYs reveals that both programmes have equalnet QALYs. Now assume that costs are equal forboth programmes, which results in equal costs-per-QALYs. And assume further that there is onlysufficient funding for one of the programmes.Because costs-per-QALY are equal, do we (maywe?) conclude that we are indifferent between allocat-ing funds to treat either the cosmetic surgery or thekidney patients?

I suggest that we are not (may not be?) indifferentbecause we are dealing with two fundamentallydifferent kinds of health care programmes; the firstof which involves the basic need to survive, while thesecond involves preferences distinguishable from thelatter basic need. Satisfying the basic needs of thekidney patients is seen here as a necessary conditionof survival, while satisfying the preferences of thecosmetic surgery patients is not a similarly necessarycondition. Since, in all probability, the medical

J7oshua Cohen 271

profession ethos would prescribe the prevention ofdeath as a first priority before satisfying the prefer-ences of patients which are unrelated to survival, it isvery likely that funds in this hypothetical case wouldbe allocated to treat the kidney patients.

It seems that in this case we are forced to stepback from the deceivingly straightforward QALYcalculus and invoke a meta-principle of some kindsuch as the principle that saving lives is judged moreimportant than improving the quality of life ofpatients.

I believe that a comparison like this involves twodistinct value dimensions that are characterised bycategory differences in kind. One of the value dimen-sions concerned relates to a basic need to survive,while the other does not; it relates to preferences.The implication of this sort of judgment is that thescale on which the quality of life adjustment factor ismeasured may not be represented as a continuousequally spaced interval scale on which there is only onevalue dimension.

For technical reasons, QALY analysis mustassume that the transformed category-rating scalementioned above conforms to the properties of aninterval scale. It has to do this in order to ensureinterpersonal comparability of preference values. Ascale which exhibits interval scale properties is oneon which the values attached by respondents tonumerically equidistant intervals on the scale (forexample, intervals {0-1, 0} and {0 3, 02}) areequal.3' In other words, the value attached to a moveon the scale from 0-1 to 0 is the same as the valueattached to a move from 0 3 to 0-2. This is a contro-versial claim, but an empirically testable one. Itwould be interesting to discover whether in thecontext of the hypothetical example above, theinterval {03, 0} is perceived of as "equivalent" tothe interval {0 7, 0 4}. Is a move from a low qualityof life (03) to death given the same weight as a movefrom a "good" quality of life (0 7) to a lesser qualityof life (0-4)? Nord32-34 suggests, on the basis of anumber of empirical findings, that in making com-parisons similar to the one above, ie, comparingprogrammes which save lives to programmes whichimprove quality of life, it is likely that differentweights will be attached to the intervals considered.As a result, what appear to be equally spaced inter-vals, {0-3, 0} and (07, 04}, may turn out to beunequally valued intervals. Nord argues that becauseQALY analysis measures health states in isolation, itdoes not adequately take account of moves betweenhealth states. Placing this into the context of thehypothetical example above, it can be said thatQALY analysis does not adequately assess the valueof survival, that is, the severity of moving from 0 3on the quality of life adjustment factor scale to 0.

Wiggins35 and Lockwood27 argue that the claim apatient has on the health care system should be afunction of a patient's health needs as opposed to theamount of benefit (based on preferences in QALY

analysis) that the health care system can provide.In other words, it is in order that health needs be sat-isfied, that we have a positive right to health care. Inparticular, it is in order that the basic need of beingfunctional (which includes the basic need ofsurvival) be satisfied, that we have a right to healthcare. I agree with this in principle. Moreover, I thinkthat the satisfaction of the basic need of beingfunctional should be given priority in health policy.However, it should be clear what demarcates beingfunctional from being dysfunctional. The QALYapproach of measuring health state preferences willnot suffice for the task of finding out where thedemarcation point lies. QALY analysis only tells ushow people (patients, doctors, nurses, non-patients)judge health states in isolation. Perhaps continuingalong the path that Nord34 35 has set out whileinvestigating how people evaluate moves from onehealth state to another, might assist in finding such ademarcation point. How this might work can beillustrated using the hypothetical example above. Ifsaving the lives of the kidney patients is given moreweight than an improvement in quality of life forcosmetic surgery patients - that is, if the interval{0 3, 0} is given more weight than the interval (0.7,0.4} - then this would indicate that the health statecorresponding to the quality of life adjustment factor0.3, is functional. If, on the other hand, the interval(0.3, 0} is given less weight than the interval {0 7,0 4}, then this would suggest the health state corres-ponding to the quality of life adjustment factor 0-3 isdysfunctional.

AcknowledgementThis paper was written during my stay as VisitingFellow at the Department of History of Science atHarvard University (fall term 1995). I would like tothank the Dutch Foundation for Academic Research(NWO) and the Tinbergen Research Institute fortheir financial support of this fellowship.

