The Dream of Consensus: Finding Common Ground in a Bioethical Context

TOM KOCH and MARY ROWELL

THE DREAM OF CONSENSUS: FINDING COMMON GROUND INA BIOETHICAL CONTEXT

ABSTRACT. Consensus is the holy grail of bioethics, the lynch pin of the assumptionthat well informed, well intentioned people may reach generally acceptable positions onethically contentious issues. It has been especially important in bioethics, where advan-cing technology has assured an increasing field of complex medical dilemmas. This paperresults on the use of a multicriterion decision making system (MCDM) analyzing groupprocess in an attempt to better define hospital policy. In a pilot program at The Hospitalfor Sick Children, Toronto, a series of small scale focus groups was constituted to examinecriteria defining organ transplant eligibility. Criteria were organized hierarchically usingthe Analytic Hierarchy Process, an MCDM approach, and the resulting data was analyzedusing Expert Choice 9.0, software designed to facilitate AHP analysis. Qualitative andquantitative analysis map barriers to practical consensus in a way not previously possible.

KEY WORDS: analytic hierarchy process, bioethics, consensus, transplantation, groupdecision making

INTRODUCTION

Consensus is the holy grail of liberal ethics and of bioethics, “the touch-stone of truth and the guarantor of correctness in matters of belief and ofadequacy in matters of decision and action.”1 Dilemmas would disappearand paradoxical formulations evaporate if only everyone agreed upon thenature of the problems we face, and upon the mechanisms which wouldbest resolve them. Indeed some perceive consensus as a necessary mech-anism for resolving uncertainty across the plain of a liberal democracy.2 AsMark Kuczewski says in his review of Jonathan Moreno’s recent book onbioethics and consensus, “liberal democracy requires consensus becausemoral uncertainty must be managed.”3 From this perspective, the imper-ative to consensus in bioethics is a subset of the greater mechanismsby which democratic society defines an appropriate position in areas ofpersistent disagreement.

In bioethics as in democracy at large, some would argue that onlyconsensus built upon discussion can lead to the resolution of the moraland social concerns which are its focus. Consensus-based decision making

Theoretical Medicine and Bioethics20: 261–273, 1999.© 1999Kluwer Academic Publishers. Printed in the Netherlands.

262 TOM KOCH AND MARY ROWELL

is, its advocates insist, the last, best hope for those who seek an informedand acceptable resolution to the dilemmas that face us. This has led somemedical ethicists – for example, Stephen Post4 – to argue for a Habermas-style “discourse ethics”5 as a means of approximating if not truth, thena generally accepted and informed position in areas of medico-legaluncertainty.

The observed failure of consensus in some contexts, and the sometimesdisastrous results where it has been achieved, lead others to doubt its poten-tial. “Consensus,” as Nicholas Rescher insists, for example, “can be andoften is no more than an agreement in folly.”6 The best that can be hopedfor, he argues, is a mechanism that limits conflict when unresolvable moraldisagreements inevitably occur. Critics of consensus can point to a richfield of dilemmas in which agreement remains illusive and disagreementseems intractable. Abortion remains a subject for which no general, socialconsensus seems possible, for example. Similarly, the debate surroundingthe appropriateness of euthanasia and assisted suicide continues to divideNorth American societies, as does the evolving debate on genetic engin-eering, and more generally, the ethics of evolving genetic science and itspractice. More darkly, detractors may point to the early popular consensuson eugenics in both North America and Europe,7 and the resulting useof that agreement to justify genocide and forced experimentation on thedisenfranchized in Germany and the US. If consensus can lead to a generalagreement to commit atrocious acts, then one might question its suitabilityas a trustworthy goal.

The questions are at once epistemological and phenomenological. Whatis the moral relevance of consensus? Perhaps more basically, is consensuspossible, and where achieved, how can the accuracy of its conclusionsbe examined? Where consensus seems impossible, what factors inhibitits achievement? Partial, tentative answers to some of these questionsoccurred to us as we analyzed data on the applicability of multicriteriondecision making models (MCDM) in a pilot program at The Hospital forSick Children, Toronto. Our original intent was not to address the problemof consensus in bioethical decision making. Rather, we sought to test whatwe hoped would be a better way to develop guidelines and possibly policyin areas of bioethical uncertainty. Still, we believe our work has clearrelevance to the ongoing debate on both the nature of consensus and itspotential application in at least some areas of bioethics.

