Andrew J Anderson1, Stuart K Gardiner2. 1. Department of Optometry and Vision Sciences, The University of Melbourne, Melbourne, Australia. Electronic address: aaj@unimelb.edu.au. 2. Discoveries in Sight, Devers Eye Institute, Legacy Health, Portland, Oregon.
Abstract
PURPOSE: Risk factors for visual field progression in glaucoma can affect both eyes, meaning that progression rates (in decibels per year) between eyes likely are correlated. We investigated whether incorporating information about the progression rate in one eye can improve the estimate of the rate in the fellow eye. DESIGN: Prospective cohort analysis. PARTICIPANTS: We analyzed series of 10 visual field results drawn from the Portland Progression Project and the Rotterdam Eye Study. METHODS: We determined visual field progression rates using linear regression of the summary index mean deviation (MD) over time. We determined best-fitting linear models for the rates over the entire series based on the rate across n visual fields (n = 3 to 6) and used an analysis of variance to determine if incorporating the fellow eye's level of visual field damage (MD) or rate significantly improved these models. MAIN OUTCOME MEASURE: Visual field progression rate (in decibels per year). RESULTS: Visual field progression rates were correlated positively between eyes for the 262 individuals analyzed (r = 0.51; P < 0.0001). For short series (n = 3 to 4), predicting the rate over the entire 10 fields was improved significantly by incorporating rate information from the fellow eye (P < 0.05), but not by incorporating knowledge about the level of damage in the fellow eye (P > 0.61). The fellow eye no longer aided predictions for n = 5 or 6 fields (P = 0.11 and P = 0.42, respectively). Although the coefficient quantifying the influence of the fellow eye's rate changed relatively little for n = 3 to 5 fields, the coefficient for the rate of the eye under consideration increased markedly with the n value. CONCLUSIONS: The long-term rate of visual field change in an eye is, in part, predicted by the rate in the fellow eye, particularly when only a few visual field results are available for each eye. Because of this, apparent rapid progression in an eye is more likely to be real if the fellow eye also seems to be progressing rapidly.
PURPOSE: Risk factors for visual field progression in glaucoma can affect both eyes, meaning that progression rates (in decibels per year) between eyes likely are correlated. We investigated whether incorporating information about the progression rate in one eye can improve the estimate of the rate in the fellow eye. DESIGN: Prospective cohort analysis. PARTICIPANTS: We analyzed series of 10 visual field results drawn from the Portland Progression Project and the Rotterdam Eye Study. METHODS: We determined visual field progression rates using linear regression of the summary index mean deviation (MD) over time. We determined best-fitting linear models for the rates over the entire series based on the rate across n visual fields (n = 3 to 6) and used an analysis of variance to determine if incorporating the fellow eye's level of visual field damage (MD) or rate significantly improved these models. MAIN OUTCOME MEASURE: Visual field progression rate (in decibels per year). RESULTS: Visual field progression rates were correlated positively between eyes for the 262 individuals analyzed (r = 0.51; P < 0.0001). For short series (n = 3 to 4), predicting the rate over the entire 10 fields was improved significantly by incorporating rate information from the fellow eye (P < 0.05), but not by incorporating knowledge about the level of damage in the fellow eye (P > 0.61). The fellow eye no longer aided predictions for n = 5 or 6 fields (P = 0.11 and P = 0.42, respectively). Although the coefficient quantifying the influence of the fellow eye's rate changed relatively little for n = 3 to 5 fields, the coefficient for the rate of the eye under consideration increased markedly with the n value. CONCLUSIONS: The long-term rate of visual field change in an eye is, in part, predicted by the rate in the fellow eye, particularly when only a few visual field results are available for each eye. Because of this, apparent rapid progression in an eye is more likely to be real if the fellow eye also seems to be progressing rapidly.