Joshua Cohen, MA Econ, is a PhD student in theDepartment ofEconomics, University ofAmsterdam, theNetherlands.

References1 Marshall A. Principles of economics. London:MacMillan, 1890: 56-7.

2 Jevons WS. The theory of political economy. London:Penguin Books, 1970 (originally published in 1871):103.

3 See reference 2: 106.4 See reference 2: 107.5 Menger C. Principles ofeconomics. New York: New York

University Press, 1981 (originally published in 1871).6 See reference 5: 124.7 Braybrooke D. Meeting needs. Princeton: Princeton

University Press, 1987.8 Outka G. Social justice and equal access to health care.

J'ournal of Religious Ethics 1974; 2: 11-32.

272 Preferences, needs and QALYs

9 Willard LD. Needs and medicine. J7ournal of Medicineand Philosophy 1982; 3: 259-74.

10 See reference 7.11 Rosenberg A. Economics - mathematical politics or science

of diminishing returns. Chicago: University of ChicagoPress, 1992: 167-8.

12 Veblen T. The theory of the leisure class. London:Penguin Books, 1967 (originally published in 1899):85.

13 See reference 9: 272.14 See reference 9: 261.15 See reference 7: 253.16 Rosser R, Kind P. A scale of valuations of states of

illness. International Journal of Epidemiology 1978; 7:347-58.

17 Rosser R, Kind P. Valuation of quality of life: some psy-chometric evidence. In: Jones-Lee M, ed. The value oflife and safety. Amsterdam: North Holland, 1982:159-70.

18 Williams A. Economics of coronary artery bypassgrafting. British Medical Journal 1985; 291: 326-8.

19 Gafni A, Birch S. Economics, health and healtheconomics: HYEs versus QALYs. J7ournal of HealthEconomics 1993; 11: 325-39.

20 Guyatt G, Feeny D, Patrick D. Measuring health-related quality of life. Annals of Internal Medicine 1993;118: 622-9.

21 van Busschbach J, Hessing D, de Charro F. The utilityof health at different stages in life: a quantitativeapproach. Social Science and Medicine 1993; 37: 153-8.

22 van Busschbach J, Hessing D, de Charro F.Observations on one hundred students filling in theEuroQol questionnaire. Kuopio University. EuroQolConference Proceedings. Helsinki: Kuopio UniversityPublications, 1993: 55-62.

23 van Busschbach J. De Validiteit van QALYs. Arnhem:Gouda Quint, 1994.

24 Ashmore M, Mulkay M, Pinch T. Health and efficiency:a sociology of health economics. Milton Keynes: OpenUniversity Press, 1989.

25 Danis M, Patrick D, Southerland L, Green M.Patients' and families' preferences for medical intensivecare. Journal of the American Medical Association 1988;260: 797-802.

26 See reference 25: 801.27 Lockwood M. Quality of life and resource allocation.

In: Bell J, Mendus S, eds. Philosophy and medicalwelfare. Cambridge: Cambridge University Press, 1988:33-55.

28 Culyer A. Commodities, characteristics of commodi-ties, characteristics of people, utilities, and the qualityof life. In: Baldwin S, Godfrey C, Propper C, eds.Quality of life: perspectives and policies. London:Routledge, 1990: 9-27.

29 Williams A. Welfare economics and health statusmeasurement. In: van der Gaag J, Perlman M, eds.Health, economics and health economics. Amsterdam:North-Holland, 1981: 271-81.

30 Williams A. Ethics and efficiency in the provision ofhealth care. See reference 27: 111-26.

31 Gescheider G. Psychological scaling. Annual Review ofPsychology 1988; 39: 169-200.

32 Nord E. Methods for quality adjustment of life years.Social Science and Medicine 1992; 34: 559-69.

33 Nord E. An alternative to QALYs: the saved young lifeequivalent (SAVE). British Medical Journal 1992; 305:875-7.

34 Nord E. The relevance of health state after treatment inprioritising between different patients. J7ournal ofMedical Ethics 1993; 19: 37-42.

35 Wiggins D. Claims of need. In: Honderich T, ed.Morality and objectivity. London: Routledge, 1984:149-201.

News and notes

III World Congress of Bioethics

The III World Congress of Bioethics will be held in SanFrancisco, California, USA from 20-25 Novemberthis year. It is being organised by the InternationalAssociation of Bioethics in conjunction with theAmerican Association of Bioethics.Topics include: Feminist approaches to bioethics;

Studying human genetic diversity: can we doit right?, and The globalisation of bioethics:

international human rights and health professionals.For further information please contact, as soon as

possible: Congress Secretariat, III World Congress ofBioethics, Pacific Center for Health Policy andEthics, University of Southern California, Los Angeles,CA 90089-0071, USA. Tel: (213) 740-2541; fax:(213) 740-5502. World Wide Website http://www.usc.edu/dept/law*llb/bioethics/world/congress.html.