Shengsen Chen1, An Cui1, Kangkang Yu1, Chong Huang1, Mengqi Zhu1, Mingquan Chen2. 1. Department of Infectious Diseases, Huashan Hospital, Fudan University, Shanghai, China. 2. Department of Infectious Diseases, Huashan Hospital, Fudan University, Shanghai, China. Electronic address: bagews@163.com.
Abstract
BACKGROUND: Meningitis is a serious complication occurring after neurosurgical operations, which can result in severe morbidity and high mortality. This retrospective cohort study aimed to determine the risk factors of postoperative meningitis in a large clinical center of neurosurgery in China. METHODS: Patients who underwent a neurosurgical procedure between January 2014 and December 2015 were selected, and 1016 cases were included our final analysis. On the basis of propensity scores, 84 patients with postoperative meningitis were successfully matched to 84 patients without postoperative meningitis. RESULTS: After propensity score matching, age, hospitalization duration, intraoperative use of corticoids, intraoperative use of antibiotics, external ventricular drainage, lumbar drainage, enteral nutrition, duration of surgery, major craniotomy, and transsphenoidal surgery were associated with postoperative meningitis. Furthermore, preoperative use of corticoids, intraoperative use of corticoids, intraoperative use of antibiotics, external ventricular drainage, lumbar drainage, and major craniotomy were independent predictors of postoperative meningitis in propensity score-matched cohort. A prediction model including these 6 independent predictors was established. Finally, receiver operating characteristic curve and discriminant analysis confirmed that this model has strong predictive power for evaluating postoperative meningitis. CONCLUSIONS: The prediction model built in our study can be an excellent tool for predicting meningitis after neurosurgical procedures.
BACKGROUND:Meningitis is a serious complication occurring after neurosurgical operations, which can result in severe morbidity and high mortality. This retrospective cohort study aimed to determine the risk factors of postoperative meningitis in a large clinical center of neurosurgery in China. METHODS:Patients who underwent a neurosurgical procedure between January 2014 and December 2015 were selected, and 1016 cases were included our final analysis. On the basis of propensity scores, 84 patients with postoperative meningitis were successfully matched to 84 patients without postoperative meningitis. RESULTS: After propensity score matching, age, hospitalization duration, intraoperative use of corticoids, intraoperative use of antibiotics, external ventricular drainage, lumbar drainage, enteral nutrition, duration of surgery, major craniotomy, and transsphenoidal surgery were associated with postoperative meningitis. Furthermore, preoperative use of corticoids, intraoperative use of corticoids, intraoperative use of antibiotics, external ventricular drainage, lumbar drainage, and major craniotomy were independent predictors of postoperative meningitis in propensity score-matched cohort. A prediction model including these 6 independent predictors was established. Finally, receiver operating characteristic curve and discriminant analysis confirmed that this model has strong predictive power for evaluating postoperative meningitis. CONCLUSIONS: The prediction model built in our study can be an excellent tool for predicting meningitis after neurosurgical procedures.