Predictors of 30-Day Mortality for Surgically Treated Patients with Spontaneous Supratentorial Intracerebral Hemorrhage and Validation of the Surgical Swedish Intracerebral Hemorrhage Score: A Retrospective Single-Center Analysis of 136 Cases

OBJECTIVE: We aimed to identify independent risk factors of 30-day mortality in patients with surgically treated spontaneous supratentorial intracerebral hemorrhage (ICH), validate the Surgical Swedish ICH (SwICH) score within Polish healthcare system, and compare the SwICH score to the ICH score. METHODS: We carried out a single-center retrospective analysis of the medical data juxtaposed with computed tomography 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 crossvalidation and 1000× bootstrap procedure was used to create new scoring system. Finally predictive potential of these scales were compared. RESULTS: The most important predictors of 30-day mortality were: ICH volume (P < 0.01), Glasgow Coma Scale 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 [area under the curve {AUC}: 0.789 (95% confidence interval {CI}: 0.715-0.863) vs. AUC: 0.757 (95% CI: 0.677-0.837)]. Moreover, based on the analyzed characteristics, we developed our score (encompassing: age, ICH volume, anticoagulants status, Glasgow Coma Scale at admission), [AUC of 0.872 (95% CI: 0.815-0.929)]. This score was significantly better than previous ones. CONCLUSIONS: Differences in health care 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.