The following section briefly reviews the problem we originally facedand the method we choose to address it. Our results as they relate to issuesof consensus are then described. Finally, the potential of these lessons both

THE DREAM OF CONSENSUS 263

for future research and for consensual decision making in bioethics aresuggested.

ORGAN TRANSPLANT ALLOCATION

The chronic scarcity of suitable human organs available for transplant-ation, coupled with the increasing demand for such organs by peoplerequiring transplantation, is a bioethical ‘nightmare.’ Who among themany potential recipients will receive this heart, this liver, or this kidney?How will hospitals decide between equally needy claimants whose socialand medical contexts differ radically, person to person. Should criminalshave an eligibility status equal to that of the general citizen? Should aperson with Down Syndrome be considered as equal to another person ofaverage intelligence for purposes of transplant organ allocation?

Following a series of highly publicized cases in North America inwhich questions of organ transplant eligibility were clearly in dispute, wefirst examined the problem of transplant allocation generally,8 and thendefined a research program which would allow us to analyze the criteriaused to decide who would and who would not be accepted as a potentialorgan transplant recipient at our hospital.9 Because both analysis of theproblem of organ transplant eligibility and the reasons for our approachhave been previously presented,10 both are only briefly summarized hereto provide a context for our current analysis.

A careful review of both the technical literature11 and the public debatesurrounding controversial transplant decisions in Canada and the US12

yielded a series of criteria generally deemed critical in deciding who wouldand would not be considered a potential organ transplant recipient. Thesecriteria, summarized here as Appendix 1, were then organized hierarchic-ally and entered into a multicriterion decision making (MCDM) softwareprogram, Expert Choice 9.0 utilizing the Analytic Hierarchy Process.13

Level-1 elements were commonly accepted criteria (consensus, survival,etc.) while level-2 elements were chosen to more clearly define the super-ordinate criteria. For the level-1 criteria “probability of post-transplantsurvival,” for example, the level-2 criteria attempted to define survival interms of a specific, relevant time span.

In general, MCDM recognizes that few decisions are made on the basisof a single, quantifiable criterion. It thus seeks to evaluate the relation ofmultiple criteria to a specific problem or goal. The Analytic HierarchyProcess facilitates the quantification of group decision making through aseries of paired comparisons of criteria organized hierarchically. Criteriaare compared in a pair-wise fashion as to their importance or contribution


to a specific goal, in this case organ transplant eligibility. Comparisonscan be given on either verbal or numerical scales. In groups, individualdecisions are averaged using the geometric mean. Participant comparisonsof paired criteria yield a relative scale of the importance assigned indi-vidual criteria. In the simplest terms, the procedure creates a judgmentmatrix whose mathematical resolution is facilitated by the software. Theresulting data permits inter and intra-group comparisons.

A series of focus groups was constituted to consider the hierarchy webelieved described the field of criteria commonly used in organ trans-plant eligibility decision making. Their size ranged from five to elevenpersons. These included two hospital groups (HSC1 and HSC2) whosemembers were drawn from surgery, nursing, psychiatry, general pediatrics,social work, pastoral care, and bioethics departments. The groups wereheterogeneous rather than department specific to see if a demonstrableconsensus could be reached within the diverse, hospital culture. Anotherfocus group was composed solely of members of a local chapter of theDown Syndrome Family Association. The greater frequency with whichpersons with Down Syndrome require organ transplantation, a fact madeevident in two cases of public controversy,14 made them critical stake-holders in the debate. Finally, a control group of citizens drawn from theToronto Beaches community was asked to consider the organ transplanthierarchy. None of the control group’s members had a medical backgroundor personal experience with transplantation.

Prior to meeting, members of each group were given backgroundmaterials describing in general terms the issues surrounding organ trans-plant eligibility. Reference to then current, publicly contentious cases wasincluded to assure a body of practical as well as general informationacross all groups. Also included in the preparatory materials was both ageneral summary of the Analytic Hierarchy Process and a description ofhow it would be employed in group discussions. When each group met,permission to tape record the sessions was secured from all members.

For each pair-wise comparison, individuals were asked first to comparetwo criteria and to say if they were equal, or if one or another wasmore important to the group. For example, “With respect to the goal oforgan transplant eligibility, is ‘probability of post transplant survival’ moreimportant, less important, or equally important than the criteria ‘publicrecognition.’ ” Where no consensus could be reached after discussion –a rare occurrence – both criteria were ranked as equal. Where agree-ment was reached on relative importance, a second question was asked todefine the intensity of that judgment on a scale of one (virtually equal) tonine (extremely more important). Because each set of pair-wise questions


required debate, discussion of the full hierarchy required between 3.5 and4 hours of debate.

