Factors related to the involvement of nurses in medical end-of-life decisions in Belgium: a death certificate study

BACKGROUND: Although nurses play an important role in end-of-life care for patients, they are not systematically involved in end-of-life decisions with a possible or certain life-shortening effect (ELDs). Until now we know little about factors relating to the involvement of nurses in these decisions. OBJECTIVE: To explore which patient- and decision-characteristics are related to the consultation of nurses and to the administering of life-ending drugs by nurses in actual ELDs in institutions and home care, as reported by physicians. METHOD: We sampled at random 5005 of all registered deaths in the second half of 2001--before euthanasia was legalized--in Flanders, Belgium. We mailed anonymous questionnaires to physicians who signed the death certificates and asked them to report on ELDs, including nurses' involvement. RESULTS: Response rate was 59% (n=2950). Physicians reported nurses involved in decision making more often in institutions than at home, and more often in care homes for the elderly than in hospitals (OR 1.70, 95% CI 1.15, 2.52). This involvement was more frequently when physicians intended to hasten the patient's death than when they had no such intention (institutions: OR 2.05, 95% CI 1.41, 2.99; home: OR 2.04, 95% CI 1.19, 3.49). In institutions, this involvement was also more likely where patients were of lower rather than higher education (OR 2.95, 95% CI 1.49, 5.84). The administering of life-ending drugs by nurses, as reported by physicians was also found more frequently in institutions than at home, and in institutions more frequently with lower rather than higher educated patients (p=.037). CONCLUSIONS: These findings raise questions about physicians' perception of the nurse's role in ELDs, but also about physicians' skills in interacting with all patients. Education and guidelines for physicians and nurses are needed to optimize good communication and to promote a clearer assignment of responsibilities concerning the execution of those decisions.     

Nurses' involvement in end-of-life decisions

In Dutch healthcare, patients and physicians are responsible for medical end-of-life decisions. These include nontreatment decisions, withholding or withdrawing parenteral hydration and nutrition where the patient can no longer drink or eat, relieving pain and other symptoms with drugs that might shorten life, euthanasia, and physician-assisted suicide. The objective of this study is to investigate the views of nurses on their role with regard to discussing these kinds of decisions, as well as their actual role in the decision-making process. Nurses (n = 489) involved in palliative terminal care answered questions about their preferred and actual role in end-of-life decision-making processes. Nurses want to be involved in making end-of-life decisions, but this depends partly on the kind of decision that has to be taken. When caring for terminal patients, 62% of the nurses usually talk about such decisions with patients or their families. Three-quarters of the nurses had been involved in an end-of-life decision-making process in the previous 2 years, mostly by talking with the physician and the patient's family. It is concluded that physicians should discuss these decisions with nurses more often. The finding that characteristics of the nurses influence their role in end-of-life decision-making processes emphasizes the importance of developing mechanisms to ensure that end-of-life decisions are made in a consistent manner and do not depend on the demographic characteristics of nurses.     

Attitudes of nurses towards euthanasia and towards their role in euthanasia: A nationwide study in Flanders,Belgium

BackgroundNurses have an important role in caring for terminally ill patients. They are also often involved in euthanasia. However, little is known about their attitudes towards it.ObjectivesTo investigate on a nationwide level nurses’ attitudes towards euthanasia and towards their role in euthanasia, and the possible relation with their socio-demographic and work-related characteristics.Design and participantsA cross-sectional design was used. In 2007, a questionnaire was mailed to a random sample of 6000 of the registered nurses in Flanders, Belgium. Response rate was 62.5% and after exclusion of nurses who had no experiences in patient care, a sample of 3321 nurses remained.MethodsAttitudes were attained by means of statements. Logistic regression models were fitted for each statement to determine the relation between socio-demographic and work-related characteristics and nurses’ attitudes.ResultsNinety-two percent of nurses accepted euthanasia for terminally ill patients with extreme uncontrollable pain or other distress, 57% accepted using lethal drugs for patients who suffer unbearably and are not capable of making decisions. Seventy percent believed that euthanasia requests would be avoided by the use of optimal palliative care. Ninety percent of nurses thought nurses should be involved in euthanasia decision-making. Although 61% did not agree that administering lethal drugs could be a task nurses are allowed to perform, 43% would be prepared to do so. Religious nurses were less accepting of euthanasia than non-religious nurses. Older nurses believed more in palliative care preventing euthanasia requests and in putting the patient into a coma until death as an alternative to euthanasia. Female and home care nurses were less inclined than male and hospital and nursing home nurses to administer lethal drugs.ConclusionsThere is broad support among nurses for euthanasia for terminally ill patients and for their involvement in consultancy in case of euthanasia requests. There is, however, uncertainty about their role in the performance of euthanasia. Guidelines could help to make their role more transparent, taking into account the differences between health care settings.     

