Intracerebral hemorrhage in intensive care unit : early prognostication fallacies. A single center retrospective study

BACKGROUND: Intracerebral hemorrhage (ICH) admitted to intensive care is deem of poor prognosis. To compare observed and predicted 30-days mortality and to evaluate long term functional outcome in a consecutive ICH cohort. METHODS: Retrospective analysis of prospectively collected data of ICH patients managed in a Neuro-ICU from 2012 to 2015. RESULTS: 136 consecutive patients; 34 (25%) had "withholding of life-sustaining treatment" (WLST) order and 102 (75%) received a "full treatment" (FT). WLST cohort: median (IQR): 72 (70-77) years old, Glasgow Coma Scale (GCS) 4 (3-4) at admission, ICH volume 114 cm3 (68- 152); all patients died during neuro-ICU recovery, 28 (82%) patients had brain death diagnosis and 15 (54%) of these were organ donors. FT cohort: 67 (51-73) years old, GCS 9 (6-12) at admission, ICH volume 46 (24-90) cm3, neurosurgery for clot removal in 65 (64%) (p <0.05 vs. WLST cohort for each of previously listed variables); 13 (13%) patients died during neuro-ICU recovery, of these 11 (85%) patients had brain death diagnosis and 4 (36%) of them were organ donors. Overall 30- days observed mortality for FT group was 18% (95% CI: 11%-26%). Patients with ICH score 1, 2, 3, 4+ had 0%, 10%, 16% and 26% 30-days mortality, respectively (p<0.0001 vs. ICH score). Full treatment group 180-days mortality was 32% (95% CI: 24%-42%). Modified Rankin Scale (mRS) after one year was ≤3 in 35 (35%), i.e. good recovery, and >3 in 64 (65%). Neurosurgery for clot removal was associated with a lower 30 and 180-days mortality (p=0.0005 and p=0.0268, respectively) and along with GCS at admission it was an independent significant prognostic factor. CONCLUSIONS: Mortality and functional outcome is less severe than predicted in patients with ICH receiving a full medical and/or surgical treatment