Risk Prediction Model Development For Intravenous Thrombolysis Hemorrhage In Elderly Patients With Ischemic Stroke

Objective1.Risk assessment of hemorrhage after intravenous thrombolysis in elderly patients with ischemic stroke was determined through expert consultation and retrospective case analysis.2.To access the risk factors and hemorrhage outcomes of intravenous thrombolysis in ischemic stroke.The risk predictive model of intravenous thrombolysis hemorrhage in elderly patients with ischemic stroke was established to provide the basis for predicting the outcome of hemorrhage.MethodsThe first part was the study of risk factors of intravenous thrombolysis in elderly patients with ischemic stroke:Relevant literatures on risk assessment of intervenous thrombolysis in ischemic stroke published in domestic and foreign databases were retrieved to construct a draft of risk assessment for intravenous thrombolysis.Then,the research group discussed the draft risk assessment of intravenous thrombolysis and finally conducted expert consultation to determine the risk factors of intravenous thrombolysis.The second part was the construction and evaluation of the risk predictive model for intravenous thrombolytic hemorrhage in elderly patients with ischemic stroke:Based on the risk factors determined by expert consultation,the information of600 patients admitted to a three-class hospital and its three branches in Henan province from January 2018 to January 2020 was recorded by means of convenient sampling.According to the evaluation order and the principle of 70%:30%,they were divided into model group(431 cases)and verification group(169 cases).Univariate and multivariate Logistic regression analysis was performed on the model group samples to determine the independent risk factors for bleeding and construct a risk prediction model.Hosmer-lemeshow chi-square was used to test the fitting degree of the model,and Receiver Operating Characteristic Curve(ROC)was used to test the distinguishing ability of the model.ResultsThe first part:After literature review,discussion by the research group and consultation by 10 experts,32 risk factors at 4 levels including general conditions,basic signs,laboratory tests and past medical history were finally included.The second part:600 patients were included,including 259 patients with hemorrhage,the bleeding rate is 43.2%.Univariate and multivariate analysis showed that sex,low-density lipoprotein and D-dimer were independent predictors of intravenous thrombolysis in elderly patients with ischemic stroke(P<0.05).A risk prediction model of intravenous thrombolysis in elderly patients with ischemic stroke was established,logit P=-0.450+0.523×sex+0.260×LDL-0.053×D-dimer.The area under the ROC curve was 0.747 and 0.714,respectively.When the Yoden index was 0.356,the sensitivity and specificity were 88.3%and 49.7% respectively.H-l goodness of fit test(c~2=9.917,P=0.271).According to the critical point,the sensitivity and specificity of the model were 77.9%and 66.3%respectively.ConclusionThe study screened the risk factors of hemorrhage after intravenous thrombolysis in elderly patients with ischemic stroke through expert consultation and other methods.Independent factors influencing the occurrence of intravenous thrombolysis hemorrhage were determined by regression analysis,including sex,low density lipoprotein and D-dimer.Quantitative assessment of these risk factors can,to a certain extent,predict the incidence of hemorrhage and help health care workers to ensure the safety of patients.