At each stage of the discussion, individual judgments were noted on aspread sheet to assure a future record of individual decisions by membersof each group. Upon completion of a session, the judgments of groupmembers on each question were combined using the geometric mean. Thisresulted in a group value that was then entered into the Expert Choicesoftware. The resulting ratios ranking individual criteria for each group,separately calculated by Expert Choice, were used to make intra-groupcomparisons. An index of consistency then was calculated for each groupin an attempt to determine the relative degree of internal agreement (Figure2), and to facilitate comparison between groups. This was generated usingthe software’s built in module.15

Finally, results from all groups were combined, yielding a consensualportrait of relative valuation. To assure the cumulative results would beacceptable to individual group participants, a general statement basedupon it was distributed to all participants with a questionnaire seekingtheir response to it. Of the 60% who responded, none disclaimed thegeneral conclusions of the final statement. Indeed, most respondents atthis stage were enthusiastic about the process and its results, urging thepilot program’s expansion to include additional focus groups. Thus onthe surface it would appear that this diverse collection of sequential focusgroups was able to reach a broad consensus in an area that has proved prob-lematic in other contexts. This apparent agreement masked certain areas ofdisagreement and confusion, however, which are of primary importance.

RESULTS

While there were certain points of clear consensus, there were also import-ant differences between groups. A summary of the first-level criteria, groupby group, presented as Figure 1, summarizes the judgments of all groupsin regard to primary, Level-1 criteria.

Three groups, for example, ranked probability of post-transplantsurvival as the single most important criteria. The fourth, HSC1, ranked itas only slightly less important than level of post-transplant activity, insist-ing that “survival” as an independent value must be qualified by somequality of life statement. Overall, however, survival was strongly valuedover all other criteria. A review of individual judgments within each groupsaw it almost universally ranked as “extremely more important” whencompared with all criteria. Even with the qualification of HSC1’s use of


Figure 1. Group summary.

the criterion “activity” as a qualifier of survival, there was strong unanimityamong individuals and across groups with regard to this value.

As one commentator has noted, this in itself was significant. Thisconsensus both predated and offered support for the 1996 announcementby the US United Network for Organ Sharing (UNOS) that it would givepriority for transplantation to those with the best chance for survival.16 Thispolicy was a departure from the more traditional, “sickest first” approach inwhich transplant centers were urged to consider the urgency of the patient’scondition as the principal allocation criterion irrespective of the probabilityof post-transplant survival or other criteria.

In our study, there was a general consistency in the ranking of a rangeof criteria. Disagreement governed, however, in discussion of the criterion“compliance.” As Figure 1 makes clear, it was very strongly valued bymembers of the citizen group, who ranked it second in overall importance,and also by the Down Syndrome family association discussion group. Incontrast, members of the hospital groups – who would have been expectedto insist upon its importance – devalued it strongly.


Figure 2. Inconsistency index.

A review of individual judgments within each group showed a highdegree of conformity on the relative importance of compliance relat-ive to other criteria. The disparate valuations between groups, however,seemed to question what was otherwise a generally consensual perspect-ive. The Citizen control group’s strong valuation of this criterion actuallyskewed its index of inconsistency creating a larger than acceptable rating,shown here in Figure 2. When that group’s valuation of “compliance” wasdevalued experimentally, however, the level of inconsistency for this groupdecreased to acceptable limits. Further, by decreasing the valuation givencompliance by the Down Syndrome group, its level of inconsistency wasimproved to match those of the hospital groups. What happened? Why wasconsensus on this one criteria so difficult to achieve? What did it mean tohave one area of disagreement affecting what appeared otherwise to be abroadly consensual point of view?

Consensus

A review of the taped discussions of the various groups revealed a seriesof diverse definitions of consensus resulting in this apparent anomaly.Although all received the same preparatory materials in which this andother criteria were defined, members of each group spontaneously adop-ted idiosyncratic definitions that affected their overall judgment in pair-wise comparisons. HSC1 members defined compliance as a measure ofcommunication between staff, patients, and patient families. In a pediatricsetting where the child is dependent on parental decisions and parentalcompliance, participants saw the application of compliance as potentiallypenalizing a child for parental actions (or inactions) beyond the child’scontrol. Further, few participants defined compliance as necessarily relev-ant to patient survival, and thus devalued it as a purely clinical criterion inrelation to the goal of transplant eligibility.


