Asimina Mataftsi1, Dimitrios Koutsimpogeorgos2, Periklis Brazitikos2, Nikolaos Ziakas2, Anna-Bettina Haidich3. 1. 2nd Department of Ophthalmology, Aristotle University of Thessaloniki, Thessaloniki, Greece. amatafts@auth.gr. 2. 2nd Department of Ophthalmology, Aristotle University of Thessaloniki, Thessaloniki, Greece. 3. Department of Hygiene, Social-Preventive Medicine and Medical Statistics, Aristotle University of Thessaloniki, Thessaloniki, Greece.
Abstract
PURPOSE: To assess if conversion of decimal visual acuity (VA) to logMAR is reliable for clinical and research purposes. METHODS: Right eye VA of 74 consecutive patients was measured with 5 min interval, using (a) the ETDRS chart (VAlog), (b) a decimal chart at 6 m (VA6m), and (c) a decimal chart at 4 m (VA4m). VA was the smallest line whereby four of five optotypes were correctly identified. Decimal scores were converted to logMAR using the formula logMAR = -log(decimal acuity). The agreement between VAlog, VA4m, and VA6m was assessed by the Bland-Altman method. RESULTS: Linear regression analysis of the difference between VAlog and VA6m or VA4m showed a significant slope (p = 0.001), with greater disagreement at higher VA values, i.e., poorer acuity. There was considerable lack of agreement, with discrepancies of up to 0.2 logMAR when VA was measured around 0.5 logMAR. CONCLUSION: Converting decimal VA to logMAR produces overestimation of its true value, especially in lower acuities.
PURPOSE: To assess if conversion of decimal visual acuity (VA) to logMAR is reliable for clinical and research purposes. METHODS: Right eye VA of 74 consecutive patients was measured with 5 min interval, using (a) the ETDRS chart (VAlog), (b) a decimal chart at 6 m (VA6m), and (c) a decimal chart at 4 m (VA4m). VA was the smallest line whereby four of five optotypes were correctly identified. Decimal scores were converted to logMAR using the formula logMAR = -log(decimal acuity). The agreement between VAlog, VA4m, and VA6m was assessed by the Bland-Altman method. RESULTS: Linear regression analysis of the difference between VAlog and VA6m or VA4m showed a significant slope (p = 0.001), with greater disagreement at higher VA values, i.e., poorer acuity. There was considerable lack of agreement, with discrepancies of up to 0.2 logMAR when VA was measured around 0.5 logMAR. CONCLUSION: Converting decimal VA to logMAR produces overestimation of its true value, especially in lower acuities.
Authors: Roy W Beck; Pamela S Moke; Andrew H Turpin; Frederick L Ferris; John Paul SanGiovanni; Chris A Johnson; Eileen E Birch; Danielle L Chandler; Terry A Cox; R Clifford Blair; Raymond T Kraker Journal: Am J Ophthalmol Date: 2003-02 Impact factor: 5.258