Abstract

Aim: To detect optimum biometric cutoffs from the Chennai glaucoma study (CGS) dataset and assess its ability to detect primary angle closure disease (PACD) in a separate clinical dataset
Methods: Clinical arm-Prospective recruitment, phakic adults. Noted axial length (AXL), anterior chamber depth(ACD), lens thickness (LT). Calculated additional parameters. ROC curves to determine CGS optimal biometric cutoff for PACD detection. Determined ability of parameters from CGS to detect PACD in clinical arm
Results:Clinical-942, CGS 1844 eyes. ACD had highest AUC(0.86-0.89) among all parameters across all datasets. Sensitivity- 63.59%, specificity 93.21% using ACD cut off from CGS rural dataset (ACD≤2.82mm) to detect PACD in the clinical dataset. Accuracy at 10% prevalence 91.54% at urban cut off (ACD ≤2.70mm), at 30% prevalence- 84.32% at rural cutoff
Conclusion: Biometry may be considered with due caution to flag patients requiring further evaluation for PACD in resource constrained areas