HSC2 at first defined “compliance” as a critical criteria contributingdirectly to the probability of post-transplant survival. A discussion of theefficacy of compliance as a predictive criterion, and of HSC1’s reserva-tions, led them to accept the first group’s definition. HSC2 participantsagreed that if compliance could be used as a predictor of patient survivalit would be second only to survival as a criterion. If, however, it wasprimarily a measure of difficulty in communication, its importance to thegoal would be devalued. “This changes everything,” said one member whooriginally saw it as an indicator of post-transplant survival.

Because its utility as a predictor was discussed in both hospital groups,it is important to note that compliance appears to be a generally poorpredictor of post-transplant behavior or survival. In one study of alcoholuse by liver transplant recipients, for example, those with a history of alco-hol abuse reported a lower rate of post-transplant drinking than did livertransplant recipients without a similar history.17 More generally, Cham-bliss suggests the figure of a least periodically non-compliant hospitalpatients is as high as 50% of all patients.18 He argues – and members ofHSC1 agreed – its utility as a medical designation often represents notclinical but social and personal judgments by staff members. In the end,compliance was seen in both hospital groups not as a strong predictor ofsurvival, or as an important independent value in its own right, but as ameasure of staff ability in dealing with and explaining complex medicalissues to lay patients. Its relation to the goal, organ transplant eligibility,was thus diminished.

For the Down Syndrome Family Association group, however, adher-ence to medical directives was the “buy in,” the social responsibility ofpatients seeking a scarce public good. Those who refused the advice ofmedical professionals and still insisted on a transplant were seen as abusingthe system. Thus as a negative indicator it assumed greater importance inthis group. For the Citizen Group, compliance was a requirement defin-ing patient trust of the medical professionals, and perhaps of the publicmedical system at large. “You don’t know,” said one person. “So youhave to trust the doctors.” For both groups, compliance was thus whollydivorced from the question of clinical outcome, representing instead socialvalues related to patient participation in the publicly funded health caresystem, and trust in its administrators.

In discussion, members of these two groups agreed that medicalpersonnel had a responsibility to explain clinically relevant requests topatients. For them, compliance as a criterion necessarily included a processof staff explanation, discussion, and perhaps negotiation with patients. “Ofcourse,” said one participant, health providers had to be able to substan-


tiate and justify requests. All assumed this regularly occurred. In a contextof disclosure, therefore, compliance was for them a necessary, criticalelement of patient trust in the medical science whose benefits they sought,and in the persons who would provide it. Those who rejected this trustshould, they said, be rejected as transplant candidates.

DISCUSSION

These results offer a new perspective on what has been primarily a concep-tual debate. Perhaps its greatest boon is methodological. By permittingthe quantification of individual judgments, group aggregates of thosejudgments, and comparison between focus groups (or individual groupmembers), clarification of concepts and definitions was brought forward.In addition, intra-group areas of agreement and disagreement can beclearly noted. When combined with recordings of the discussions them-selves, the process by which consensus is either facilitated or stymied inthe context of small group discussions is more clearly revealed.

In the evolving literature of bioethics and consensus it is assumedreflexively that people share a common data base and a common perspect-ive. Agreement, therefore, a unified vision of correctness, if not truth,should be reached through sustained discussion. Certainly, that may occur.In our study there was strong consensus on the importance of a seriesof values across the range of discussants. These included the primacyof post-transplant survival as a primary consensus, for example, and thedevaluation of a person’s public notoriety and past contribution as criteriain transplant allocation decision making.

Consensus may be inhibited, however, when peoples discussing a singlegoal do so from the perspective of divergent definition sets and knowledgebases. Further, these differences may persist despite shared data sets – inthis case the preparatory materials – as individuals and groups spontan-eously redefine individual criteria to match their own social or professionalperspectives. Put another way, consensus may only be possible in a contextwhere primary definitions and the parameters of acceptable data have beenpreviously and self-consciously regularized. To the extent consensus is agoal in bioethics, information theory and sociology – not philosophy andpolitical science – may be the critical fields in which its potential is beststudied.

Differences at the level of definition or perspective may illuminatesignificant barriers. They may also offer the key to eventual agreement. Toaccommodate the diverse definitions of all participants, in our final report“compliance” was split from a single criterion into two distinct criteria.


