Predictive effect of Nomgram model on the risk of hematoma enlargement in patients with acute cerebral hemorrhage after operation
Huang Haining
Guo Bingquan
Han Yuedong
Abstract：Objective To explore the predictive effect of Nomgram model on the risk of hematoma enlargement in patients with acute cerebral hemorrhage after operation.Methods A total of 282 patients with acute cerebral hemorrhage received surgical treatment,and were divided into hematoma stabilization group(n=204)and hematoma enlargement group(n=78)according to whether hematoma enlargement occurred or not.To analyze the risk factors of postoperative hematoma enlargement in patients with acute cerebral hemorrhage,construct a Nomgram prediction model,and evaluate the clinical value of the Nomgram prediction model.Results There were significant differences in age,diabetes history,hypertension history,systolic blood pressure,diastolic blood pressure,coagulation functional composite score(CFCS),platelet count,plasma prothrombin time,fibrinogen,international normalized ratio(INR),hematoma volume at admission and bleeding sign data between the hematoma stabilization group and the hematoma enlargement group(P<0.05).After the 12 indexes were further screened by Lasso regression,the variables finally included in the model were diabetes history,hypertension history,CFCS,INR,hematoma volume at admission and bleeding sign.Logstic regression analysis showed that diabetes history,hypertension history,CFCS,INR,hematoma volume at admission and bleeding sign were all independent risk factors for postoperative hematoma enlargement in patients with acute cerebral hemorrhage(P<0.05).The value was assigned according to the weights of the main risk factors screened by multivariate analysis,and a nomogram model affecting postoperative hematoma enlargement in patients with acute cerebral hemorrhage was constructed.The larger the hematoma volume at admission,the lower the INR,and the higher the CFCS,the higher the score of the model.The history of diabetes was 3 points,the history of hypertension was 2 points,and the bleeding sign was 10 points.The ROC curve showed that the AUC of Nomgram model in predicting the postoperative hematoma enlargement of patients with acute cerebral hemorrhage was 0.897(95%CI 0.828-0.966).The results of calibration curve analysis show that the predicted probability and actual probability of the nomogram model affecting the postoperative hematoma enlargement of patients with acute cerebral hemorrhage were basically consistent.Conclusion Diabetes history,hypertension history,CFCS,INR,hematoma volume at admission and bleeding sign are all independent risk factors for postoperative hematoma enlargement in patients with acute cerebral hemorrhage.Nomgram model has high clinical value in predicting postoperative hematoma enlargement in patients with acute cerebral hemorrhage.
Keywords：cerebral hemorrhagebrain hematomanomgram modelpredicted value
Publication Date：2023-12-30
Online Publishing Date：2025-08-15(First online date of this platform, not the publication date of the document)
Pages：6( 104-109 )