Establishment of a Risk Prediction Model for Non-arteritic anterior ischemic optic neuropathy(NAION) in the contralateral eye in unilateral AION patients

Abstract

Abstract Purpose To explore the risk factors for contralateral eye involvement in non-arteritic anterior ischemic optic neuropathy (NAION) and construct a clinical prediction model based on these factors. Methods This study included 151 NAION patients who were hospitalized at the Eye Center of the Second Hospital of Hebei Medical University from October 2018 to December 2021. Clinical data were collected and patients were followed up until December 2022 to determine whether the contralateral eye was affected. The least absolute shrinkage and selection operator (LASSO) regression analysis and K-fold (10-fold in this study) cross-validation were used to screen for predictive factors, and a multivariate logistic regression analysis was applied to establish the prediction model. The receiver operating characteristic curve (ROC), Hosmer-Lemeshow (H-L) test, and decision curve analysis (DCA) were used to evaluate the predictive model and its clinical utility. Results Through LASSO regression analysis, 9 predictive factors were selected from 23 variables: age, visual acuity at admission, acute ischemic stroke (AIS), diabetes mellitus (DM), fasting blood glucose (FBG), high blood pressure (HBP), total cholesterol (TC), optic disc drusen (ODD) in the contralateral eye, and cup/disc ratio (c/d). The model constructed with these 9 predictive factors showed good predictive ability with an ROC area under the curve of 0.85 and a non-significant H-L test result (P = 0.2386). The DCA curve indicated that the model had high clinical utility within a patient's risk threshold range of 5–94%. Conclusion Diabetes mellitus, high blood pressure, total cholesterol, optic disc drusen, and small cup/disc ratio are independent risk factors for contralateral eye involvement in NAION patients. Age, visual acuity at admission, acute ischemic stroke, and fasting blood glucose are closely related to the contralateral eye involvement in NAION. The constructed prediction model based on these factors has good predictive performance.