In the first it was used as an independent criterion representing the patient– hospital staff relationship. Its Level-2 components described both staffand patient responsibility as independent sub-criteria. While staff wereresponsible for informing patients of the reasons for procedural requests,patients had the responsibility of discussing staff demands. Further, theyhad to understand that rejection of staff recommendations might carryserious consequences. Secondly, however, compliance was included as asubset of the criterion, “survival.” As physicians in the hospital groupsemphasized, there are cases – for example, continued intravenous drug use– where non-compliance will clearly affect a transplant recipient’s poten-tial for post-transplant survival. In this manner, the views of all groupswere accommodated and a general consensus on these issues achieved.These distinctions were accepted by all participants who reviewed thepost-study report.

CONCLUSION

While consensus may be a central goal of practical ethics, it is one whichhas remained illusive. Proponents insist that through honest discourse andwith time, unanimity may be achieved. Critiques suggest that while desir-able, perhaps, consensus is a chimera. Disagreement is inevitable. Ourwork steers a middle course through this mire. We believe the degree ofagreement found between these groups argues for the potential of con-sensus as a tool for clarification in bioethical decision making, at least atthe level of policy for a hospital region. Where it is thwarted, the failure toachieve unanimity may result from differences in social vision and clinicalvalue reflected in the definition of criteria typically discussed as if theywere similarly defined.

For us, these differences offer an opportunity for conciliation if diver-gent assumptions can be regularized in a hierarchy, rebalanced to permitmultiple perspectives within the multiplicity of sometimes competingcriteria stakeholders bring to any single problem. What results is not“truth,” however, an immutably correct position. Indeed, this approachhighlights the limits of principled consensus as well as its potential. At itsbest, consensual positions are always conditional, their durability based notonly on participant unanimity but, more importantly, on the set of shareddefinitions and antecedent perspectives that are the underpinning of anyindividual or group’s conclusion.


APPENDIX 1: HIERARCHY OF CRITERIA

GOAL: Transplant Organ Eligibility

Intelligence: Measured by the Stanford Binet scale as a criteria for organtransplant allocation decisions.

a. GENIUS:>4140 pointsb. GIFTED: 115–140 pointsc. NORMAL: 80–115 pointsd. DIMINISHED: 55–80 pointse. LOW IQ:<55 points

Survival: Assumed likelihood of patient survival following a transplant in years.

a. >10 yearsb. 5–10 yearsc. 1–5 yearsd. <1 year

Physical Independence: Anticipated life style outcome for a patient following asuccessful transplant.

a. FULL: Patient will likely become fully self-sufficient.b. FAMILY: Patient will require ongoing, long-term family supportc. INSURER: Patient will require long term financial support from an insurer,

or other non-governmental source.d. SOCIAL: Patient will require long term social support for the foreseeable

future.e. COMBINAT: Patient will require a package of two more support areas (i.e.

Family and Social) for the foreseeable future.

Activity: Anticipated level of possible patient activity following a successfultransplant.

a. WORKING: Patient will be able to return to a full schedule at office, school,etc.

b. HOME: Patient will be home-bound, but otherwise able to pursue a full rangeof professional and personal activities.

c. HOME CARE: Patient will remain home-bound, requiring help and supportin a restricted work and activity environment.

d. LONG TERM CARE: Patient will continue to require in-hospital care andsupport.

e. ICU: Patient will continue to require ICU-level care and monitoring.f. ICU–support: Patient will require both ICU-based care and life support

services following a successful transplant.

Recognition: Social recognition as a measure of patient public service.

a. SUPERIOR: Patient is internationally recognized on one or more fields.


b. EXCELLENT: Patient is nationally or regionally recognized in one or morefields.

c. NORMAL: Patient makes “average” contributions in one or more fields. Forexample, student who works in the schools recycling program and is on theschool newspaper.

d. NORMAL: Patient receives no recognition for contributions in any field orcommunity.

e. NEGATIVE: Patient’s activities are perceived as anti-social. Consider thediagnosed psychopath; sociopath; career criminal.

Compliance: Perceived level of patient compliance with medical staff directions.

a. FULL: Patient is perceived as following staff directions, and is expected tocontinue to do so post-transplant.

b. PARTIAL: While there is some friction between the patient and staff, areasonable working relationship exists, and is expected to continue post-transplant.

c. MINIMAL: Staff-patient relationships are sometimes strained. Patient isperceived as being uncooperative and expectation for patient compliance withrecommended post-transplantation monitoring are minimal.

