Assessment of Risk Factors for Early Seizures following Surgery for Meningiomas Using Logistic Regression Analysis

Abstract

This study analysed the influence of clinical factors on early postoperative seizures in patients with meningiomas and constructed a logistic regression equation for assessing risk factors. Clinical data from 222 patients with meningiomas were collected. The odds ratios (ORs) for independent variables were determined: the ORs for preoperative seizure history and movement disorder were > 1, whereas the OR for prophylactic therapy was < 1. A logistic regression analysis was then performed to select potential risk factors for early postoperative seizures. Five variables (preoperative seizure history, movement disorder, tumour location, primary location of initial tumour and prophylactic therapy) were introduced into the regression model. A logistic regression equation was then constructed that had a positive predictive value of 66.65% and a negative predictive value of 84.95%. This suggested that the five variables introduced in the equation were closely associated with early postoperative seizures, with preoperative seizure history and movement disorder as potential risk factors and prophylactic therapy as a protective factor.