Journal of medical ethics, 1996; 22: 292-299

Ethical theory, ethnography, and differencesbetween doctors and nurses in approaches topatient careDavid W Robertson University of Toronto Medical School, Toronto, Canada

AbstractObjectives-To study empirically whether ethicaltheory (from the mainstream principles-based,virtue-based, andfeminist schools) usefully describesthe approaches doctors and nurses take in everydaypatient care.Design-Ethnographic methods: participantobservation and interviews, the transcripts of whichwere analysed to identify themes in ethical approaches.Setting-A British old-age psychiatry ward.Participants-The more than 20 doctors and nurses onthe ward.Results-Doctors and nurses on the ward differed intheir conceptions of the principles of beneficence andrespect for patient autonomy. Nurses shared with doctorsa commitment to liberal and utilitarian conceptions ofthese principles, but also placed much greater weight onrelationships and character virtues when expressing thesame principles. Nurses also emphasised patientautonomy, while doctors were more likely to adovatebeneficence, when the two principles conflicted.Conclusion-The study indicates that ethical theorycan, contrary to the charges of certain critics, be relevantto everyday health care - if it (a) attends to socialcontext and (b) is flexible enough to draw on variousschools of theory.

I. Introduction and purposeCan mainstream, principles-based ethical theoryusefully describe the approaches doctors and nursestake to patient care? Many doctors and nurses, andeven some writers on health care morality,'-3 chargethat mainstream ethical theory is too removed fromlived reality to shed light on the way health careprofessionals do their work. This study explores thequestion empirically in a psychiatric ward, usingethnography. Ethnography is a sociological method-ology for exploring everyday life, in which theresearcher characteristically becomes a participant

Key wordsMedical ethics; ethnography; combining empiricalresearch and ethical theory.

observer, studying the sub-culture under examinationas a member of it.4 I observed the ward in the rolesfirst of junior nurse, and later of medical student.

I begin by describing two elements of the study'sfoundation: the aspects of health care observed, andthe nature of the ethical theory explored. First, inexamining empirically the ethical perspectives of staff,what aspects of health care are relevant to observe?Bioethics most typically focuses on such momentousquestions as abortion, euthanasia and withholdinglife-sustaining treatment. Certain writers in moralphilosophy5 and medical sociology,67 however, arguethe ethical significance of all aspects of health care,including less dramatic decisions, such as what dopatients eat for breakfast and how much opportunitydo they have to talk with their doctors. Such issuesaffect the quality of care for all patients, not only thosewhose cases imply life-and-death considerations. Thepresent study focuses on such everyday aspects ofcare; aspects largely neglected in the ethics literature.

Second, the theoretical approaches exploredincluded mainstream, principles-based health careethics, characteristically presented by Beauchampand Childress8 (who give a mostly liberal-utilitarianinterpretation of the principles) and Gillon9 (whoseapproach is more pluralist), and founded on three orfour principles (respect for patient autonomy,justice, beneficence and - often - non-maleficence).This dominant theory has incurred criticism fromproponents of casuistic, virtue-based, and feministethical theory, for its abstraction, focus on universal-ity, and paucity of terms for describing moralcharacter and the morality of relationships.Beauchamp and Childress8 have responded to thecritics by advocating a theoretical amalgam based ontheir four principles but including aspects of theircritics' theories, to reflect the range of moral frame-works at play in Western societies and even withinindividuals. The present study therefore alsoexplores the relevance of neo-Aristotelian virtuetheory with its account of personal character,'0 andfeminist theory, which focuses on relationships. "I

In sum, the study used ethnography to explorewhether an amalgam of ethical theories sheds usefullight on patient care. This article describes first the

copyright. on 8 S

eptember 2018 by guest. P

rotected byhttp://jm

e.bmj.com

/J M

ed Ethics: first published as 10.1136/jm

e.22.5.292 on 1 October 1996. D

ownloaded from

David W Robertson 293

literature on the ethical approaches of practitioners,then this study's methods, results, and recommen-dations for practice, education and further research.

II. Literature on the ethical approaches ofpractitionersResearch into the ethical approaches of health careprofessionals generally falls into two categories:attitudinal surveys, and ethnographic studies ofpractice. The present study aims to incorporatestrengths of both. The attitudinal surveys, whetherthrough questionnaires with simulated cases,12-14 orobserving role-playing exercises,'5 allow researchersto construct simulated situations structured accord-ing to theoretical ethics (as is also true of studies ofdecision-making ethics from clinical notes).'6 All ofthese methods, though, stand at one remove frompractice.

