Editor—Approximately 10% of all critically ill patients in the USA and in Europe are denied intensive care unit (ICU) admission because of bed shortage.1Strauss MJ LoGerfo JP Yeltatzie JA Temkin N Hudson LD Rationing of intensive care unit services. An everyday occurrence.J Am Med Assoc. 1986; 255: 1143-1146Crossref PubMed Scopus (198) Google Scholar Such a refusal may be associated with a modification of patient care objectives and increased mortality.2Metcalfe MA Sloggett A McPherson K Mortality among appropriately referred patients refused admission to intensive-care units.Lancet. 1997; 350: 7-11Abstract Full Text Full Text PDF PubMed Scopus (143) Google Scholar The post-anaesthesia care unit (PACU) is an alternative setting for the care of critically ill patients.3Ziser A Alkobi M Markovits R Rozenberg B The postanaesthesia care unit as a temporary admission location due to intensive care and ward overflow.Br J Anaesth. 2002; 88: 577-579Abstract Full Text Full Text PDF PubMed Scopus (37) Google Scholar However, caring for the critically ill in the PACU might jeopardize the overall quality of care given to patients after elective surgery. Indeed, PACUs are not designed, equipped, or staffed to serve as ICUs. Furthermore, there is no French recommendation regarding the organization of care for critically ill patients hospitalized in the PACU. To our knowledge, no data exist in France on the frequency of admissions of critically ill patients in the PACU, and on the level of care provided. We therefore conducted a cross-sectional survey to estimate the prevalence and conditions in which critically ill patients are cared for in the PACU. A questionnaire with 112 items was addressed to 175 heads of PACUs from the main hospitals registered in the ‘Fédération Hospitalière de France’ network. Questions were asked regarding hospital activities, the estimated prevalence and type of critically ill patients hospitalized in the PACU, and the perceived satisfaction of the respondents regarding the organization of the PACU. Critically ill patients were described as patients suffering, or at risk of suffering, from one or more life-threatening acute organ failures requiring the initiation of artificial life support.46188-90, Journal Officiel de la République Française. Décret n° 2002-466 du 5 avril 2002 relatif aux conditions techniques de fonctionnement auxquelles doivent satisfaire les établissements de santé pour pratiquer les activités de réanimation, de soins intensifs et de surveillance continue et modifiant le code de la santé publique (troisième partie: Décrets simples).Google Scholar Only such patients hospitalized in the PACU for at least 4 h were taken into account. We obtained 101 responses (response rate of 57.7%). 80.6% of the respondents were the heads of the anaesthesia departments, 25.7% worked in a tertiary care academic hospital, and 14.9% were located in the Paris area. At least one critically ill patient was admitted per week to the PACU in 40.6% of the responding units. The organization of care provided and the main characteristics of the PACUs are summarized in Table 1. The supportive care administered to critically ill patients in the PACU were: central venous line/arterial line access (81.8% of the patients), non-invasive ventilation (66.7%), para-clinical examinations (CT scan, cardiac echography, bronchoscopy, gastrointestinal endoscopy; 51%), continuous haemodynamic monitoring (30%), or arterial embolization (29.3%).Table 1Characteristics of PACUs and care of the critically ill. Data are presented as n (%) or as median (range). PACU, post-anaesthesia care unit. The numbers of respondents may vary slightly because of missing dataTotal (n=101)Number of beds in the PACU: daytime10 (1–52)Paramedical/bed ratio: daytime0.25 (0.1–0.6)Implementation of a protocol describing the way mechanically ventilated patients should be monitored28 (30.8)Implementation of a protocol describing the prevention of pressure ulcers in mechanically ventilated patients18 (19.8)Intermediate care unit in the same hospital89 (89.9)Critically ill patients remaining for >24 h in the PACU12 (14.5)Relatives allowed to visit critically ill patients in the PACU31 (32.6)Waiting room in the PACU dedicated to relatives of the critically ill28 (29.5)Dedicated room to conduct interviews with relatives of the critically ill18 (18.9)Decision to forgo life-sustaining treatments49 (55.7) Open table in a new tab The typical management of a critically ill patient in the PACU was, by order of frequency to remain in the PACU (68.4% of units reported being able to keep their critically ill patients), to be transferred to another hospital (56.5%), or to be exchanged for a lower acuity patient from the ICU (49.3%). Less than half of the respondents were satisfied about the way the critically ill (41%) or their relatives (46%) are cared for, or about the way end of life decisions are handled (41.8%) in the PACU. Lastly, few respondents were satisfied about the way the fee-for-service policy reimburses the care of the critically ill provided in PACU (19%). This descriptive study describes the current practices regarding the care of critically ill patients outlying in French PACUs. The reported number of critically ill patients admitted in the PACU is significant. Existing practices vary greatly from one unit to another. Although data regarding patients’ outcomes are lacking, our study underlines the lack of formal guidelines supporting a frequent practice and the inadequate resources allocated to these patients. Additional studies aimed at assessing morbidity rates in these patients are necessary. Lastly, both improved economic incentives and updates of French national guidelines dealing with the care of critically ill patients in the PACU are needed. None declared.

