Risk Factors And Establishment Of Predictive Model For The Hemorrhagic Progression Of Contusion

Objective:To explore the risk factors of the Hemorrhagic Progression of Contusion(HPC),and to establish a clinical prediction model.Methods:Retrospective analysis the Second Affiliated Hospital of University of South China date of 225 case of cerebral contusion patients who were admitted to our hospital from May 2015 to April2019.According to whether HPC occurred,the patients were divided into the progress group and non-progress group.Univariate analysis was used to screen the influencing factors with statistical differences,and then multivariate Logistic stepwise regression analysis was used to screen the risk factors of HPC.Finally,the prediction model of HPC was established based on regression coefficient.And the clinical data of 102 patients with cerebral contusion diagnosed in the the Second Affiliated Hospital of University of South China from May 2019 to December 2019 were used to verify the effectiveness of the model,in which the predictive model differentiation was evaluated by the area under Receiver Operator Characteristic(AUC),and the calibration was evaluated by drawing Calibration Plot and Hosmer-Lemeshow test.Results:(1)The univariate analysis showed that there were significant differences in the age,Long-term smoking history,Systolic blood pressure and mean arterial pressure at admission,Ionized calcium,combined with traumatic subarachnoid hemorrhage(tSAH),traumatic subdural hematoma(tSDH)and Initial hematoma volume between the two groups(P<0.05).(2)Logistic regression showed that,Long-term smoking history,Ionized calcium<1.12mmol/L,combined with traumatic subarachnoid hemorrhage and Initial hematoma volume>4.0ml were the independent risk factors for HPC(P<0.05).(3)The long-term smoking history,serum ionized calcium,combined tSAH and initial hematoma volume were used as predictive factors to establish a predictive model.The area under the ROC curve of the modeling data group was 0.823(95%CI0.768-0.878).The area under the ROC curve of the verified data set is 0.838(95%CI0.748-0.927),which shows that the model has a good degree of discrimination.The evaluation of the calibration degree of the prediction model through the Hosmer-Lemeshow test showed modeling group?~2=2.195,P=0.988>0.05,Verification group?~2=8.450,P=0.490>0.05,manifest that there was no significant difference between the predicted value of the model and the actual predicted value,and the predictive model had better calibration ability.Conclusion:(1)Long-term smoking history,serum ionized calcium<1.12mmol/L,combined with traumatic subarachnoid hemorrhage,initial hematoma volume>4.0ml were high risk factors for the Hemorrhagic Progression of Contusion.(2)Long-term smoking history,serum ionized calcium,traumatic subarachnoid hemorrhage and initial hematoma volume were used as predictive factors to establish a predictive model,which is valuable for the prediction of HPC.