Abstract
Abstract: :
Purpose: To evaluate the ability of various glaucoma screening tests to detect glaucomatous optic nerve damage in a population–based sample of Latinos.Methods: The Los Angeles Latino Eye Study (LALES) is a population–based study of eye disease in Latinos aged 40 years and older. Study participants underwent an interview and a clinical examination that included Humphrey visual field testing SITA Standard (HVF) and Frequency Doubling Perimetry (FDP), measurement of intraocular pressure (IOP) and central corneal thickness (CCT), and cup/disc ratio. The gold standard for glaucoma was the presence of a glaucomatous optic nerve appearance by a consensus of three glaucoma specialists. Chi–square analysis and stepwise logistic regression were used to evaluate the ability of the following parameters to detect glaucoma: HVF result, fixation losses (FL>50%), false positives (FP>50%), false negatives (FN>33%), mean deviation (MD, p<5%), pattern standard deviation (PSD, p<5%), abnormal glaucoma hemifield test (GHT), FDP, IOP > 21mmHg, and CCT (<510µ, 510–580, >580). Sensitivity and specificity were calculated for HVF, FN, MD, PSD, GHT, FDP, IOP, CCT, and prior diagnosis of glaucoma. Receiver operating curves were calculated for MD, PSD, and IOP. Results: Sensitivity and specificity were as follows: HVF result = 63.3%, 89.0%, FN = 4.1%, 98.3%, MD = 78.8%, 64.3%, PSD = 66.2%, 77.9%, GHT = 76.0%, 71.6%; FDP = 61.6%, 78.9%, IOP = 18.0%, 97.0%, CCT = 9.6%, 93.8%, and prior diagnosis = 26.5%, 98.0%. The areas under the curve (AUC) were as follows: MD = 0.79, PSD = 0.79, and IOP = 0.72. Conclusions: In this population–based study of Latinos, HVF parameters and FDP parameters had moderate sensitivity and specificity. IOP, CCT and prior diagnosis had poor sensitivity and good specificity. In a population–based screening for glaucomatous optic nerve damage, neither IOP, FDP nor HVF are good screening tests.
Keywords: clinical (human) or epidemiologic studies: treatment/prevention assessment/controlled