Nurses' involvement in 'do not resuscitate' decisions on acute elder care wards

AIM: This paper reports the involvement of nurses in 'do not resuscitate' decision-making on acute elder care wards and their adherence to such decisions in the case of an actual cardiopulmonary arrest. BACKGROUND: Previous literature showed that nurses are involved in half or less than half of 'do not resuscitate' decisions in hospitals, but their involvement in this decision-making on acute elder care wards in particular has not been investigated. METHOD: A questionnaire was sent in 2002 to the head nurses of all acute elder care wards in Flanders, Belgium (n = 94). They were asked whether nurses had been involved in the last 'do not resuscitate' decision-making process on their ward and whether nurses 'never', 'rarely', 'sometimes', 'often' or 'always' started resuscitation in case of cardiopulmonary arrest of patients with 'do not resuscitate' status and of those without. RESULTS: The response rate was 86.2% (n = 81). In 74.7% of the last 'do not resuscitate' decisions on acute elder care wards in Flanders, a nurse was involved in the decision-making process. For patients with 'do not resuscitate' status, 54.3% of respondents reported that cardiopulmonary resuscitation was 'never' started on their ward, 'rarely' on 39.5% and 'sometimes' on 6.2%. For patients without 'do not resuscitate' status, nurses started cardiopulmonary resuscitation 'rarely' or 'sometimes' on 22.2% of all wards, and 'often' or 'always' on 77.8%. CONCLUSION: To make appropriate 'do not resuscitate' decisions and to avoid rash decision-making in cases of actual cardiopulmonary arrest, nurses should be involved early in 'do not resuscitate' decision-making. If institutional 'do not resuscitate' guidelines were to stress more clearly the important role of nurses in all kinds of end-of-life decisions, this might improve the 'do not resuscitate' decision-making process.     

Euthanasia and physician-assisted suicide in the Dutch homecare sector: the role of the district nurse

AIM: This paper is a report of the findings of a study into the role of district nurses in euthanasia and physician-assisted suicide in homecare organizations, conducted as part of a study into the role of nurses in medical end-of-life decisions. BACKGROUND: Issues concerning legislation and regulation with respect to the role of nurses in euthanasia and physician-assisted suicide gave the Minister for Health reason to commission a study into the role of nurses in medical end-of-life decisions in hospitals, nursing homes and homecare organizations. This is the first quantitative study from the perspective of nurses. Previous quantitative studies were conducted under physicians and information on the role of nurses was obtained indirectly. METHOD: A questionnaire was sent in 2003 to 500 district nurses employed in 55 homecare organizations. The absolute response rate was 86.0% and 81.6% (408) could be used for analysis. RESULTS: In 22.3% of 278 cases, the district nurse was the first with whom patients discussed their request for euthanasia or physician-assisted suicide. In about half (49.8%) of 267 cases nurses were not involved in the general practitioner's decision-making process, and in only 13.3% of 264 cases, did they attend the administration of the lethal drugs. District nurses had provided some degree of aftercare to the surviving relatives in 80.3% of 264 cases. CONCLUSION: Collaboration between general practitioners and district nurses needs improvement, particularly in relation to decision-making. Our Dutch data could help nurses in other countries to define their (future) role in euthanasia and physician-assisted suicide.     

Advance Care Planning and Dying in Nursing Homes in Flanders,Belgium: A Nationwide Survey

ContextIn Belgium, data on actual advance care planning (ACP) in nursing homes (NHs) are scarce.ObjectivesTo investigate the prevalence and characteristics of documented advance directives and physicians' orders for end-of-life care in NHs, and the authorization of a legal representative in relation to the residents' demographic and clinical characteristics and care received.MethodsThis was a retrospective cross-sectional study, including all NH residents deceased during September and October 2006 in all 594 NHs in Flanders, Belgium. Structured mail questionnaires about the resident's characteristics, hospital transfers, palliative care delivery, ACPs, and authorization of legal representatives were completed via the NH administrators and nurses involved in the care of the resident.ResultsAdministrators of 318 NHs (53.5%) reported 1303 deaths. Nurses provided information about 1240 (95.2%) of these deaths. At the end of life, NH residents often had dementia (65.2%) and were severely dependent (76.1%). Almost half (43.1%) had at least one hospital transfer during the last three months of life and two-thirds received palliative care. Half had an ACP, predominantly a physician's order and less often an advance directive. Having advance directives or physician's orders was associated with receiving palliative care. Residents with a physician's order more often died in the NH. Nine percent had an authorized legal representative.ConclusionPrevalence of ACPs and formal authorization of a legal representative was low among the deceased NH residents in Flanders, Belgium. There was a higher prevalence of physicians' orders, often established after the resident had lost capacity. Initiatives should be developed to stimulate more advance discussion on care options and making end-of-life decision with the residents while they retain capacity.     