In contrast, ethnographic studies of practice' 17-19that address, or come close to addressing, healthcare ethics, while rare,20 ' have the advantage ofimmediacy to practice but the disadvantage that theygenerally (with the exception of Frohock and Foxand Swazey) lack even a passing explicit reference toa philosophical ethical framework. Jennings2'proposes uniting the ethical articulation that attitudesurveys can accommodate and the immediacy ofethnography, since "most writings in bioethics stilldraw faintheartedly and unimaginatively, or not atall, on social scientific studies that address thesetting, institutional context, and cultural forcesrelating to the bioethical problem at hand". Thepresent study takes up Jennings's challenge, employ-ing the following method.

III. MethodThe study's methodology combines an analyticalframework from ethics and ethnography'stechniques for studying everyday practice. I willexplain first the setting, second the ethnographicmethod of observation, and third the ethicalanalytical framework.

SettingThe study examined old-age psychiatry, which isrife with ethical tensions between professionalbeneficence and respect for patient autonomy. Themandate of the ward studied was to stabilise andassess patients from a socio-economically diversepopulation, prior to referral back to their homes orto an institution providing longer term care. Patientswere men and women over 65 years of age,diagnosed with either a depressive disorder ordementia. The staff-to-patient ratio was high:typically four nurses and between two and fourdoctors for a maximum of twelve patients. The ranksof both the medical and the nursing teams includedmen and women, and a range of ages.

ObservationDuring the ten weeks of ethnographic empiricalobservation, I worked 21 shifts on the ward, withdoctors (whom I accompanied, as a medical student,in their daily work, including interactions withpatients and professional meetings) and nurses (withwhom I worked under an honorary contract as anursing assistant). Fieldnotes of ethically relevantevents were taken during informal interviews,meetings of staff and observations of patient care.Events were defined as specific actions or verbalcommunications made by staff and indicatingethically laden decision-making and attitudes, forexample, a verbal exchange between two doctorsdebating the administration of medication against apatient's will, or a nurse's explanation that a patientshould be encouraged to brush her own teeth tomaximise self-sufficiency. In all, approximately 400events were recorded, about 20 per shift.

AnalysisOne of the tenets of ethnography is that theresearcher may begin the study with certain researchquestions, but ought not to impose a pre-setanalytical framework on the data: the frameworkshould arise from the data, as the researcher beginsto discern patterns in it. This study began with thequestion: can various schools of health care theoryusefully describe decision-making and attitudes on apsychiatric ward? Although I had surveyed theliterature on these ethical schools before beginningfieldwork, I waited until the fieldwork was under waybefore assembling the analytical framework in thefollowing manner. Throughout the fieldwork, Itested the relevance of the three bodies of theory byassessing which theory (or combination of theories)best described a sample of events. In this way, anoriginal framework aptly applicable to all the eventsevolved in tandem with the fieldwork. The followingexamples of events observed on the ward illustratehow feminist and virtue theory helped refine themainstream principle of beneficence, as reflected inthe final framework (see figure 1).

1. A doctor considering the treatment of apsychiatric out-patient who likely has cancer:

"I think she should be endoscoped. I've beencaught before thinking that it's a carcinoma when it'snot." A firm diagnosis would be helpful since if "shewon't be here next Christmas", ie, it is inoperable,palliative care should be envisioned: the "GP couldgive a bit of MST (morphine sulphate tablets) ... ifshe's in pain; a little MST might do the trick. But ifshe's going to live another ten years, you don't wanther addicted to morphine".

2. A nurse explaining his philosophy of care:"By the end of the day, a person has to make a

moral choice about (how much to care). Somethingto do with general outlook and receptivity, which I

copyright. on 8 S

eptember 2018 by guest. P

rotected byhttp://jm

e.bmj.com

/J M

ed Ethics: first published as 10.1136/jm

e.22.5.292 on 1 October 1996. D

ownloaded from

294 Ethical theory, ethnography, and differences between doctors and nurses in approaches to patient care

Figure 1: Ethicalframework adaptedfor the study

Commitment to justice- egalitarian- utilitarian

Commitment to beneficence- utilitarian- virtue- or relationship-based

Commitment to autonomy- rights-based- rationality-based- non-rationality abilities-based- relationship-based

Tension between principles- between autonomy and beneficence- between justice and beneficence- between justice and autonomy- between staff self-interest and the three principles

don't think can be encapsulated in a trainingschool." The nurse went on to talk about the impor-tance of "engaging" with the patient.

