In-hospital mortality due to subarachnoid hemorrhage in the "Barcelona Stroke Registry"

OBJECTIVES: To determine clinical predictors of in-hospital mortality in patients with nan-traumatic subarachnoid hemorrhage. PATIENTS AND METHODS: Data de 184 patients with subarachnoid hemorrhage were obtained from consecutive stroke included in the prospective <<Barcelona Stroke Resgistry>>. Demographic, anamnestic, clinical, neuroimaging and outcome variables in the subgroup of patients who died were compared with those in the surviving subgroup, The independent predictive value of each variable an the development of death was asessed with a logistic regression analysis. Three predictive models were constructed. A first model was based on demographic and clinical variables (total 10 variables). A second model was based on demographic, clinical and neuroimaging variables (total 17). A third model was based on demographic, clinical, neuroimaging and outcome variables (total 21). RESULTS: In-hospital death was observed in 44 patients (24%). Transient neurological deficit (OR = 13.92; 95% CI: 1.01-191.95), progressive deficit (OR = 4.21; 95% IC: 1.28-13.86), limb weakness (OR = 3.24; 95% IC: 1.49-7.08) and age (OR = 1.05; 95% CI: 1.02-1.09) appeared to be independent prognostic factors of in-hospital mortality in the first predictive model. In addition to these variables, intraventricular hemorrhage (OR = 5.51; 95% CI: 1.94-16.04) was selected in the second predictive model. Transient neurological deficit (OR = 41.2; 95% CI: 1.61-1056.2), neurological complications (OR = 11.04; CI del 95%: 3.85-31.74), carotid aneurysm (OR = 6.61; 95% CI: 1.23-35.43), intraventricular hemorrhage (OR = 5.51; 95% CI: 1.65-18.4), progressive deficit (OR = 5.35; 95% CI: 1.11-25.90) and hemispheric intracerebral hemorrhage (OR = 4.32; 95% CI: 1.35-13.90), appeared to be independent prognostic factors of in-hospital mortality in the third model. CONCLUSIONS: Clinical features easily obtained at the patient's bedside in additon to neuroimaging data easily obtained in routine neuroimaging studies help clinicians to predict in-hospital mortality in patients with subarachnoid hemorrhage. Transient neurological deficit prior to definitive subarachnoid hemorrhage was the main clinical predictor of in-hospital mortality.