Continuous deep sedation until death in Belgium: a survey among nurses

ContextContinuous deep sedation (CDS) is a subject of important debate, but until now nurses have rarely been questioned about their involvement and perceptions.ObjectivesTo study the communication process between nurses and patients, relatives, or physicians before starting CDS, and how nurses perceive this end-of-life practice.MethodsIn 2007, we surveyed 1678 nurses in Flanders, Belgium, who, in an earlier survey, had reported caring for one or more patients who received an end-of-life decision within the previous year. Nurses were surveyed about their most recent case.ResultsThe response rate was 75.8%: 250 nurses reported a case of CDS (64.4% hospital, 18.4% home, and 17.2% nursing home). In, respectively, 25.8% and 75.4%, the patient and relatives had communicated with the nurse about the CDS. In 17.6%, there was no communication between the nurse and the physician about the CDS; in 29.1%, the physician and nurse only exchanged information; and in 23.4%, they made the decision jointly. Making the decision jointly was associated with a more positive evaluation of the cooperation with the physician (adjusted odds ratio 10.9 and 95% confidence interval 3.0, 39.2). Nurses perceived CDS as partly intended to hasten death partially in 48.4% and explicitly in 28.4% of cases, estimating possible or certain life shortening in 95.6%.ConclusionNurses in different health care settings are often involved in communication about CDS. They see it mainly as a practice intended to hasten death, with a life-shortening effect; guidelines should recommend clear discussions between caregivers in which the physician states the purpose and estimated effect of the decision.     

Do-not-resuscitate decisions in six European countries

OBJECTIVE: To study and compare the incidence and main background characteristics of do-not-resuscitate (DNR) decision making in six European countries. DESIGN: Retrospective. SETTING: We studied DNR decisions simultaneously in Belgium (Flanders), Denmark, Italy (four regions), the Netherlands, Sweden, and Switzerland (German-speaking part). In each country, random samples of death certificates were drawn from death registries to which all deaths are reported. The deaths occurred between June 2001 and February 2002. PARTICIPANTS: Reporting physicians received a mailed questionnaire about the medical decision making that had preceded death. The response percentage was 75% for the Netherlands, 67% for Switzerland, 62% for Denmark, 61% for Sweden, 59% for Belgium, and 44% for Italy. The total number of deaths studied was 20,480. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Measurements were frequency of DNR decisions, both individual and institutional, and patient involvement. Before death, an individual DNR decision was made in about 50-60% of all nonsudden deaths (Switzerland 73%, Italy 16%). The frequency of institutional decisions was highest in Sweden (22%) and Italy (17%) and lowest in Belgium (5%). DNR decisions are discussed with competent patients in 10-84% of cases. In the Netherlands patient involvement rose from 53% in 1990 to 84% in 2001. In case of incompetent patients, physicians bypassed relatives in 5-37% of cases. CONCLUSIONS: Except in Italy, DNR decisions are a common phenomenon in these six countries. Most of these decisions are individual, but institutional decisions occur frequently as well. In most countries, the involvement of patients in DNR decision making can be improved.     

The use of drugs with a life-shortening effect in end-of-life care in neonates and infants

Objective The purpose was to describe the use of drugs with a possible or certain life-shortening effect in end-of-life care in infants and to evaluate the possibly lethal effect.Design For 292/298 deaths of live born infants (<1 year), in a 1-year period (between 1 August 1999 and 31 July 2000) in Flanders, Belgium, the attending physician could be identified and was sent an anonymous questionnaire. The questionnaires relating to deaths directly preceded by the administration of drugs were reviewed by a multi-disciplinary panel.Results The response rate was 86.6% (253/292). In 57 cases (22.5%), drugs were administered directly before death. In 17/57 cases, the physician explicitly intended to hasten death. In 16/17 cases information about the drug(s) was available: opioids were administered in 14, a muscle relaxant in 5 and potassium chloride in 3 cases. In 13 cases where the lethal effect could be evaluated, the panel judged that the drugs were effective in hastening death in 10 cases. In most cases the estimated life-shortening was <24 h. In 40/57 cases the physician administered drugs to alleviate pain and/or symptoms, taking into account a possible life-shortening effect without explicitly intending it. Opioids were administered in all 30 cases where information about the drug(s) was supplied. In 13 cases the lethal effect could be evaluated, and in 6 cases the panel judged that the drugs had hastened death.Conclusions When life-shortening was explicitly intended, (dosages of) drugs were likely to be lethal. Drugs administered also clearly hastened death in some cases where life-shortening was not explicitly intended.     

