Analysis Of Risk Factors And Establishment Of Forecast Model Of Hemorrhagic Transformation After Cerebral Ischemic Stroke

Objective To establish a predictive model and provide reference basis for clinical doctors in screening patients with a high risk of Hemorrhagic transformation(HT)by analyzing of the related risk factors of HT after ischemic stroke.Methods Making retrospective analysis of the clinical data of the 327 cases with ischemic stroke,collected in the neurology of the affiliated hospital of Qingdao university from which in January 2015 to September 2016.New prediction model was established according to the Logistic regression analysis.The receiver-operating characteristic curve(ROC)was used to obtain the boundary value,higher than it for the HT high-risk group,lower than it for the HT low-risk group.Results The patients were divided into hemorrhagic transformation group(HT group)and non-HT group(NHT group)according to the CT/MRI imaging results.Compared with NHT group,the atrial fibrillation,systolic blood pressure,blood glucose,urine protein,white blood cell count(WBC),prothrombin time(PT-S),international normalized ratio(INR),thrombolysis,large area infarction,the National Institutes of Health Stroke Scale(NIHSS)score more than 10 points of the HT group were increased significantly,the platelet count(PLT),cholesterol,aspirin and clopidogrel treatment were decreased significantly,the differences were statistically significant(P < 0.05).The four groups of systolic blood pressure were compared layer by layer.There was significant difference between the 140-159 mm Hg group and the 160-179 mm Hg group and there was no statistical difference between other layers.The three groups of blood glucose were also compared layer by layer.There was significant difference between the groups of lower than 7.0mmol/L and 7.0-11.0mmol/L.The systolic blood pressure(<160mm Hg,?160mm Hg)and blood glucose(<7.0mmol/L,?7.0mmol/L)were newly grouped and brought into regression analysis.The factors that influenced by human factors significantly,such as aspirin and clopidogrel treatment,and the similar factors,such as PT-S with INR,were excluded.Multivariate logistic regression analysis showed that atrial fibrillation(OR=4.940,95%CI: 1.879~12.988),systolic blood pressure higher than 160 mm Hg(OR=2.762,95%CI: 1.175~6.494),massive cerebral infarction(OR=3.796,95%CI: 1.495~9.635),NIHSS score more than 10 points(OR=5.665,95%CI: 2.289~14.021),thrombolysis treatment(OR=3.086,95%CI: 4.981~96.257)are independent risk factors for HT.The prediction model and the ROC established according to the results of regression analysis.Prediction modle: Logit P=-3.588+1.579×X1+1.016×X2+1.334×X3+1.734×X4+3.086×X5.The AUC is 0.912(95%CI: 0.873~0.950).Take the largest youden index to be the boundary value(the cut-off value).When the boundary value is-2,the sensitivity is 88.33%,specificity is 82.02%.Conclusions The atrial fibrillatio,systolic blood pressure higher than 160 mm Hg,massive cerebral infarction,NIHSS score more than 10 point,thrombolysis treatment are independent risk factors for HT.The patient will have a high risk of HT,when the score is greater than-2.