Both 1 and 2 express beneficence. In 1, the doctoris concerned with balancing the benefits of painpalliation and the risks of addiction: a typical utili-tarian deliberation. However, 2 includes anotherimportant facet of beneficence: that of professionalvirtue and the relationship between doctor andpatient, seen in the emphasis on moral attitude or

outlook, and engaging in relationships with patients.I therefore labelled 1 "utilitarian beneficence" (fromthe mainstream framework) and 2 "virtue- or

relationship-based beneficence" (reflecting virtueand feminist theory). Both labels figure as sub-categories in the adaptation of the four-principlesframework. I determined the rest of the study'sethical framework similarly; the result was the threeprinciples schema in figure 1.A methodological comment is necessary here for

readers familiar with ethnography. Ethnographicdata about a given social group typically give rise toan analytical framework that helps explain the groupand its attributes. In contrast, the present studyfitted elements of pre-existing ethical theory toethnographic data. Some ethnographers mighttherefore criticise the study on methodologicalgrounds, for not allowing the data to lead totheir own framework without reference to previoustheory. However, the different approach takenhere has the advantages of (a) putting existingethical theory to a systematic and practical test, and(b) systematically describing the ethical approachesof doctors and nurses. Also, the categories of theethical framework used to analyse the data were

determined only after the data were collected. Thecategories use ethical theory, but only aspects of theoryfound applicable to observations on the ward. Theorywas therefore not imposed on the data; the data andthe theory both informed the analytical framework.Once the fieldwork ended and the analytical

framework had been finalised, the fieldnotes were

transcribed, coded and analysed according to thecategories of the framework, using a content analysis

software package, HyperRESEARCH. The softwarecounted the number of events falling under eachcategory, generating frequency statistics and indicat-ing the salience of each category for doctors andnurses, respectively. No further statistical analysiswas performed. The frequency statistics ensure thatthe study's data were methodically collected and itsconclusions justified. The validity of any moresophisticated statistical analysis would be tenuous,given relatively low N values, and the fact thatethnographic data do not allow the same controlsthat, for example, survey methods permit. It alsoallowed easy reference to the text of the events. Thestudy's conclusions are therefore based on bothquantitative and qualitative data. The codingprocess was verified by two colleagues, along linesrecommended by Miles and Huberman22 andBurnard.23 Each colleague coded two observationtranscripts. I compared my colleagues' coding withmine and discussed with them any points where wediffered. Our coding accorded closely, except fortwo points which led me to adjust the analyticalframework to its present form. I also presented theresults and conclusions to individuals on the wardand to the group as a whole. Their responseconfirmed the validity of the findings, in that theyrang true with the experience of staff.

Before proceeding with the results, I will brieflydescribe a final aspect of the method. Ethnographyemphasises the relevance of social context to socialpatterns observed by the researcher.1 One can inferthat an ethnographic study of ethics on a hospitalward should provide an account of whether thesocial context of the ward was relevant to ethicalpatterns observed. Therefore, simultaneously to theethical case study, and using the same fieldnotes, Iconducted a sociological case study, for which Iconstructed a second analytical framework, drawingon the work of Guillemin and Holmstrom'7 andassessing, most importantly, the professional goals ofstaff. This article, for reasons of space, will discussonly the highlights of the sociological case study, todemonstrate that social context does indeed helpaccount for ethical patterns. The following twosections outline the principal results.

IV. ResultsThis section will first present the highlights of thesociological case study, and then, in more detail,those of the ethical case study. The most salientsociological results, described here for brevity's sakewithout corroborating qualitative and quantitativeevidence, pertained to differences between doctorsand nurses in their professional goals and attitudestowards technology. Doctors tended to put far moreemphasis than nurses on three goals: (i) clinicalproblem-solving (defined by the steps of assessinga situation, pinpointing a problem, then determiningand pursuing a remedy); (ii) assessing and

copyright. on 8 S

eptember 2018 by guest. P

rotected byhttp://jm

e.bmj.com

/J M

ed Ethics: first published as 10.1136/jm

e.22.5.292 on 1 October 1996. D

ownloaded from

David W Robertson 295

maximising organic function, and (iii) research.Nurses appeared to place a higher priority ondaily care and fostering patient normality andindependence. The sociological case study alsorevealed that the "technological imperative", ortendency to use technology (specifically drugs, elec-troconvulsive therapy, and diagnostic tests) andtechnological explanations rather than social orpsychological alternatives, was especially evidentamongst doctors on the ward. These results helpexplain the ethical findings, presented next.

