Construction and application of a prediction model for postoperative epilepsy with a column chart after craniocerebral surgery
Objective To construct a predictive model for postoperative epilepsy using a column chart and analyze its application value.Methods A retrospective analysis of clinical data was conducted.86 patients with postoperative epilepsy admitted to the Second Affiliated Hospital of AFMU from December 2021 to December 2023 were selected as the observation subjects,who were included in the epilepsy group(n=86).During the same period,115 patients without epilepsy who underwent craniocerebral surgery in our hospital were selected as the no epilepsy group(n=115).General information,clinical data,and treatment methods of the two groups of patients were collected and compared.Multivariate Logistic regression analysis was used to analyze the influencing factors of postoperative epilepsy.Establish a risk prediction column chart model for postoperative epilepsy using R4.1.2 software package and MS program package,and evaluate the value of the column chart model in predicting postoperative epilepsy using receiver operating characteristic(ROC)curve.Results The preoperative GCS score of epilepsy patients was 13-15 points,and the proportion of light reflex(normal)was significantly lower than that of the no epilepsy group.The proportion of mechanical ventilation,brain herniation,brain contusion,and intracranial hematoma was significantly higher than that of the no epilepsy group(P＜0.05).The proportion of self repair materials in epilepsy patients was significantly higher than that in the no epilepsy group(P＜0.05).Multivariate Logistic regression analysis showed that a GCS score of 13-15 was a protective factor for postoperative epilepsy(P＜0.05),while mechanical ventilation,cerebral hernia,and intracranial hematoma were risk factors for postoperative epilepsy(P＜0.05).According to the results of multiple Logistic regression analysis,a column chart model for postoperative epilepsy was drawn using R software.ROC curve analysis showed that the area under the curve(AUC)of the column chart model for predicting postoperative epilepsy was 0.860(95％CI:0.778-0.941),with a sensitivity of 91.53％and a specificity of 70.00％.Conclusion Mechanical ventilation,cerebral herniation,and intracranial hematoma are risk factors for postoperative epilepsy.A GCS score of 13-15 can reduce the risk of epilepsy.Establishing a column chart model for postoperative epilepsy based on these factors has certain value in predicting postoperative epilepsy.Early evaluation can guide treatment plans.
NETL
NSTL
万方数据
