Do-not-resuscitate decisions in a large tertiary hospital : differences between specialties and results of a hospital-wide intervention

Background: Despite the advent of palliative care, the quality of dying in the hospital remains poor. Differences in quality of end-of-life practice between hospital wards are well known in clinical practice but rarely have been investigated. Methods: A prospective observation of do-not-resuscitate (DNR) decisions was conducted in patients dying in a Belgian university hospital before (115 patients) and after (113 patients) a hospital-wide intervention consisting in informing hospital staff on the law on Patient's Rights and the introduction of a new DNR form. The new DNR form puts more emphasis on the motivation of the DNR decision and on the description of the participants in the decision-making process. Results: The completion of DNR forms improved after the intervention: physicians better documented who participated in DNR decisions (for participation of family: 63% after the intervention vs. 44% before the intervention, p=0.022, for nurses: 27% vs. 14%, p=0.047) and the motivation for these decisions (59% vs. 32%,p=0.001). However, there was no difference in referral to the intensive care unit (ICU) at the end of life (in 40% of patients after and 37% before the intervention). Furthermore, the number of patients dying without DNR form on the wards was similar (13% and 8%). Surgical patients and patients with non-malignant diseases were more often referred to ICU at the end of life (71% in surgical vs. 35% in medical patients, p<0.001 and 49% in patients with non-malignant diseases vs. 23% in patients with malignancy, p