This discussion of the ethical case study drawsconclusions about doctors' and nurses' commit-ments to justice, beneficence and patient autonomy,and about tensions between these principles.JUSTICE denotes fairness, desert, entitlement, andequitableness. Its prime relevance in health care is inthe allocation of scarce resources to needy patients,and the most common perspectives on justice areutilitarian, libertarian, communitarian, andegalitarian. In the study, although the utilitarian andcommunitarian perspectives may be discernedbehind National Health System policies, day-to-dayresource allocation within the ward was observed tobe discussed exclusively along egalitarian lines,concerned with the equal worth and entitlement ofpatients. (It is possible that tacit rationing alongutilitarian lines occurred, but I observed no evidenceof this.)

Doctors and nurses exercised this egalitarianapproach in an environment where resources wereclearly insufficient to respond to all patient needsand a "beds crisis" was a constant condition ofpractice, a phenomenon illustrated by the followingexchange between a doctor and a group of socialworkers from the ward's catchment area.

Doctor: The bed situation is very tight ... sixteenpatients waiting for purely financial [bed shortage]reasons. The directorate will really have to call acrisis meeting to decide what to do about this.Perhaps we could shuffle some beds.Social worker: We have been trying to negotiate withthe geriatrician.Second SW: This is very bad for patients.Doctor: The situation is just about getting togridlock [so] we should put people's names down fortreatment if they need it, but ... [doctor implies theunlikelihood that all the people whose names wereput down would actually receive care].

Commitment to BENEFICENCE was central forall staff. Figure 2 indicates that doctors and nurseswere each observed to value beneficence highly in alarge number of events observed on the ward (47 fordoctors; 56 for nurses). The comments and actionsof both doctors and nurses suggested they weremotivated to similarly high degrees by a utility-basedconception ofbeneficence (44 events for doctors; 40 fornurses). Events illustrating this conception conveyed

Doctors Nurses

Utility-based 44 40

Virtue-/relationship-based 3 16

Total 47 56

the idea of maximising utility in weighing benefits oftwo or more possible outcomes. The most commonexample was the weighing of risks and benefits ofover- and under-medicating patients in the pursuitof an ideal mean. However, a purely utilitarian con-ception of beneficence was not always adequate:many events involving beneficence were charac-terised by an emphasis not so much on achievinggood outcomes as on being a benevolent practitioner(an idea better described by virtue theory) and on

fostering good relationships with patients (a conceptelaborated in feminist relationship theory). Inobservations, virtue and relationship conceptions ofbeneficence effectively coincided. Virtue- andrelationship-based conceptions of beneficence were farmore often voiced amongst nurses (16 events) thandoctors (three events). One nurse implied such a

perspective on beneficence in the following accountof good nursing; note especially the mention ofattitudes and the reference to caring and strugglingalong with the patient.

"Many qualities needed to become a good nurse [asopposed to a good doctor] have more to do withcharacter and morality.... Somebody with a brightkind of attitude with regard to other human beingscould become a good nurse, no matter how un-

academically developed. If you don't care, if youdon't struggle ... you won't be very much use andwon't be a comfort to suffering people."

Engaging in close relationships with patients,however, was not always a matter simply of com-passion and affection. It sometimes also involvedtension and negative or mixed feelings about difficultpatients, an observation already made by Anspach.24A nurse related to me, for example, how one suchpatient had become "an emotional burden" to thenursing team, a perception that had contributed tothe nurses' decision to urge the doctors to refer thepatient to another ward.

RESPECT FOR PATIENT AUTONOMY alsooccupied a prominent place in the moral commit-ments of staff (figure 3). Both occupational groupswere highly committed to rights- and rationality-based views of autonomy, conceptions which are a

mainstay of liberal health care ethics. A commitmentto patient autonomy qua patient rights commonto both doctors (13 events) and nurses (22 events)

Figure 2: Conceptions ofbeneficence by occupational group

Number ofevents

copyright. on 8 S

eptember 2018 by guest. P

rotected byhttp://jm

e.bmj.com

/J M

ed Ethics: first published as 10.1136/jm

e.22.5.292 on 1 October 1996. D

ownloaded from

296 Ethical theory, ethnography, and differences between doctors and nurses in approaches to patient care

Figure 4: Conflict between autonomy and beneficence

Within occupational groups Number of events

Doctors 15

Nurses 24

Between occupational groups

Doctors advocating beneficence and nurses advocatingautonomy 8

Doctors advocating autonomy and nurses advocatingbeneficence

was evident, for example, in the following commentby a consultant during a ward round, where heconveys his respect for the right of patients of soundmind to make binding advance treatment directivesregarding non-psychiatric treatment:

"I was told on the telephone about an advancedirective [regarding resuscitation] which had beenwritten a few years ago [by a patient who at the timeof writing the directive had been undebatably ofsound mind] .... It has no legal validity but I think itshould be respected...."

