Risk Factors and a Prediction Model for Hemorrhagic Transformation in Acute Ischemic Stroke With Atrial Fibrillation.

Abstract

Objectives: To identify the risk factors of hemorrhagic transformation (HT) and to establish a prediction model for HT in patients with acute ischemic stroke (AIS) and atrial fibrillation (AF).

Methods: From January 2015 to December 2018, patients with AIS and AF were enrolled. Demographics, lesion features, and blood test results were collected. Univariate and multivariate logistic regression analyses were used to identify the independent risk factors of HT. The receiver operating curve (ROC) curve was utilized to determine the cutoff values and the efficiency of the variables. A predictive model was subsequently developed based on the identified independent risk factors.

Results: A total of 259 patients were included. Age [odds ratio (OR): 1.094; 95% CI: 1.048-1.142; P <0.001], LDL-C (OR: 0.633; 95% CI: 0.407-0.983; P =0.042), uric acid (OR: 0.996; 95% CI: 0.991-0.999; P =0.031), Alberta Stroke Program Early CT Score (ASPECTS) (OR: 0.700; 95% CI: 0.563-0.870; P <0.001), cerebral cortex infarction (OR: 0.294; 95% CI: 0.168-0.515; P <0.001), and massive cerebral infarction (OR: 3.683; 95% CI: 3.025-5.378; P <0.001) were independently associated with HT. We have developed a model incorporating these variables. The area under the curve of the predictive model was 0.87 (95% CI: 0.83-0.92), demonstrating satisfactory predictive ability with a sensitivity of 83.5% and a specificity of 76.4%.

Conclusions: Our predictive model, which integrates age, LDL-C, uric acid, ASPECTS, cerebral cortex infarction, and massive cerebral infarction, can be used to predict HT after AIS in patients with AF, thereby facilitating the mitigation of adverse outcomes.