Hemorrhagic transformation risk in stroke thrombolysis
Year: 2026
Type: article
Study design: Observational Study
Abstract: This study aimed to investigate risk factors for hemorrhagic transformation (HT) following intravenous thrombolysis in acute ischemic stroke (AIS) and to develop a novel nomogram for individualized risk prediction. A total of 346 AIS patients who received reperfusion therapy between September 2023 and September 2025 were included, taking HT as the primary endpoint, and randomly assigned to training ( n = 242) and validation ( n = 104) sets in a 7:3 ratio. Demographic, clinical, laboratory, and imaging data were collected. In the training set, patients were stratified by HT occurrence. Least absolute shrinkage and selection operator regression and multivariate logistic regression analysis were employed to identify predictive variables. Subsequently, three machine learning models—Random Forest, Gradient Boosting Machine, and Support Vector Machine—were constructed. Model performance was assessed by area under the receiver operating characteristic curve (AUC) for discrimination, calibration curves with Brier scores for calibration, and decision curve analysis (DCA) for clinical utility. The Shapley Additive Explanations (SHAP) method was utilized to interpret the optimal model. HT occurred in 59 of 346 patients (17.05%). Six core predictors were identified: age, history of atrial fibrillation, extent of early ischemic changes in middle cerebral artery territory, baseline NIHSS score, blood glucose level, and platelet count. Among the models, Random Forest demonstrated superior performance, achieving an AUC of 0.871 (95% CI: 0.794–0.948) in the training set and 0.788 (95% CI: 0.650–0.926) in the validation set. The Random Forest model’s calibration curve showed strong agreement between predicted and observed outcomes. DCA revealed that the Random Forest model provided a higher net benefit than “treat-all” or “treat-none” strategies within a threshold probability range of 5% to 35%. SHAP analysis of our single-center cohort indicated that baseline NIHSS score and platelet count exerted the strongest predictive effects on HT risk within this patient population. The developed nomogram for predicting HT after intravenous thrombolysis in AIS shows good predictive accuracy and clinical utility in this single-center cohort. It may help clinicians screen patients with elevated HT risk for close monitoring and individualized risk-stratified management within our single-center cohort, while multi-center external validation is mandatory before generalized clinical application. Not applicable.
Source: BMC Neurology
Authors: Qianqian Qu, Lijuan Yan, Mingyue Jiang, Qian Cui
Institution: Affiliated Hospital of Xuzhou Medical College
Language: English
Subfield: Epidemiology
Field: Medicine
Domain: Health Sciences
Open Access status: gold
Estimated APC paid: $3,090
OpenAlex ID: W7225180176