REFERENCES

1. Rescher, Nicholas.Pluralism: Against the Demand for Consensus. NY: OxfordUniversity Press, 1993; 185.

2. Moreno, Jonathan D.Deciding Together: Bioethics and Moral Consensus. NY:Oxford University Press, 1995.

3. Kuczewski, Michael. Review of Moreno, Jonathan D.Deciding Together: Bioethicsand Moral Consensus. NY: Oxford University Press, 1995. InCambridge Quarterlyof Healthcare Ethics. 6: 3, 358–359.

4. Post, Stephen.The Moral challenge of Alzheimer Disease. Baltimore, MD. 1995.For an insightful review of Post’s Book see McCurdy, David B. Alzheimer diseaseMaking the Rounds in Health, Faith & Ethics. Chicago, III. Parkridge Center, March25, 1996, 5.

5. Habermas, Jurgen.Justification and Application: Remarks on Discourse Ethics.Cambridge, Mass: MIT Press: 1993; Habermas, Jurgen.Moral Consciousness andCommunicative Action. Cambridge, Mass: MIT Press, 1990.

6. Rescher, Nicholas.PluralismOp. Cit. note 1, 185.7. Pasternack, Martin S.The Black Stork; Eugenics and the Death of ‘Defective’ Babies

in American Medicine and Motion Pictures since1915. NY: Oxford University Press,1996; Binding, Karl and Hoche, Alfred. Permitting the destruction of unworthy life.Walter Wright, trans.Issues in Law and Medicine1992; 8: 2231–2268.

8. Koch, Tom. Normative and prescriptive criteria: The efficacy of organ transplantationallocation protocols.Theoretical Medicine1996; 17: 1, 75–93.

9. Koch, Tom and Rowell, Mary. A pilot study on transplant eligibility criteria: Valuingthe stories in numbers.Pediatric Nursing1997; 23: 2, 160–166.


10. Koch, T. Normative and Prescriptive Criteria. Op. Cit. Note 8. Koch, Tom and Rowell,Mary, ibid.

11. Including Cook, RD, Staschak, S, Green, WT. And Vargas, LG, A method to allocatelivers for orthotopic transplantation: An application of the analytic hierarchy process.Proceedings of the International Conference on Multiple Criteria Decision Making.Bangkok, December 6–8 1989; Corley, ME and Sneed, G. Criteria in the selection oforgan transplant recipients.Heart and Lung. 1994; 23: 6, 453–454; Obrisch, ME &Levenson, JL Psychosocial evaluation of heart transplant candidates; an internationalsurvey of process, criteria, and outcomes.Journal of Heart Lung Transplant. 1991;948–955.

12. These included the initial rejection of persons with Down Syndrome as suitablecandidates and the transplantation of both former baseball player Mickey Mantleand of actor Larry Hagman. These cases are reviewed in Koch, T.Normative andPrescriptive Criteria, op. cit. note 10.

13. Fernandez, AA. Expert choice: Pro Version 9.0 for Windows confirms product as anoutstanding choice for addressing complex, multicriteria problems.OR/MS Today,1996; 23, 4.

14. Koch, T. The Canadian question: What’s so great about intelligence?CambridgeQuarterly of Healthcare Ethics. 1996; 5: 2, 307–310.

15. For a brief review of the method used in this program, see Saaty, Thomas L.DecisionMaking for Leaders: The Analytic Hierarchy Process for Decisions in a ComplexWorld. Vol. II. Pittsburgh: RWS Publications, 82–84.

16. Ott, Barbara B. Commentary on Koch and Rowell article: Changes in liver transplantpolicy. Pediatric Nursing. 1997; 23: 2, 167–168.

17. Osorio, RS, Ascher, NL, Advery, M, et al. Predicting recidivism after orthotopic livertransplantation for alcoholic liver disease.Hematology1994; 20: 1: 1, 105–110.

18. Chambliss, Daniel F.Beyond Caring: Hospitals, Nurses, and the Social Organizationof Ethics. Chicago: University of Chicago Press, 1996, 138.

Hospital for Sick Children,Department of Bioethics,515 University Ave.,Toronto, OntarioCanada M5G 1X8

The Dream of Consensus: Finding Common Ground in a Bioethical Context

Documents

Transcript of The Dream of Consensus: Finding Common Ground in a Bioethical Context