• Edouard Ferrand, Patricia Jabre, Claire Vincent-Genod, Régis Aubry, Michel Badet, Philippe Badia, Alain Cariou, Françoise Ellien, Valérie Gounant, Roger Gil, Samir Jaber, Sylvie Jay, Elena Paillaud, Philippe Poulain, Bernard Regnier, Jean Reignier, Gérard Socie, Bernard Tardy, François Lemaire, Christian Brun-Buisson, Jean Marty, and French Mort-a-l'Hôpital Group.
• Department of Anesthesiology and Intensive Care, Henri Mondor Hospital, Assistance Publique-Hôpitaux de Paris, Paris 12 University, 94010 Créteil CEDEX, France. edouard.ferrand@hmn.aphp.fr
• Arch. Intern. Med. 2008 Apr 28; 168 (8): 867-75.

BackgroundIn developed countries at present, death mostly occurs in hospitals, but the circumstances and factors associated with the quality of organization and care surrounding death are not well described.MethodsWe designed a large multicenter cross-sectional study to analyze the setting and clinical course of each patient on the day of death. We included 2750 clinical departments of 294 hospitals. Of these, 1033 departments (37.6%) of 200 hospitals (68.0%) contributed to the Mort-a-l'Hôpital survey. Data were collected prospectively by the bedside nurse of each patient within 10 days of the occurrence of death. Main outcome measures included circumstances of death in hospitalized patients; secondary outcomes, nurses' perceptions of quality of end-of-life care.ResultsOf the 1033 participating departments, 420 recorded no deaths during the study period and 613 declared at least 1 death. In the 3793 patients who died and were included for assessment, only 925 (24.4%) had loved ones present at the time of death; 70.1% had respiratory distress during the period before death; and only 12.0% were in pain. Written protocols for end-of-life care were available in 12.2% of participating departments. Only 35.1% of nurses judged the quality of dying and death acceptable for themselves. Principal factors significantly associated with this perception were availability of a written protocol for end-of-life care, anticipation of death, informing the family, surrogate designation, adequate control of pain, presence of family or friends at the time of death, and staff meeting with the family after the death.ConclusionsThis large prospective study identifies nonoptimal circumstances of death for hospitalized patients and a number of suggestions for improvement. A combination of factors reflected in the nurses' satisfaction may improve the quality of end-of-life care.

30 keywords...

hide…