The surrogate for inpatientsThe French legal framework of the surrogate has been defined by a law passed in 2002 concerning the patients' rights, in response to the absence of prior rights of the incompetent patient. The surrogate is designated only by a competent major patient.In the case of competent patient, the surrogate may support the patient throughout the course of care, including during the hospitalizations or consultations.In the case of incompetent patient, the surrogate must be involved in the decision-making process.A poor designation and a lack of the surrogate's involvement emerge from different French studies since 2002, especially in the end-of-life decisions, despite a specific low passed in 2005, which reinforced the surrogate's role in this context. The evolution of the patients' rights, in France as in most of the industrialized countries, should lead to specific actions to improve the surrogate's involvement, in the respect of the ethical principles of autonomy, beneficence and non-maleficence.

OBJECTIVE:To assess the impact of an educational program on the quality of the end-of-life decision (EOLD).METHODS:Prospective study for 3 months in a surgical Intensive Care Unit (ICU) involving: staff training conferences and guidelines for documenting level-of-care staff conference; audit before and at 3 months; analysis of records for deceased patients. The main outcome measures the proportion of treatment-limitation in dying ICU patients; and the secondary outcomes the decision-making process and nurses' satisfaction.RESULTS:Eighty-three patients were included; among them, 14 with EOLD. Pre-death palliative strategy increased from 51 % to 85 % with a persisting improvement of practices after 2 years. All steps of EOLD decision-making processes were traced in all such cases, 85 % being based on the proposed guidelines. Nursing team's satisfaction rate almost doubled to 70 %.DISCUSSION:The study demonstrate staff members' capacity to quickly improve their procedures for palliative care when provided with appropriate tools to think about the process and come to a decision. Our data suggest the potential benefice to extend this program to the other specialties involved in the end-of-life process.

Background. - Strongly marked ideological positions on the impact of palliative care and limited hard data plague the debate on physician-assisted death.Methods. - A national cross-sectional study on the requests to hasten death (RHD) was conducted among 789 French palliative care organizations. Data were collected for all patients with RHD encountered during year 2010. Data on patients' characteristics, medical, psychological and social context, symptoms, nature of palliative management, patient's evolution, palliative care team's interpretation of the request were obtained.Findings. - A majority of centers responded and 342 teams provided descriptions of 783 RHD, 476 by a patient, 258 by relatives or close friends, and 49 by the nursing staff. Cancer was the most frequent pathology (72 %) and 68 % of the patients had entered terminal stage. Patients rarely appeared with uncontrolled pain (3.7 %), but had difficulties with feeding (65 %), moving (54 %), excretion (49 %), or were cachectic (39 %); 31 % were considered to be anxio-depressive; 79 % did not give physical reasons for their request; 37 % of RHD were maintained and 24 % fluctuated despite provision of regular follow-up by a palliative care team to 83 % of all cases; 68 % of patients died within a month; the interpretation of RHD by the staff was a wish for relief (69 %), patient's inextricable situation (44 %), actual desire not to continue living (36 %), or to be helped to die (30 %).Interpretation. - The large number of described cases provides, for the first time, comprehensive hard data on the evolution of RHDs in a country that has not legalized euthanasia. Whatever the way RHD are expressed, they are frequently maintained despite adequate palliative care with suitable control of pain and psychological support by specialists. (C) 2012 Elsevier Masson SAS. All rights reserved.

