Predicting the recurrence of facial synkinesis after epineurectomy of facial nerve trunk using logistic regression model.

Abstract

This study aimed to investigate the risk factors affecting epineurectomy of the facial nerve trunk for facial synkinesis and use them to establish a prediction model to assess the recurrence of post-operative facial synkinesis. A total of 68 patients with synkinesis after facial paralysis were enrolled in this study. They were randomized to the training and testing sets. All patients underwent standard surgical procedures and their clinical data were collected and analyzed. The condition of facial synkinesis was evaluated at 12 months after the operation. Univariable and multivariable logistic regression analysis was performed to identify independent risk factors and establish a model and related nomogram to predict the recurrence of post-operative facial synkinesis. Area under receiver operating characteristic curve (AUC) and calibration curves were employed to assess the predictive accuracy of the nomogram. Multivariate logistic regression analysis indicated that older age and higher preoperative score of synkinesis may be the potential factors for the recurrence of post-operative facial synkinesis at 12 months. The prediction model showed a good discrimination with mean AUC of 5-fold cross-validation of 0.781. The accuracy of the nomogram predicting the recurrence in the testing set reached 78.57%, respectively. Bias-corrected curve revealed a strong consistency between the actual observation and prediction. The study illustrated that the proposed logistic regression model based on the age and preoperative score of synkinesis of 2 potential factors could be a promising tool to predict the recurrence of facial synkinesis after the surgery.