Development and validation of a predictive model for blindness after lens extraction surgery in acute primary angle...

To develop and validate a Cox regression-based nomogram for predicting blindness in patients with acute primary angle closure(APAC) treated with lens extraction(LE) surgery.

This multicenter, retrospective longitudinal study included 254 eyes (217 patients) in the training cohort and 111 eyes (103 patients) from three centers in the validation cohort. Univariate and multivariate Cox regression analyses were performed to identify predictors of blindness. Nomogram performance was assessed using time-dependent receiver operating characteristic(ROC) curves, area under the curve(AUC), concordance index(C-index), and calibration curves. Kaplan-Meier curves were used to compare the blindness incidence between risk groups.

Cox regression identified significant predictors of blindness including logMAR visual acuity(VA) at the acute attack(HR, 1.935; 95% CI, 0.966-3.879), use of intravenous mannitol during the acute phase(HR, 0.453; 95% CI, 0.231-0.887), timing of surgery(linear term: HR, 0.771; 95% CI, 0.315-1.887; quadratic term: HR, 0.304; 95% CI, 0.144-0.646; cubic term: HR, 0.792; 95% CI, 0.419-1.498), preoperative logMAR VA(HR, 1.886; 95% CI, 1.174-3.028), and preoperative intraocular pressure(HR, 1.047; 95% CI, 1.022-1.073). The C-index was 0.860 and 0.826 in the training and validation cohorts, respectively. The 1-year and 5-year AUC values were 0.906 and 0.875 in the training cohort, and 0.947 and 0.894 in the validation cohort, respectively. The calibration curves closely follow the ideal diagonal. Kaplan-Meier analysis revealed significant differences in blindness probability between low- and high-risk groups(P < 0.0001).

The nomogram provides an effective tool for predicting blindness in APAC patients following LE surgery, enabling risk stratification and aiding clinical decision-making.