Development and validation of the HIGHT prediction model for early hematoma expansion in spontaneous intracerebral hemorrhage

Spontaneous intracerebral hemorrhage (sICH) is associated with high mortality and disability, with early hematoma expansion (EHE) being a key factor in poor prognosis. The most effective strategies for predicting EHE and improving patient outcomes remain unclear. This study systematically reviewed clinical studies on EHE in sICH from the PubMed database (2013-2023) to identify key predictors for a prediction model. Key factors included: Hematoma hypodensity, Island sign, Glasgow Coma Scale, Hematoma in ventricular, and Taking anticoagulant(H-I-G-H-T). The HIGHT model was developed based on these predictors and validated using receiver operating characteristic (ROC) curves, calibration curves, and decision curve analysis. A dynamic nomogram (DN) was created for clinical use. Results: The model was tested in a retrospective cohort of 532 patients, yielding a good fit (AUC = 0.706, P < 0.0001). In a prospective cohort of 83 patients, the AUC was 0.642 (P = 0.048), with sensitivity of 93.75% and specificity of 32.84%.Conclusion: The HIGHT prediction model demonstrated It has shown a certain extent predictive effectiveness in both retrospective and prospective evaluations. And the online DN is now available for clinical use.