Do-not-resuscitate (DNR) orders in patients with intracerebral hemorrhage


  • Conflicts of interest: KS: No disclosures. AM: Compensation for consultancy with Boehringer Ingelheim DS: No disclosures. JP: travel expenses and honoraria from Boehringer Ingelheim, Genzyme, and Orion Pharma (all modest). MK: honoraria and travel expenses for participating in the steering committee meetings of the PERFORM, CEPO, MCI-184-E04 and DIAS-4 trials, and for serving as a consultant for Boehringer Ingelheim, Servier, Mitsubishi Pharma Europe Ltd., Siemens AG, Merck, and H. Lundbeck A/S (all modest). TT: scientific advisory boards for Boehringer Ingelheim and Mitsubishi Pharma, consultant to Boehringer Ingelheim, PhotoThera, BrainsGate, Schering Plough, H. Lundbeck A/S, Sanofi Aventis, and Concentric Medical (all modest), and research contracts with Boehringer Ingelheim, PhotoThera, BrainsGate, Schering Plough, H. Lundbeck A/S, Sanofi Aventis, Concentric Medical, Mitsubishi Pharma (all significant).
  • Funding: This study was supported by the Helsinki University Central Hospital Research Funds (EVO). Additional support was received from the Sigrid Jusélius Foundation (AM, TT), and the Academy of Finland (TT).


Background and purpose

Do-not-resuscitate orders may be associated with poor outcome in patients with intracerebral hemorrhage because of less active management.


We sought to characterize the practice of issuing do-not-resuscitate orders in intracerebral hemorrhage. We also aimed to identify possible differences in care according to do-not-resuscitate status.


We conducted a retrospective study of all consecutive intracerebral hemorrhage patients admitted to the Meilahti Hospital of the Helsinki University Central Hospital between January 2005 and March 2010. Data obtained from medical records allowed comparison of characteristics of patients and care of do-not-resuscitate and non-do-not-resuscitate patients as well as patients with early (within 24 h) and late (>24 h) do-not-resuscitate decisions. Logistic regression was used to identify factors independently associated with do-not-resuscitate decisions.


Of our 1013 patients, a do-not-resuscitate order was issued in 368 (35%), of which 262 (73%) occurred within 24 h from admission. Advanced age (odds ratio 1·06 per year; 95% confidence interval 1·04–1·08), more severe stroke (1·09 per National Institutes of Health Stroke Scale point; 1·06–1·13), and deterioration soon after admission (5·12, 3·33–7·87) had the strongest associations with do-not-resuscitate decisions. Patients with do-not-resuscitate orders received recommended care including stroke unit care (43% vs. 64%; P < 0·001) and prophylaxis for deep venous thrombosis (45% vs. 54%; P = 0·027) less often than non-do-not-resuscitate patients. This was especially the case when the do-not-resuscitate order was issued early.


In addition to confirming the role of known intracerebral hemorrhage prognostic factors in do-not-resuscitate decision-making, our results demonstrate that do-not-resuscitate orders led to less active care of intracerebral hemorrhage patients.