Measurement and analysis of visual acuity in multicenter randomized clinical trials in the United States: Findings from a survey

Abstract

Purpose To describe visual acuity (VA) testing and scoring methods used in multicenter randomized clinical trials in ophthalmology in the United States (USA) sponsored by the National Eye Institute (NEI). Methods A survey was conducted among multicenter studies in the US that included one or more randomized clinical trials and were sponsored by the NEI, National Institutes of Health. To be included in the survey, a study had to have VA reported in one or more publications or patient eligibility based on VA, a majority of study subjects 13 years of age or older, and a patient population primarily or exclusively from the US. A standard survey form was completed for each study based on information presented in the study manual of procedures and publications. Findings were summarized and displayed with frequency distributions. Results The survey included 24 studies, each with enrollment initiated in the period 1972 to 1999.VA was one of the study eligibility criteria for 19 (79%) studies, and VA or change in VA was the primary outcome in 12 (50%) studies. ETDRS charts have been employed in 16 of 19 studies initiated after the charts were published in 1982. All studies but one specified in the documents reviewed that VA testing was performed at multiple VA test distances. For studies that used ETDRS charts, methods of converting VA scores obtained at different test distances to a common scale fell into two categories: 11 studies used a method similar in nature to that first reported in the Macular Photocoagulation Study, and 7 studies used a method similar in nature to that first reported in the Krypton-Argon Regression of Neovascularization Study. Conclusion The development of the ETDRS charts and a custom light box has led to more uniformity in VA test charts and chart illumination. However, details of VA measurement protocols vary widely from study to study. To assure comparable data across studies and ocular conditions, it would be useful to have a standard VA testing and scoring protocol with provisions for testing different levels of visual acuity that recommends a single method for converting scores from different test distances to a common scale and standard conversion of “off the chart”VA levels for calculation of means and changes in VA.