To validate and compare three outcome prediction scores for patients with intracerebral hemorrhage (ICH) in an Argentine cohort of patients with this condition. 

We retrospectively reviewed clinical, laboratory and neuroimaging data of consecutive cases of ICH admitted to our institution between 2012 and 2019. Patients with secondary causes of ICH were excluded. We defined poor outcome as a modified Rankin Scale [mRS] >= 4. Three prognostic scales were evaluated: the ICH Score, Max ICH Score and FUNC Score. Empirical ROC curves and their corresponding area under the curve (AUC) were computed for each score, and differences between each pair were calculated using DeLong’s method. 

A total of 101 ICH patients were admitted in this study (median age 71,08 [standard deviation [SD] 12,88], 45.5% (n=46)  female), of which 37,6% (n=38) were deep, 56,4 % (n=57) lobar and 6% (n=6) infratentorial. Median hospitalization time was 10.5 days [interquartile range [IQR] 5.75, 21.75] , mortality was 6,9% (n=7), and 47.5% (n=48 patients) had poor functional outcome (mRS >=4). The Max-ICH Score (AUC 0.87 [0.80-0.94]) performed better than the ICH Score (0.81 [0.72-0.90]) and the FUNC Score (0.74 [0.64-0.85]) (p=0.03 and p=0.001, respectively). 

The Max-ICH Score showed an excellent performance for predicting functional outcome after ICH in this Argentine cohort. The ICH Score and the FUNC Score performed well.  

The validation of these tools in our population facilitates communication with patients and families regarding possible recovery and goals of care.