Abstract
PURPOSE: To investigate the agreement of optic nerve head evaluations and initial diagnoses of primary open-angle glaucoma (POAG) by general ophthalmologists and glaucoma specialists in Shanghai, China. METHODS: This multicenter, cross-sectional study involved the outpatients from the general ophthalmology departments of four top eye hospitals in Shanghai. The participants underwent ocular examinations, including intraocular pressure, fundus photography, corneal thickness, refractometry, visual acuity, visual field and gonioscopy. General ophthalmologists and glaucoma specialists performed the diagnoses and classified them as non-glaucoma, POAG suspects, and POAG. The consistency of initial diagnosis between general ophthalmologists and glaucoma specialists was measured using the weighted kappa coefficient. Logistic regression analysis was used to detect the risk factors for the reliability of POAG diagnosis. RESULTS: In 922 participants, the initial diagnosis rates of POAG and non-glaucoma were much higher in the glaucoma specialist group than in the general ophthalmologist group, while the initial diagnosis rates of POAG suspects were higher in the general ophthalmologist group. The weighted kappa coefficient between the two groups was 0.831±0.027 (95% confidence interval, 0.779-0.884). Logistic regression analysis of the risk factors for the reliability of POAG diagnosis showed that the independent risk factors were intraocular pressure (OR 8.363, 95% CI: 4.27-16.37) and vertical cup-to-disc ratio (OR 3.459, 95% CI: 1.54-7.76). CONCLUSION: The diagnosis consistency between the general ophthalmologists and the glaucoma specialists was similar among outpatients in the area of Shanghai. However, general ophthalmologists tended to classify the indefinite subjects as POAG suspects, and their accuracy in diagnosing POAG was low. By paying more attention to the risk factors of POAG diagnosis, general ophthalmologists could improve the diagnosis accuracy.