A commitment to rationality-based autonomywas also shared by doctors (13 events) and nurses

(22 events). For instance, readiness for dischargehome, especially for patients recovering fromdepression, was often perceived to depend upon

rationality. With such patients, nurses persistentlyencouraged "appropriate" or "normal" -behaviour,by which terms they referred to a standard of rationalautonomy and implied a recovered capacity for thepatient to live outside a mental health institution.

Rationality, however, is a very strenuous criterionfor autonomy and, as Buchanan and Brock25 havenoted, there is a whole "complex of ideas givingcontent to the concept of individual self-determina-tion", or autonomy. Staff on the ward, especiallynurses (58 events, compared to six for doctors), alsohad a strong conception of autonomy as definedby abilities falling short of rational decision-making. For example, nurses painstakingly encour-

aged patients with dementia to be as independent as

possible in eating, washing, shaving, etc, with the

goal of maintaining maximal autonomy, long afterrational decision-making had become an impossiblestandard.

Consistent with Susan Sherwin's26 assertion thatGilligan'stt relationship ethics can often beintegrated with the principles of more traditionalapproaches to health care ethics, it was apparent on

the ward that nurses (34 events) were also stronglycommitted to patient autonomy as constitutedin relationships; doctors (three events), far less.For the nurses, patients' ability to be self-governing,indeed patients' personhood, appeared to dependon their ability to "meet another person", as onenurse put it. Perhaps a helpful image in explainingthis idea of relationship-based autonomy is that ofthe autistic person, largely incapable of formingrelationships, who is therefore in a crucial sense

not autonomous. This conception of autonomy isillustrated in a nurse's comment that one sexagenar-ian patient with rapidly progressing dementia,expressed his selfhood and individuality through"reaching out toward the other [other people] ...

[the patient] still does that: he's been able toexpress great tenderness and humour" despite hisdysphasia.

TENSION BETWEEN AUTONOMY ANDBENEFICENCE was the most salient pattern ofethical tension on the ward (see figure 4). Bothdoctors and nurses were observed to experience thistension, within and between occupational groups.Within occupational groups (15 events for doctors;24 for nurses), most of the observed situations were

manifestations of tension between rights-basedautonomy and utilitarian beneficence. This islogical, since rights protect entitlements of indivi-duals against intrusions (such as those motivated bybeneficence) from society or other individuals. Oneexample of such a situation was a doctor's delibera-tion over whether to admit an apparently suicidaland possibly depressed patient against the patient'swill, under the Mental Health Act provision foradmission without consent. The doctor was notwithout doubt that the patient was depressed, butdecided to admit her on the grounds that the risk tothe patient's health posed by depression justifiedplacing utilitarian beneficence above the patient'sautonomy rights.

In cases where nurses and doctors found them-selves divided along occupational lines in situationsof tension between autonomy and beneficence,nurses were more likely to be advocates ofautonomy; doctors, of beneficence. In suchsituations, nurses were observed on eight occasionsto be the prime advocates of autonomy, whiledoctors had this role on only one occasionobserved. The conflict was typically between rights-based autonomy and utility-based beneficence. Inone such case, a severely demented patient con-

sistently and violently resisted intimate care. Nurses

Figure 3: Conceptions ofautonomy by occupational group

Number ofeventsI~~~~~~~~~~~~~~~~~~~~

Doctors Nurses

Conception of autonomy

Rights-based 13 22

Rationality-based 13 22

Abilities-based 6 58

Relationship-based 3 34

Total 35 136

copyright. on 8 S

eptember 2018 by guest. P

rotected byhttp://jm

e.bmj.com

/J M

ed Ethics: first published as 10.1136/jm

e.22.5.292 on 1 October 1996. D

ownloaded from

David W Robertson 297

Figure 5: Tension between justice and beneficence

Within occupational groups Number ofevents

Doctors 19

Nurses 1

Between occupational groups 0

had to restrain her to wash, dress and toilet her. Thenurses, who felt "as ifwe're frog-marching her" werevery concerned that they were breaching an inherentright to autonomy, legal rights aside. Doctors weremuch less concerned, likely because they were at oneremove from actually carrying out the violation ofthe patient's autonomy.

