Study On Short-term Death Level And Death Risk Prediction Of Ischemic Stroke Patients In China

Background and objectivesStroke has the characteristics of high morbidity,high mortality and high disability,which has imposed a heavy economic burden on society.Stroke can be divided into ischemic stroke and hemorrhagic stroke according to the pathological mechanism,in which ischemic stroke is the most common type of stroke,accounting for 69.6%-70.8%in China.To give better treatment to patients,doctors and clinical researchers often need to assess patients' short and long-term risk of death.The effect of ischemic stroke treatment not only depends on the treatment measures of the doctors,but also depends on the patients and their families.Only by providing them with accurate risk prediction can patients better avoid risks and improve their confidence in treatment.At the same time,reliable risk analysis of stroke death can help doctors better arrange supportive care and allocate resources more effectively.Therefore,the risk analysis of death in patients with ischemic stroke is very meaningful.Although there have been many previous studies on the etiology or risk factors of ischemic stroke,no large-scale multi-center stroke cohort study based on the national coverage has been reported in recent years.In order to explore short-term death level and risk analysis of stroke patients in China and establish death prediction model,we matched data from China National Stroke Registry-III(CNSR3)and The China Disease Surveillance Points System,that also provides a new idea for similar large-scale multi-center clinical research.MethodsCNSR3 study included 15166 eligible patients in 201 hospitals from August 2015 to March 2018.In our article,12112 patients enrolled in the CNSR3 study from 2015 to 2017 who has completed the 90-day follow-up were selected as the research ob.jects,and were matched all the cases with The China Disease Surveillance Points System from 2015 to 2017.Through database matching,we obtained study data,including clinical information of stroke patients and death outcome information.We used Cox proportional hazards model to analyze the relevant influencing factors of death of ischemic stroke patients at 90 days after the onset.All statistical analyses in this study were completed with SAS softwareResultsA total of 12112 patients were included in this study,with the mean age was 62.25±11.32 years old and 3743 patients were female,accounting for 30.9%.164 ischemic stroke patients died within 90 days after onset,with a case fatality rate of 1.35%.Multivariate Cox model was used to select the variables with statistical significance in the univariate analysis,including age,gender,education level,time from onset to hospital,emergency vehicle,history of heart disease,mRS score within 24 hours after admission and NIHSS score at admission.The variables were removed with P<0.05 as the standard.Finally,multivariate analysis results were obtained:Age 65-75 years(HR=2.08,95%CI:1.41-3.07,P<0.001,regression coefficient=0.732);Age>75(HR=4.79,95%CI:3.24-7.06,P<0.001,regression coefficient=1.566);History of heart disease(HR=1.92,95%CI:1.37-2.70,P<0.001,regression coefficient=0.653);NIHSS>3 scores at admission(HR=2.48,95%CI:1.76-3.48,P<0.001,regression coefficient=0.906)were independently correlated with the mortality of patients with ischemic stroke at 90 days after onset.According to the regression coefficient(?)corresponding to the risk factors,a scoring system with different risk scores assigned to different levels of risk factors was developed,that is,age 65-75=1 point;age>75=2points;History of heart disease=1 point;NIHSS at admission>3=1 point,total score:4 points.The area under the ROC curve was used to evaluate the prediction ability of the model,and AUC=0.73(95%CI:0.79-0.77).The prediction model established in this study had a good prediction effect.ConclusionDatabase matching research method was used to obtain more complete and accurate information of death outcomes with ischemic stroke patients.Through the data of large-scale multi-center stroke cohort study the death level and related influencing factors of patients with ischemic stroke at 90 days after onset were analyzed,and the death prediction model was established,so as to provide important reference for ischemic stroke prevention and management in China.