TIÊN LƯỢNG CƠN GÓC ĐÓNG CẤP DỰA TRÊN CÁC THÔNG SỐ SIÊU ÂM SINH HIỂN VI

Abstract

Objective: To develop a logistic regression model to predict acute angle closure (AAC) using ultrasound biomicroscopy (UBM) parameters. Methods: A cross-sectional study was conducted on 30 patients (60 eyes) with unilateral AAC at Ho Chi Minh City Eye Hospital. Following resolution of the acute episode, both eyes underwent UBM with measurements of ACD, LP, TIA, LT, ICPD, IC, IT750, and IT2000. Univariate and multivariable logistic regression analyses were performed with checks for multicollinearity and ROC analysis. Results: On univariate analysis, ACD and TIA were significantly associated with AAC risk (ACD: OR = 0.0053; 95%CI: 0.00013 – 0.210; p < 0.01; TIA: OR = 0.63; 95%CI: 0.47 – 0.84; p = 0.002), while LP showed only a borderline trend (p=0.056). In the final multivariable model, only TIA remained an independent predictor. The regression equation was: Logit(p) = 5.86 – 1.39×ACD – 0.34×LP – 0.47×TIA. ROC analysis showed an AUC of 0.867; at the optimal cutoff (p = 0,61), both sensitivity and specificity were 83.3%. Conclusion: Narrow trabecular–iris angle (TIA) was identified as the strongest independent predictor of AAC. The logistic regression model demonstrated good discriminative ability, suggesting potential utility for screening and prophylaxis, although further validation in larger populations is required.