All observed cases of TENSION BETWEENJUSTICE AND BENEFICENCE in both occupa-tional groups, involved an egalitarian conception ofjustice and a utilitarian conception of beneficencetowards individuals. I observed 19 such casesamongst doctors, 11 amongst nurses, and nonewhere doctors and nurses took opposing sides(figure 5).

For both doctors and nurses, the most typicalconflicts between justice and beneficence concernedthe ongoing "beds crisis": egalitarian justice requiredthem to be vigilant for opportunities to dischargepatients who had recovered, freeing their beds forothers. However, the resulting pressure likelymeant that on occasion patients were dischargedsooner, or admitted later (if at all), than they wouldhave been had there not been a resource shortage.Tensions between justice and autonomy andbetween staff self-interest and the three principleswere observed too seldom to draw substantiveconclusions.

VI. DiscussionThe study's ethnographic methodology proved anapt means of examining everyday ethics.Ethnographically-gathered data confirmed, forthe ward in question, Sedgwick's and Fulford'stheoretical conclusion that all aspects of health care,even the most uncontroversial, were morallycharged. It effectively explored the study'shypothesis that ethical theory can shed light on theapproaches of doctors and nurses to patient care.Indeed, the combination of ethnography and ethicaltheory identified important differences betweendoctors and nurses in their approaches to patientcare. These differences paralleled the study'ssociological results, which reveal that members ofthe two professions, on the ward studied, emphasisedifferent professional goals.The sociological results help explain the ethical

finding that, while both doctors and nurses valuedutility-based BENEFICENCE, nurses more than

doctors also tended to express commitment tovirtue- and relationship-based beneficence. Theprime professional goals of nurses were daily careand helping patients live as normally and indepen-dently as possible; these goals were pursued throughongoing relationships whose sustenance demandedthe demonstration of character virtues. Doctors'most important goals were systematic problem-solving, improving organic function, and research;commitments that emphasise beneficial consequencesand fit more readily into an unadulterated utilitarianmould. The doctors' "technological imperative" alsoconforms to the consequentialist, utilitarian model:the technologies' goal was improved function,regardless of virtue or relationships. Differentprofessional roles, therefore, appear to account herefor differences in ethical approaches. It should benoted, however, that, although doctors spoke lessoften of virtue, one cannot necessarily infer thatdoctors have a weaker conception of benevolence asa characteristic of a virtuous doctor: it may be thatdoctors are highly motivated by such a conceptionbut that their socialisation process, directed alongscientific, outcome-based lines, teaches them tomask this conception with utilitarian expressions ofbeneficence.

Also, the closer relationships at the root of nurses'apparently different conception of beneficence neednot always signify superior concern for patientwelfare: nurses' frustration with problem-patients onthe ward sometimes contributed to the precipitatetransfer of patients or led nurses to avoid patients.Doctors maintained more distance from patientsand correspondingly expressed less frustration withdifficult ones. Thus there appear to be beneficence-related advantages and disadvantages both ofnurses'close relationships with patients, and of doctors' dis-engagement. Nurses can foster patients' social well-being and are attuned to their individual needs,while doctors are assiduous in trying to solveproblems of dysfunction; nurses risk makingdecisions coloured by frustration with problem-patients, while doctors risk lacking sensitivity to indi-vidual cases. The advantage of the doctors'perspective is the converse of the disadvantage of thenurses', and vice versa, suggesting that open com-munication and shared decision-making betweennurses and doctors is in patients' best interests.The sociological case study also sheds light on the

results regarding RESPECT FOR PATIENTAUTONOMY. Nurses' relative emphasis ofautonomy based on abilities other than rationalitycan be partly explained by the fact that nurses' mostsalient professional goal, patient normality andindependence, is defined by patients' abilities,including, but never limited to, rationality. Similarly,relationships between carers and patients, and there-fore relationship-based conceptions of autonomy,are far more likely to be built through daily care -another of the nurses' characteristic professional

copyright. on 8 S

eptember 2018 by guest. P

rotected byhttp://jm

e.bmj.com

/J M

ed Ethics: first published as 10.1136/jm

e.22.5.292 on 1 October 1996. D

ownloaded from

298 Ethical theory, ethnography, and differences between doctors and nurses in approaches to patient care

goals - than through the more episodic patientcontact typical of doctors.

