| Background:With the development trend of China’s aging population,the number of stroke patients increases at a rate of about 8.7% every year,among which more than 70% patients lose their ability to work to varying degrees,which poses a major threat to the health and life safety of the whole country and brings heavy economic burden to patients’ families and society [1-3].Among them,Acute ischemic stroke(AIS)accounts for 69.6% of stroke,characterized by a high incidence and easy recurrence,and poses a serious threat to people’s life and health [4].At present,alteplase is considered to be the most effective treatment for AIS in China,which can dissolve thrombus and thus make blood vessels recanalize to achieve the purpose of treatment.However,there is a risk of bleeding in the early stages of thrombolysis,among which intracranial hemorrhage is one of the most serious complications.hemorrhagic transformation(HT),once intracranial hemorrhage transformation(HT)occurs,will aggravate the degree of brain tissue injury and affect the thrombolytic effect,and their clinical prognosis is poor,which is the main obstacle to the promotion of intravenous thrombolytic therapy at present.In order to prevent the occurrence of HT after intravenous thrombolysis,many scholars have studied the risk factors that can affect the occurrence of HT,but the results are not exactly the same.Therefore,it is of great significance to identify independent risk factors related to HT and accurately establish a model to predict the occurrence of HT,which can provide guidance for clinicians to treat AIS,so as to better perform perioperative intervention of intravenous thrombolysis and achieve the goal of improving prognosis and reducing mortality.Objective:To identify independent risk factors for bleeding conversion after intravenous thrombolysis in acute ischemic stroke,and to develop and validate a simple and effective tool for predicting the probability of bleeding conversion.Method:(1)The medical records of patients who received intravenous thromolytic therapy with alteplase for the first time in AIS from January 2019 to January 2023 in the Department of Neurology,Henan University Huaihe Hospital and from August 2020 to January 2022 in the Department of Neurology,Kaifeng People’s Hospital were retrospectively collected.Patients presented at two different hospitals and enrolled in the study were classified as the cohort study group and the external validation group.(2)Whether AIS patients developed HT after intravenous thrombolysis was collected in the study group,and the patients were divided into HT group and non-HT group according to the occurrence of HT.SPSS26.0 software was used to analyze the association between different variables and the occurrence of HT.Significant variables with P < 0.1 in univariate analysis were selected for multivariate regression analysis to determine the independent risk factors affecting the occurrence of HT.(3)After the independent risk factors were determined,the "rms" package in R software(version4.2.2)was used to establish the histogram prediction model of HT after intravenous thrombolysis according to the regression coefficients of multivariate regression analysis of independent variables.(4)The ROC curve was plotted using R software in the cohort study group.The area under the ROC curve(AUC)was calculated to evaluate the differentiation of the prediction model internally.The fit of the prediction model was tested by Hosmer-Lemeshow,and the net benefit of the prediction model to clinical decision-making was evaluated internally through decision curve analysis.(5)External validation of the prediction model was performed with external validation group medical records.SPSS26.0 software was used to analyze the differences between different variables in the queue study group and the external validation group.R software was used to combine the medical records of the external validation group with the prediction model to calculate the AUC of the ROC curve to evaluate the differentiation of the prediction model.Hosmer-Lemeshow tested the calibration of the external evaluation prediction model.Results:(1)Patient inclusion and exclusion study results:A total of 356 patients meeting the experimental requirements were included in the cohort study group,including 72 patients(20.2%)in the HT group and 284 patients(79.8%)in the non-HT group.A total of 68 patients were included in the external validation group,including 11 patients(16.2%)in the HT group and 57 patients(83.8%)in the non-HT group.(2)AIS study results of HT and non-HT groups occurring in different parts in the formation study group:In HT group,most infarcts were located in subcortical region and least in cerebellar region.In the non-HT group,most infarcts were located in the cortical area,and the least were located in the subcortical area.(3)Univariate analysis results of correlation indexes between HT group and non-HT group in the platoon study group:Univariate analysis showed that age(P= 0.013),albumin(P<0.0001),blood glucose level before thrombolysis(P=0.024),history of atrial fibrillation(P<0.0001),history of anticoagulant and antiplatelet drugs(P=0.025),BUN/Cr(P=0.014),NIHSS in the HT group and the non-HT group The difference between(P=0.009)and infarct location(P<0.0001)was statistically significant,suggesting that it might be an influencing factor for HT,but the interference of confounding factors was not controlled.Therefore,multi-factor regression analysis was needed to find independent influencing factors for HT.(4)The results of multivariate Logistic regression analysis on independent influencing factors of HT in the platoon study group were as follows:Analysis showed age(P=0.003,OR=1.042,95%CI=1.014-1.071),NIHSS score(P=0.011,OR=1.086,95%CI =1.019-1.158),BUN/Cr(P=0.016,OR=1.062,95%CI=1.011-1.116),history of atrial fibrillation(P=0.021,OR=2.922 95%CI 1.174-7.269),albumin(P=0.040,OR=0.877,95%CI =0.774-0.994),infarcts located subcortical versus cortical(P= 0.006,OR=3.435,95%CI =1.434-8.230)was an independent factor for intracranial HT(P < 0.05).Among them,albumin was an independent protective factor,while other indexes were independent risk factors.(5)Analysis of independent risk factors for HT in the platoon study group:The optimal critical value,area under the curve 0.599,sensitivity 81.9% and specificity 41.2%for predicting HT by NIHSS score were 6.5.The optimal critical value of BUN/Cr ratio for predicting HT was 16.355,the area under the curve was 0.613,the sensitivity was 83.3% and the specificity was 40.8%.The optimal threshold value,area under the curve,sensitivity and specificity of HT were 56.5,0.595,90.3% and 30.3% respectively.The optimal threshold value,area under the curve,sensitivity and specificity of albumin for HT prediction were 36.38,0.675,59.7%,and 80.3%.(6)According to the independent influencing factors in the study group,a line graph prediction model was constructed and the results of the prediction model were evaluated:According to the independent risk factors of HT after AIS thrombolysis screened by multivariate Logistic regression analysis,R software was used to build a line graph prediction model.The ROC curve was used to evaluate the area under the curve of the model as 0.736,indicating that the model had good differentiation,and the calibration curve showed that the model had good consistency.Clinical Decision Curve Analysis(DCA)suggested that this model had high net benefit.(7)External validation of prediction models and evaluation of models:Univariate analysis showed no differences between the cohort and external validation groups in variables such as age,body mass index,NIHSS score,history of atrial fibrillation,BUN/Cr,prethrombolysis blood glucose level,albumin,and history of anticoagulant and antiplatelet drug use.The area under the curve evaluated externally by the ROC curve was 0.957,indicating good differentiation performance of the model,and the external calibration curve showed good prediction efficiency of the model.External verification reflected the model’s effect in the new data set,and further confirmed that the histogram prediction model had good performance in predicting HT after AIS thrombolysis.Conclusion:(1)Age,NIHSS score,atrial fibrillation history,urea nitrogen/creatinine,subcortical infarction were independent risk factors for bleeding transformation in patients with acute ischemic stroke.Serum high albumin was an independent protective factor for bleeding transformation in patients with intravenous thrombolysis,and had a certain predictive value.(2)Statistical analysis showed that the Nomogram model had good performance. |
PDF Full Text Request