A qualitative analysis of ethical problems experienced by physicians and nurses in intensive care units in Turkey

In this qualitative study, we aimed to identify and compare the ethical problems perceived by physicians and nurses in intensive care units at Baskent University hospitals in Turkey. A total of 21 physicians and 22 nurses were asked to describe ethical problems that they frequently encounter in their practice. The data were analyzed using an interactive model. The core problem for both physicians and nurses was end-of-life decisions (first level). In this category, physicians were most frequently concerned with euthanasia while nurses were more concerned with do-not-resuscitate orders (second level). At the third level, we saw that almost all of the participants' responses related to negative perceptions about euthanasia. Communication and hierarchical problems were the second most reported main category. Nurses were more likely to cite problems with hierarchy than physicians. At the third level, a large percentage of nurses described communication problems with authority and hierarchical problems with physicians. In the same category, physicians were most often concerned with communication problems with patients' relatives. The ethical problems were reported at different frequencies by physicians and nurses. We asked the participants about ethical decision-making styles. The results show that nurses and physicians do not follow a systematic pattern of ethical decision making.     

Using potentially life-shortening drugs in neonates and infants

OBJECTIVE: To describe the frequency, background, and impact of decisions to give analgesic or other drugs that may, intentionally or unintentionally, shorten the life-span of severely ill neonates. SETTING: The Netherlands. DESIGN: Retrospective, cross-sectional study. PATIENTS: Questionnaires were mailed in The Netherlands to physicians reporting 338 consecutive deaths of infants under 1 yr of age from August through November 1995. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Questions were asked about medical end-of-life decisions preceding the death of the infant and about the decision-making process. Potentially life-shortening drugs, mostly opioids, were given in 37% of all deaths. The estimated effect in terms of the shortening of life was <1 wk in 72% of all patients in whom the administration of potentially life-shortening drugs had been the most important end-of-life decision. Most decisions to administer such drugs were discussed with parents and colleagues. The decisions were discussed regarding virtually all patients in whom the physician had intended to hasten death; doses of opioids tended to be larger in this group. CONCLUSIONS: The frequency with which drugs that may shorten life are administered before the death of severely ill infants confirms the important role of modern medicine in dying in neonatology. Most physicians caring for neonates feel that palliative medication may be warranted in dying infants, even if it shortens life. A distinction between intentionally ending life and providing adequate terminal care by alleviating pain or other symptoms, which is important in moral and judicial terms, is probably not easily made for some of these patients.     

Forgoing artificial nutrition or hydration in patients nearing death in six European countries

Whether or not artificial nutrition or hydration (ANH) may be forgone in terminally ill patients has been the subject of medical and ethical discussions. Information about the frequency and background characteristics of making decisions to forgo ANH is generally limited to specific clinical settings. The aim of this study was to compare the practice of forgoing ANH in six European countries: Belgium, Denmark, Italy, The Netherlands, Sweden, and Switzerland. In each country, random samples were drawn from death registries. Subsequently, the reporting physician received a questionnaire about the medical decisions that preceded the patient's death. The total number of deaths studied was 20,480. The percentage of all deaths that were preceded by a decision to forgo ANH varied from 2.6% in Italy to 10.9% in The Netherlands. In most countries, decisions to forgo ANH were more frequently made for female patients, patients aged 80 years or older, and for patients who died of a malignancy or disease of the nervous system (including dementia). Of patients in whom ANH was forgone, 67%-93% were incompetent. Patients in whom ANH was forgone did not receive more potentially life-shortening drugs to relieve symptoms than other patients for whom other end-of-life decisions had been made. Decisions to forgo ANH are made in a substantial percentage of terminally ill patients. Providing all patients who are in the terminal stage of a lethal disease with ANH does not seem to be a widely accepted standard among physicians in Western Europe.     