BACKGROUND:Strongly marked ideological positions on the impact of palliative care and limited hard data plague the debate on physician-assisted death.METHODS:A national cross-sectional study on the requests to hasten death (RHD) was conducted among 789 French palliative care organisations. Data were collected for all patients with RHD encountered during year 2010. Data on patients' characteristics, medical, psychological and social context, symptoms, nature of palliative management, patient's evolution and palliative care team's interpretation of the request were obtained.FINDINGS:A majority of centres responded and 342 teams provided descriptions of 783 RHD, 476 by a patient, 258 by relatives or close friends and 49 by the nursing staff. Cancer was the most frequent pathology (72%) and 68% of the patients had entered terminal stage. Patients rarely appeared with uncontrolled pain (3.7%), but had difficulties with feeding (65%), moving (54%), excretion (49%), or were cachectic (39%); 31% were considered to be anxio-depressive; 79% did not give physical reasons for their request; 37% of RHD were maintained and 24% fluctuated despite provision of regular follow-up by a palliative care team to 83% of all cases; 68% of patients died within a month; the interpretation of RHD by the staff was a wish for relief (69%), patient's inextricable situation (44%), actual desire not to continue living (36%) or to be helped to die (30%).INTERPRETATION:The large number of described cases provides, for the first time, comprehensive hard data on the evolution of RHDs in a country that has not legalised euthanasia. Whatever the way RHD are expressed, they are frequently maintained despite adequate palliative care with suitable control of pain and psychological support by specialists.

A law relating to patients' rights in end-of-life decisions was promulgated in 2005 in France, and traced a line between "allowing to die" and active euthanasia. However, the question of legalising euthanasia remains at the centre of public debate in France.1LegifranceLoi no 2005-370 du 22 avril 2005 relative aux droits des malades et à la fin de vie. Journal Officiel de la République Française.http://www.legifrance.gouv.fr/affichTexte.do?cidTexte=JORFTEXT000000446240Date: 23 avril 2005Google Scholar It forms part of the socialist programme for the presidential campaign of 2012, which proposes legislation similar to that existing in Belgium and the Netherlands.2Parti SocialisteConvention égalité réelle: texte adopté par le Conseil National.http://www.parti-socialiste.fr/static/9040/convention-egalite-reelle-le-texte-soumis-aux-militants-par-le-conseil-national-64873.pdfGoogle Scholar Since 2002, full traceability of all shared decisions has been legally required, but the reality and magnitude of requests for euthanasia in France has never been assessed. No French study has confirmed whether end-of-life decision-making processes have improved over recent years, by contrast with several other European surveys.3Ferrand E Jabre P Vincent-Genod C et al.Circumstances of death in hospitalized patients and nurses' perception: French multicenter MAHO survey.Arch Intern Med. 2008; 168: 867-875Crossref PubMed Scopus (59) Google Scholar, 4Bosshard G Nilstun T Bilsen J et al.European End-of-Life ConsortiumForgoing treatment at the end of life in 6 European countries.Arch Intern Med. 2005; 165: 401-407Crossref PubMed Scopus (89) Google Scholar We did a survey to review all available medical and nursing records of patients who died in 2007 and 2009 in 12 departments of medicine and surgery in a French teaching hospital. We found seven (0·9%) repeated requests for euthanasia in 761 medical charts reviewed. The reasons for the request were documented in the charts, and all patients had a short-term fatal prognosis (table). However, none was referred to a psychologist or psychiatrist, and in all cases the decision-making process involved only the physician in charge, and not the patient, family, or other surrogate. Only one patient was referred to the palliative care team and comfort was assessed in only three. No active euthanasia was done, but four patients received intravenous sedation from which they did not emerge before death. No evidence of discussion or explicit statement suggested that the procedure was likely to hasten death.TableMain patients' characteristics in request for euthanasia in a French samplePatient 1Patient 2Patient 3Patient 4Patient 5Patient 6Patient 7Year of death2007200720072009200920092009Age (years)82829679656096SexMaleFemaleFemaleFemaleMaleMaleFemaleMain reason for admissionRespiratoryDeterioration of general conditionCardiovascularCardiovascularRespiratoryDeterioration of general conditionRespiratoryMalignant neoplasmYesYesNoNoYesYesNoSelf-sufficientYesNoNoNoYesNoNoMcCabe score 3*McCabe score 3 predicts death within a year.YesYesYesYesYesYesYesLength of stay (days)141262611215Designated surrogateYesYesYesNoYesYesYesTraceability of strategyYesNoYesYesYesYesNoLife-sustaining treatmentYesYesYesYesYesYesNoShared decisionNoNoNoNoNoNoNot applicablePalliative care consultationNoYesNoNoNoNoNoComfort assessmentNoYesNoYesYesNoNoTerminal sedation by opioids and mode of administrationYes, patchYes, IVYes, SCYes, IVYes, IVYes, IVNoTerminal sedation by hypnotics and mode of administrationNoYes, IVNoYes, IVYes, IVYes, IVNoIV=intravenous. SC=subcutaneous.* McCabe score 3 predicts death within a year. Open table in a new tab IV=intravenous. SC=subcutaneous. These data suggest that requests for euthanasia in France do not seem much rarer than in Belgium or the Netherlands.5Bilsen J Cohen J Chambaere K et al.Medical end-of-life practices under the euthanasia law in Belgium.N Engl J Med. 2009; 361: 1119-1121Crossref PubMed Scopus (111) Google Scholar Unmet or undeclared euthanasia are commonplace in these countries, but we have not taken into account those that have not been heard of or documented. Equally, the study underlines a lack of shared decision making associated with these requests. More importantly, it raises suspicion about life-shortening sedation, in a surprising mismatch between law and clinical practice. The results suggest that, in this context, the major risk of legalising euthanasia in France would be to involve the most vulnerable patients—those mainly awaiting compassionate and individualised care. For this reason, the priority of politicians should be to first require a high level of competence and expertise in end-of-life care, before considering any legalisation of euthanasia. We declare that we have no conflicts of interest.