Also, the observation that, when TENSIONBETWEEN BENEFICENCE AND RESPECTFOR PATIENT AUTONOMY involved doctorson one side and nurses on the other, nurses seemedmore likely to support autonomy (typically rights-based), while doctors in such situations usuallyadvocated beneficence (typically utility-based), maybe explained by nurses' closer relationships andcloser identification with patients. Doctors, on theother hand, appeared most concerned with problem-solving and improving organic function, consequen-tialist goals sometimes inconsistent with rights-basedautonomy.The germ of this study was scepticism among

health practitioners and academics that ethicaltheory, especially the principles-based mainstream,can usefully describe the approaches of doctors andnurses to patient care. The study is original inmelding ethnography and ethical theory to explorethe relevance of the various theoretical approaches,empirically. The results confirm that ethical theorycan shed light on the moral considerations present inthe everyday provision of health care. Specifically,these results usefully describe the ethical approachesof doctors and nurses and, moreover, significantdifferences between the occupational groups. Inconclusion, principles-based ethical theory can, ifsupplemented by feminist and virtue theory, describeethics in practice, and thus has potential to helpdoctors and nurses articulate and refine their ownethical values, as well as understand those of others.The following recommendations suggest that ethicaltheory tested empirically for practical relevancecould lead health-care professionals to heightenedawareness of the ethical nature of their work, a moredemocratic approach to addressing practical moraldilemmas, and, ultimately, better clinical decision-making.

VII. Considerations for doctors, nursesand researchersFor practice: Decision-making by doctors and nursescan stand to profit from an understanding of theethical priorities shared by the two groups, and of theways in which the groups' priorities diverge. Such anunderstanding could give doctors and nurses insightinto each others' strengths and weaknesses asethical-decision makers, and improve their ability tocomplement each other in making decisions,ultimately to the benefit of the patient.For teaching: The study reveals (a) that everydaypractice abounds in ethically charged situations, and(b) that health care professionals are discussing theseissues, even if not in explicitly ethical terms.Everyday practice can therefore serve as a richenvironment for medical and nursing students inwhich to learn ethics. Educators should take

advantage of this, in keeping also with other work27 28suggesting the relevance of real cases and students'clinical experience in ethics education.For future research: Future ethnographic-ethicalresearch could build on the findings of this study bypursuing the same research questions using similarmethodology in fields other than psychiatry, with aview to drawing more generally applicable conclu-sions. It would also be useful to seek to identifyoptimal conditions for the resolution of ethicaltensions in health care settings.

AcknowledgementsI am grateful to the staff and patients on the wardand to my MPhil. supervisors, Professor K W MFulford, Ms J Garcia, and Professor S Ringen, fortheir helpful and extensive comments on theresearch, which was supported by a joint scholarshipfrom the Social Science and Humanities ResearchCouncil of Canada and the Natural Science andEngineering Research Council of Canada.

David W Robertson, BASc, MPhil, was a master'sstudent in Comparative Social Research at theUniversity of Oxford while conducting this study. Atpresent he is a medical student at the University ofToronto, Canada.

References1 Fox R, Swazey J. Medical morality is not bioethics -

medical ethics in China and the United States.Perspectives in Biology and Medicine 1984; 27, 3:336-60.

2 Baron RJ. Dogmatics, empirics, and moral medicine.Hastings Center Report 1989; 19, 1: 41.

3 Hoffmaster B. Can ethnography save the life of medicalethics? Social Science and Medicine 1992; 35:1421-31.

4 I would refer readers unfamiliar with ethnography toLincoln Y, Guba E. Naturalistic inquiry. Beverly Hills:Sage, 1985.

5 Fulford K W M. Moral theory and medical practice.Cambridge: Cambridge University Press, 1989.

6 Reiser SJ, Bursztajn HJ, Appelbaum PS, Gutheil TG.Divided staffs, divided selves: a case approach to mentalhealth ethics. Cambridge: Cambridge UniversityPress, 1987.

7 Sedgwick P. Psycho politics. London: Pluto Press, 1982.8 Beauchamp TL, Childress JF. Principles of biomedical

ethics [4th ed]. New York: Oxford University Press,1994.

