Yi-Guang Mao,1,* Jia-Yu Chen,2,* Man-Li Wang,3 Ying-Jun Ma,1 Chen Jiang1 

1Department of Neurosurgery Intensive Care Unit, the Affiliated Wuxi People’s Hospital of Nanjing Medical University,Wuxi Medical Center,Nanjing Medical University, Wuxi People’s Hospital, Wuxi, Jiangsu Province, 214023, People’s Republic of China; 2Department of Neurology, the Affiliated Huishan Hospital of Xinglin College, Nantong University, Wuxi Huishan District People’s Hospital, Wuxi, Jiangsu Province, 214187, People’s Republic of China; 3Department of Intensive Care Unit, the Affiliated Southeast University of Medicine, Jiangyin People’s Hospital, Wuxi, Jiangsu Province, 214499, People’s Republic of China

*These authors contributed equally to this work

Correspondence: Chen Jiang, Department of Neurosurgery Intensive Care Unit, The Affiliated Wuxi People’s Hospital of Nanjing Medical University, Wuxi Medical Center, Nanjing Medical University, Wuxi People’s Hospital, No. 299 of Qingyang Road, Liangxi District, Wuxi, Jiangsu Province, 214023, People’s Republic of China, Tel +86-051082700778, Email jiangchenjc01@163.com Ying-Jun Ma, Department of Neurosurgery Intensive Care Unit, The Affiliated Wuxi People’s Hospital of Nanjing Medical University, Wuxi Medical Center, Nanjing Medical University, Wuxi People’s Hospital, No. 299 of Qingyang Road, Liangxi District, Wuxi, Jiangsu Province, 214023, People’s Republic of China, Tel +86-051082700778, Email yingjunmamyj@126.com

Background: Identifying risk factors associated with hematoma expansion following spontaneous intracerebral hemorrhage (ICH) is essential for improving early intervention strategies. We hope to use this predictive model in the future to comprehensively score the risk factors of hospitalized patients with cerebral hemorrhage and evaluate the possibility of hematoma enlargement. Being able to identify high-risk patients with hematoma enlargement early and take intervention measures to save their lives.
Methods: A retrospective analysis was conducted on clinical data from 226 individuals diagnosed with spontaneous ICH between December 29, 2023, and August 29, 2024. Multiple logistic regression analysis was performed to identify risk factors associated with hematoma expansion. Predictive performance of the model was assessed using ROC curve analysis and receiver operating characteristic curve analysis. Mortality rates were calculated for each group following a 7-day follow-up period.
Results: Hematoma expansion was associated with diabetes mellitus, a low Glasgow Coma Scale (GCS) score at admission, elevated systolic blood pressure at admission, coagulation abnormalities, and specific computed tomography (CT) imaging findings, such as heterogeneous density, black hole sign, swirl sign, lobulation sign, and blend sign. A prognostic model incorporating these factors demonstrated robust predictive performance, achieving an area under the curve of 0.771 (95% CI: 0.628– 0.915, p = 0.002). The model yielded a maximum Youden index of 0.489, with an optimal cutoff score of 38, a sensitivity of 54.5%, and a specificity of 94.4%. Mortality among individuals with coagulation abnormalities was 53.3%.
Conclusion: Coagulation abnormalities, GCS score, systolic blood pressure at admission, CT imaging findings, and diabetes mellitus were identified as predictors of hematoma expansion in spontaneous ICH. Individuals with coagulopathy and elevated systolic blood pressure at admission exhibited the poorest prognoses.

Keywords: cerebral hematoma, hematoma expansion, mortality, predictive model, risk factors