Predictors of 30-day mortality for surgically treated patients with spontaneous supratentorial intracerebral haemorrhage and validation of the Surgical Swedish ICH Score: a retrospective single-centre analysis of 136 cases.

We aimed to: identify independent risk factors of 30-day mortality in patients with surgically treated spontaneous supratentorial ICH, validate the SwICH Score within Polish Healthcare System, and compare the SwICH Score to the ICH Score. We carried out a single-centre retrospective analysis of the medical data juxtaposed with CT scans of 136 ICH patients treated surgically between 2008 and 2022. Statistical analysis was performed using the same characteristics as in the SwICH Score and the ICH Score. Backward stepwise logistic regression with both 5-fold cross-validation and 1000x bootstrap procedure was used to create new scoring system. Finally predictive potential of these scales were compared. The most important predictors of 30-days mortality were: ICH volume (p<0.01), GCS at admission (p<0.01), anticoagulant status (p=0.03), and age (p<0.01). The SwICH score appears to have a better predictive potential than the ICH score, although this did not reach statistical significance [AUC: 0.789 (95% CI: 0.715-0.863) vs. AUC: 0.757 (95% CI: 0.677-0.837)]. Moreover, based on the analysed characteristics, we developed our score (encompassing: age, ICH volume, anticoagulants status, GCS at admission), [AUC of 0.872 (95% CI: 0.815-0.929)]. This score was significantly better than previous ones. Differences in healthcare systems seem to affect the accuracy of prognostic scales for patients with ICH, including possible differences in indications for surgery and postoperative care. Thus, it is important to validate assessment tools before they can be applied in a new setting and develop population-specific scores. This may improve the effectiveness of risk stratification in patients with ICH.