L’hospitalisation d’un malade en réanimation est une agression physique et psychologique. Durant son séjour, le malade est dépendant d’équipements techniques et de compétences humaines pour le support d’au moins une fonction vitale. Ce contexte particulier fait que la plupart des patients ne sont pas en mesure d’être informés ou de communiquer leurs souhaits concernant leur prise en charge [1, 2]. Face à ce déficit contextuel de possibilité d’information, et en dehors des cas où le patient avait exprimé le souhait que l’équipe soignante (médecins et équipe paramédicale) ne communique pas d’information à ses proches, les équipes de réanimation s’adressent spontanément aux proches des patients [3]. Les difficultés du patient à communiquer et sa représentation par les membres de sa famille, ne doivent néanmoins pas mener les réanimateurs à se soustraire aux principes généraux et aux exigences de l’information chaque fois que possible. L’évolution actuelle des recommandations sur l’information au patient impose une mise en question de la nature « paternaliste » encore appliquée en France de la relation médecin–malade. L’autonomie du malade doit être valorisée, recherchée et prise en compte [4]. Si l’application des dispositions légales et réglementaires en matière d’information est impossible en pratique, le patient de réanimation reste un sujet de droit, et la relation médecin–malade ne doit pas être rompue pour autant [1, 2]. L’objectif de l’information répond à un triple impératif : – respecter les principes éthiques de la relation médecin–malade (bienfaisance, non-malfaisance, autonomie et justice) [5] ; – satisfaire les règles déontologiques [6] ; – répondre à l’exigence légale du consentement aux soins [7]. L’une des spécificités du contexte de la réanimation est l’élargissement du caractère théorique de la relation médecin–malade (classiquement définie comme un « sanctuaire », ou un « colloque singulier ») vers une relation « équipe médicale – patient et entourage ». Dans les pays anglo-saxons, l’implication des familles dans les décisions médicales est courante, d’autant plus qu’il s’y mêle un impératif financier puisque c’est souvent la famille qui honore la facture d’hospitalisation. La qualité de l’information apportée aux familles est ainsi intégrée dans les critères d’accréditation des services de réanimation affiliés à la Society of critical care medicine (SCCM) depuis 1994 [8]. En France, les décisions sont assumées le plus souvent par l’équipe médicale, les proches étant le plus souvent informés, quelquefois consultés, mais moins souvent impliqués dans les décisions. Le système de couverture sociale dont nous disposons conforte par ailleurs ce fonctionnement.

The law number 2005-370 of April 22, 2005 concerning the patients' rights at the end of life imposes from now the refusal of futility of care, a shared decision-making in unconscious patients and the duty of a palliative strategy. We describe a case of polytrauma, for which the shared decision-making process led to a palliative strategy after initial aggressive life-support treatments. This case underlines the need for a two-step model of decision-making process, which distinguishes between goals of care and withdrawal or withholding of life support. It suggests that implementation of written procedures could improve the quality of management at the end of life and traceability of decisions.