Open Access Open Badges Research article

Prediction of the survival and functional ability of severe stroke patients after ICU therapeutic intervention

Moussa Riachy1*, Frida Sfeir1, Ghassan Sleilaty2, Samer Hage-Chahine3, Georges Dabar1, Taha Bazerbachi1, Zeina Aoun-Bacha1, Georges Khayat1 and Salam Koussa3

Author Affiliations

1 Department of Pulmonary and Critical Care Medicine, Hotel Dieu de France, Beirut, Lebanon

2 Department of Cardiothoracic Surgery, Hotel Dieu de France, Beirut, Lebanon

3 Department of Neurology, Hotel Dieu de France, Beirut, Lebanon

For all author emails, please log on.

BMC Neurology 2008, 8:24  doi:10.1186/1471-2377-8-24

Published: 26 June 2008



This study evaluated the benefits and impact of ICU therapeutic interventions on the survival and functional ability of severe cerebrovascular accident (CVA) patients.


Sixty-two ICU patients suffering from severe ischemic/haemorrhagic stroke were evaluated for CVA severity using APACHE II and the Glasgow coma scale (GCS). Survival was determined using Kaplan-Meier survival tables and survival prediction factors were determined by Cox multivariate analysis. Functional ability was assessed using the stroke impact scale (SIS-16) and Karnofsky score. Risk factors, life support techniques and neurosurgical interventions were recorded. One year post-CVA dependency was investigated using multivariate analysis based on linear regression.


The study cohort constituted 6% of all CVA (37.8% haemorrhagic/62.2% ischemic) admissions. Patient mean(SD) age was 65.8(12.3) years with a 1:1 male: female ratio. During the study period 16 patients had died within the ICU and seven in the year following hospital release.

The mean(SD) APACHE II score at hospital admission was 14.9(6.0) and ICU mean duration of stay was 11.2(15.4) days. Mechanical ventilation was required in 37.1% of cases. Risk ratios were; GCS at admission 0.8(0.14), (p = 0.024), APACHE II 1.11(0.11), (p = 0.05) and duration of mechanical ventilation 1.07(0.07), (p = 0.046). Linear coefficients were: type of CVA – haemorrhagic versus ischemic: -18.95(4.58) (p = 0.007), GCS at hospital admission: -6.83(1.08), (p = 0.001), and duration of hospital stay -0.38(0.14), (p = 0.40).


To ensure a better prognosis CVA patients require ICU therapeutic interventions. However, as we have shown, where tests can determine the worst affected patients with a poor vital and functional outcome should treatment be withheld?