Purchase this article with an account.
Amir Marvasti, Linda M Zangwill, Ricardo Y Abe, Alberto Diniz-Filho, Carolina Gracitelli, Robert N Weinreb, Christopher A Girkin, Jeffrey M Liebmann, Felipe A Medeiros; The Relative Odds of Progressing by Structural and Functional Tests in Glaucoma. Invest. Ophthalmol. Vis. Sci. 2015;56(7 ):623.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To evaluate the relationship between disease severity, number of tests during follow-up and the odds of progressing by structural and functional tests in glaucoma.
This was an observational cohort study involving 451 eyes of 303 patients with glaucomatous optic neuropathy and 110 eyes of 59 healthy subjects. Subjects were followed for an average of 3.6 ± 0.9 years with an average of 7.7 ± 2.4 visits. At each visit during follow-up, subjects underwent visual field assessment by standard automated perimetry (SAP) with Swedish interactive threshold algorithm (SITA Standard 24-2) and retinal nerve fiber layer (RNFL) evaluation by spectral domain optical coherence tomography (Spectralis SDOCT, Heidelberg Engineering). Slopes of change were calculated for SAP mean sensitivity (MS) and SDOCT average RNFL thickness. To ensure matched specificities and to take into account age-related changes, quantile regression of longitudinal data from healthy eyes was used to obtain cut-off values for age-related expected rates of change for each test. Progression by each test for glaucoma eyes was determined if the slope of change was statistically significant and exceed the 95% confidence limit cut-offs. A logistic regression model was then used to evaluate factors associated with the relative odds of progressing by SDOCT versus SAP. The model adjusted for age and variability (standard error of slope) of SAP and SDOCT.
There were statistically significant effects of the number of tests, baseline disease severity and an interaction between these two variables on the odds of progressing by structure versus function. The odds of progressing by SDOCT were significantly higher than by SAP for a lower number of tests and at early stages of disease (Figure 1). For moderate and advanced stages of disease, the odds of progressing by SAP were higher than by SDOCT, but only if a relatively large number of tests were acquired during follow-up.
The relative odds of identifying progression by structural and functional tests varied with the number of tests acquired during follow-up and the stage of the disease.
This PDF is available to Subscribers Only