Predictive nomogram for hearing deficits after microvascular decompression treatment.

Abstract

We explored the impact of brainstem auditory evoked potentials monitoring, as well as anatomical characteristics, in relation to their influence on hearing deficits. A total of 851 patients diagnosed with idiopathic hemifacial spasm underwent microvascular decompression treatment were recruited in our study. A nomogram was developed based on the regression analysis. Nomogram performance was evaluated through receiver operating characteristic (ROC), decision curve analyses and calibration curve. The rate of positive wave V change was also higher in the hearing deficit group (71.8% vs no hearing deficit group, p < 0.001). Furthermore, greater retraction depth (0.78 ± 0.25cm vs 0.55 ± 0.12cm, p < 0.001), duration (74.43 ± 15.74min vs 55.71 ± 7.01min, p < 0.001) and retraction distance (4.38 ± 0.38cm vs 4.17 ± 0.24cm, p = 0.001) were evident in the hearing deficit patients. Multivariate logistic regression showed that positive wave V change (OR 5.43), greater retraction depth (OR 55.57) and longer retraction duration (OR 1.14) emerged as significant independent predictors of postoperative hearing deficit. The external validation cohort exhibited a favorable discrimination with an AUC of 0.88. The calibration curves further confirmed the reliability of the predicted outcome in relation to the observed outcome in the external validation cohort (p = 0.89). The decision curves demonstrated that the nomogram outperformed the All or None scheme when the threshold probability ranged from > 2% to < 60% in the external validation cohort. We constructed a nomogram, including wave V, retraction depth, and retraction duration, which can effectively predict the occurrence of hearing deficits and has good clinical applicability.