Nurse involvement in end-of-life decision making: the ETHICUS Study

Objective The purpose was to investigate physicians perceptions of the role of European intensive care nurses in end-of-life decision making.Design This study was part of a larger study sponsored by the Ethics Section of the European Society of Intensive Care Medicine, the ETHICUS Study. Physicians described whether they thought nurses were involved in such decisions, whether nurses initiated such a discussion and whether there was agreement between physicians and nurses. The items were analyzed and comparisons were made between different regions within Europe.Setting The study took place in 37 intensive care units in 17 European countries.Patients and participants Physician investigators reported data related to patients from 37 centers in 17 European countries.Interventions None.Measurements and results Physicians perceived nurses as involved in 2,412 (78.3%) of the 3,086 end-of-life decisions (EOLD) made. Nurses were thought to initiate the discussion in 66 cases (2.1%), while ICU physicians were cited in 2,438 cases (79.3%), the primary physician in 328 cases (10.7%), the consulting physician in 105 cases (3.4%), the family in 119 cases (3.9%) and the patient in 19 cases (0.6%). In only 20 responses (0.6%) did physicians report disagreement between physicians and nurses related to EOLD. A significant association was found between the region and responses to the items related to nursing. Physicians in more northern regions reported more nurse involvement.Conclusions Physicians perceive nurses as involved to a large extent in EOLDs, but not as initiating the discussion. Once a decision is made, there is a sense of agreement. The level of perceived participation is different for different regions.This article refers to the editorial .     

Euthanasia and assisted suicide in Dutch hospitals: the role of nurses

Aim. To report a study on the role of nurses in euthanasia and physician‐assisted suicide in hospitals, conducted as part of a wider study on the role of nurses in medical end‐of‐life decisions. Background. Issues concerning legislation and regulation with respect to the role of nurses in euthanasia and physician‐assisted suicide gave the Dutch Minister for Health reason to commission a study on the role of nurses in medical end‐of‐life decisions in hospitals, homecare and nursing homes. Method. A questionnaire was sent in 2003 to 692 nurses employed in 73 hospital locations. The response suitable for analysis was from 532 (76·9%) nurses. Data were quantitatively analysed using spss version 11.5 for Windows. Results. In almost half of the cases (45·1%), the nurse was the first with whom patients discussed their request for euthanasia or physician‐assisted suicide. Consultations between physicians and nurses quite often took place (78·8%). In several cases (15·4%), nurses themselves administered the euthanatics with or without a physician. It is not self‐evident that hospitals have guidelines concerning euthanasia/physician‐assisted suicide. Conclusions. In the decision‐making process, the consultation between the physician and the nurse needs improvement. In administering the euthanatics, physicians should take responsibility and should not leave these actions to nurses. Guidelines may play an important role to improve the collaboration between physicians and nurses and to prevent procedural, ethical and legal misunderstandings. Relevance to clinical practice. Nurses in clinical practice are often closely involved in the last stage of a person's life. Consequently, they are often confronted with caring for patients requesting euthanasia or physician‐assisted suicide. The results provide relevant information and may help nurses in defining their role in euthanasia and physician‐assisted suicide, especially in case these practices should become legalised.     

Knowledge of,and participation in,advance care planning: A cross-sectional study of acute and critical care nurses’ perceptions

BackgroundNurses have a core role in facilitating discussions and enacting decisions about end-of-life issues for patients in hospitals. Nurses’ own knowledge and attitudes may influence whether they engage in meaningful end-of-life conversations with patients.AimsTo determine in a sample of nurses working in acute and critical care hospital wards:1) their knowledge of advance care planning, including the authority of substitute decision-makers and legal validity of advance directives;2) their own participation in advance care planning decision-making practices; and3) associations between nurses’ socio-demographic characteristics; clinical expertise; and knowledge and behaviour in relation to advance care planning practices.DesignQuestionnaire-based, cross-sectional study.Setting and participantsThe study was conducted with 181 registered and enrolled nurses employed in acute and critical care wards of three metropolitan hospitals in Australia.ResultsNurses were least knowledgeable about items relating to the authority of medical (56%) and financial (42%) substitute decision-makers. Few nurses had prepared advance directives (10%) or appointed medical (23%) or financial (27%) decision-makers, when compared to discussing end-of-life wishes (53%) or organ donation (75%). Overall, 15% of nurses had not engaged in any advance care planning practices. Nurses who had cared for 11–30 dying patients in the last six months were more likely to have an increased knowledge score. Older nurses were more likely to participate in a greater number of advance care planning practices and an increase in shifts worked per week led to a significant decrease in nurses’ participation.ConclusionNurses have a key role in providing advice and engaging dying patients and their families in advance care planning practices. Nurses’ own knowledge and rates of participation are low. Further education and support is needed to ensure that nurses have an accurate knowledge of advance care planning practices, including how, when and with whom wishes should be discussed and can be enacted.     

