Development and validation of an improved ICH score: Integrating clinical and radiographic parameters for enhanced prediction of 30-day mortality in spontaneous intracerebral hemorrhage

The ICH score effectively predicts 30-day mortality in patients with intracerebral hemorrhage (ICH), but accuracy decreases in surgical cases. This study evaluated whether integrating clinical and radiographic parameters with the original ICH variables enhances mortality prediction.

This study reviewed charts of 368 adults with spontaneous ICH from 2014 to 2019. Stepwise logistic regression and multivariate analysis were employed to identify additional clinical parameters associated with 30-day mortality. A scoring system was developed using significantly linked factors (p < 0.05) and evaluated for predictive performance using the area under the receiver operating characteristic curve (AUROC).

Among the 368 patients, 104 (28.3 %) died within 30 days. Five variables were assigned score weights: pupil status, warfarin use, infratentorial location, Glasgow Coma Scale (GCS) score, and hematoma volume >30 mL, identified as independent risk factors. The new scoring system improved the AUROC to 0.91 in the internal validation cohort, up from 0.87 for the original ICH score. For surgical patients, AUROC values were 0.69 and 0.57, respectively.

Incorporating pupil status and warfarin use with the three ICH score parameters enhances mortality prediction in all patient cohorts, including those undergoing surgical intervention.

This work was supported by the Faculty of Medicine, Khon Kaen University (Grant number IN63334).