| Objective: To explore the related risk factors of hemorrhage transformation after intravenous thrombolysis in acute anterior circulation ischemic stroke,and establish a nomogram prediction model.Methods: From August 2018 to December 2020,patients with acute anterior circulation ischemic stroke who received rt-PA intravenous thrombolysis at the First Affiliated Hospital of China Medical University during the 4.5-hour thrombolytic time window of onset were selected,and their clinical data,biochemical indicators and CT imaging data were collected,we aim to analyze their correlation with post-thrombolysis HT(hemorrhagic transformation).Two independent sample t-test was used for univariate analysis,and factors with significant differences were included in binary logistic regression analysis to obtain independent risk factors for post-thrombolysis HT.Finally,the independent risk factors were made a nomogram to predict the risk of HT.The predictive value was analyzed by calculating C statistics and drawing a calibration curve.Results: A total of 184 patients with acute anterior circulation ischemic stroke who received recombinant tissue plasminogen activator intravenous thrombolysis were included in this study,of which 38 cases(20.65%)developed HT.Conducting a single factor analysis study,it is found that there are statistically significant differences in 5factors including atrial fibrillation history,NHISS(National Institute of Health stroke scale)score before thrombolysis,EIS(early infarction sign),HMCAS(middle cerebral artery high density sign),and moderate to severe leukoaraiosis(P<0.05),related to HT.Incorporating the five factors into the binary logistic regression model,EIS(p=0.020,OR=2.823,95% confidence interval: 1.181-6.746),HMCAS(p=0.018,OR=2.815,95% confidence interval: 1.197-6.617),moderate to severe leukoaraiosis(p=0.016,OR=4.659,95% confidence interval: 1.339-16.207),these three factors are independently related to HT(p<0.05).Establish a nomogram prediction model according to the three final independent risk factors,the calculated C statistic is0.712(between 0.71-0.90),indicating that the model has medium accuracy.By drawing the calibration curve,it can be seen that the predicted HT incidence has the same overall trend as the actual observed HT incidence,but there are some errors.According to whether the outcome is PH2 hemorrhage transformation,the nomogram and calibration curve are made again.The calculated C statistic is 0.738,which is more precise than before,but still has medium accuracy.Conclusion: 1).EIS,HMCAS and moderate to severe leukoaraiosis are independent risk factors for post-thrombolysis HT.2).The nomogram produced by EIS,HMCAS and moderate to severe leukoaraiosis has certain predictive value for post-thrombolysis HT especially for PH2,but it still needs to be improved in more patient cohorts and it needs forward-looking confirmation. |