9 Gillon R. Philosophical medical ethics. Chichester: JohnWiley and Sons, 1985.

10 See Macintyre A. After virtue: a study in moral theory.London: Duckworth, 1981.

11 See Gilligan C. In a different voice: psychological theoryand women's moral development. Cambridge,Massachusetts: Harvard University Press, 1982.

12 Sulmasy DP, Geller G, Levine D, Faden R. Medicalhouse officers' knowledge, attitudes, and confidenceregarding medical ethics. Archives of Internal Medicine1990; 150: 2509-13.

copyright. on 8 S

eptember 2018 by guest. P

rotected byhttp://jm

e.bmj.com

/J M

ed Ethics: first published as 10.1136/jm

e.22.5.292 on 1 October 1996. D

ownloaded from

David W Robertson 299

13 Sulmasy DP, Geller G, Levine D, Faden R. A random-ized trial of ethics education for medical houseofficers. Journal of Medical Ethics 1993; 19: 157-63.

14 Hebert P, Meslin E, Dunn E, Byrne N, Ried S.Evaluating ethical sensitivity in medical students:using vignettes as an instrument. Journal of MedicalEthics 1990; 16: 141-5.

15 Mitchell KR, Myser C, Kerridge IH. Assessing theclinical ethical competence of undergraduate medicalstudents. Journal of Medical Ethics 1993; 19: 230-6.

16 Sulmasy DP, Geller G, Faden R, Levine D. The qualityof mercy: caring for patients with 'do no resuscitate'orders. Journal of the American Medical Association1992; 268, 4: 471-2.

17 Guillemin JH, Holmstrom LL. Mixed blessings: intensivecare for newborns. New York: Oxford University Press,1986.

18 Anspach RR. Deciding who lives: fateful choices in theintensive care nursery. Berkeley: University ofCalifornia Press, 1993.

19 Frohock, FM. Special care: medical decisions at thebeginning oflife. Chicago: University of Chicago Press,1986.

20 Lieban RW. Medical anthropology and the compara-tive study of medical ethics. In: Weisz G, ed. Socialscience perspectives on medical ethics. Philadelphia:University of Pennsylvania Press, 1991.

21 Jennings B. Ethics and ethnography in neonatalintensive care. Social science perspectives on medicalethics. Philadelphia: University of Pennsylvania Press,1991.

22 Miles MB, Huberman AM. Qualitative data analysis: anexpanded sourcebook [2nd ed]. Thousand Oaks: Sage,1994.

23 Burnard P. A method of analysing interview transcriptsin qualitative research. Nurse Education Today 1991;11, 6: 461-6.

24 Anspach RR. Prognostic conflict in life-and-death deci-sions: the organization as an ecology of knowledge.Journal of Health and Human Behaviour 1987; 28, 3:215-31.

25 Buchanan AE, Brock DW. Decidingfor others: the ethicsof surrogate decision making. Cambridge: CambridgeUniversity Press, 1989.

26 Sherwin S. No longer patient: feminist ethics and healthcare. Philadelphia: Temple University Press, 1992.

27 Osborne LW, Martin CM. The importance of listeningto medical students' experience when teaching themmedical ethics. journal of Medical Ethics 1989; 15:35-8.

28 Hope A, Fulford KWM. Medical education: patients,principles and practice skills. In: Gillon R, ed.Principles of health care ethics. Chichester: John Wileyand Sons, 1994.

News and notes

Gene therapy - Current Clinical Status

Henry Stewart Conference studies are presenting a confer-ence on Gene Therapy - Current Clinical Status onTuesday 29 October 1996 at The Cafe Royal, LondonWI.

This one-day course has been designed as an inten-sive update on the results so far in clinical trials of genetherapy for various diseases. The speakers will addressthe issues that have arisen to date in the clinical trials,what has been learnt and where they expect progresswill be made in the future.

The event will be chaired by Professor Karol Sikoraand the international panel of speakers includes ProfessorJack Roth (University of Texas M D Anderson CancerCenter), Professor Angus Dalgleish (St George'sHospital Medical School) and Professor Jonathan Weber(Imperial College of Medicine at St Mary's).Further enquiries to Nicola McCall at Henry Stewart

Conference Studies on: Tel: +44 (0) 171 404 3040;Fax: +44 (0) 171 404 2081; e-mail:100622.3264@compuserve.com

copyright. on 8 S

eptember 2018 by guest. P

rotected byhttp://jm

e.bmj.com

/J M

ed Ethics: first published as 10.1136/jm

e.22.5.292 on 1 October 1996. D

ownloaded from