Using drugs to end life without an explicit request of the patient

A small proportion of deaths result from the use of drugs with the intention to hasten death without an explicit request of the patient. Additional insight into its characteristics is needed for evaluating this practice. In the Netherlands in 2001, questionnaires were mailed to physicians that addressed the decision making that preceded their patient's death. Cases of ending life without an explicit request of the patient were compared with similar cases from 1995 and with cases from Belgium, Denmark, and Switzerland. In the Netherlands in 2001, patients receiving life-ending drugs without their explicit request were most often 80+ years old and had cancer. Most of them were incompetent patients nearing death. Characteristics of this practice in 1995 were quite comparable, as were characteristics of this practice in Belgium, Denmark, and Switzerland. The use of drugs with the intention to hasten death without an explicit request of the patient is part of medical end-of-life practice in the studied countries, regardless of their legal framework, and it occurs in similar fashion.     

Alleviation of pain and symptoms with a life-shortening intention

This article reports the findings of a study into the role of Dutch nurses in the alleviation of pain and symptoms with a life-shortening intention, conducted as part of a study into the role of nurses in medical end-of-life decisions. A questionnaire survey was carried out using a population of 1509 nurses who were employed in hospitals, home care organizations and nursing homes. The response rate was 82.0%; 78.1% (1179) were suitable for analysis. The results show that in about half of the cases (55.8%) nurses were involved in the decision making by the physician and that nurses were frequently (81.5%) involved in administering the medication. The authors' conclusion is that alleviation of pain and symptoms with a life-shortening intention represents a ;grey' area, in which physicians and nurses act on the basis of personal ethical norms rather than legal rules, professional guidelines or shared moral values.     

Integration of critical care and palliative care at end of life

End-of-life care in the critical care environment suffers from a lack of clarity and uncertainty. Critical care nurses may often feel torn between wanting to do everything possible to sustain a patient's life and wanting to do what is in the patient's best interests. Reframing the focus of care from cure to comfort can be an uncomfortable shift for nurses, who may not be in control of when the change of goals takes place. Good end-of-life care should be a core competency for all nurses and it is our responsibility to ensure that decisions to forgo life-sustaining treatment in critical care are appropriate and timely. Futility, conflict and resources all factor in such decisions. Nurses must ensure the transition from cure to comfort does not emphasize a dichotomy between palliative care and critical care but instead focuses on the provision of the best possible end-of-life care.     

Nurse-physician communication concerning artificial nutrition or hydration (ANH) in patients with dementia: a qualitative study

Aims and objectives. To explore nurses' experiences with nurse-physician communication during artificial nutrition or hydration (ANH) decision-making in hospitalised patients with dementia. Background. Artificial nutrition or hydration decision-making often occurs in patients with dementia. Effective communication between professionals is extremely challenging in this population, because these patients are unable to communicate their treatment wishes. Design. Qualitative interview design. Methods. Between April 2008 and June 2009, we conducted 21 interviews with nurses from nine different hospitals geographically spread throughout Flanders (Belgium). Interviews were audiotaped and later transcribed. Data processing involved (1) simultaneous and systematic data collection and analysis, (2) constant forwards-backwards wave, (3) continuous dialogue with the data and (4) interactive team processes. Results. The interviews showed that communication with physicians is the central instrument the nurses used in their attempts to realise their perception of 'the best possible care'. From the nurses' perspective, we distinguished three mutually connected factors that affected the effectiveness of nurse-physician communication during artificial nutrition or hydration decision-making: the physicians' attitude towards the nurses, the nurses' attitude towards the physicians and the forms of communication used by the nurses. The complex interaction between these three factors resulted in a range of nurses' perceptions, varying from positive to negative. The direction of their perceptions depended on the extent to which they succeeded or failed to use nurse-physician communication as an instrument to realise the 'best care'. Conclusion. Nurse-physician communication was the most important instrument determining whether nurses succeeded or failed to actively act as a patient's representative and whether nurses achieved the best possible care in co-operation with physicians. Relevance to clinical practice. To reach optimal care and nurse job satisfaction, nurse-physician communication during artificial nutrition or hydration decision-making should be an open dialogue characterised by mutual respect and understanding.