Measurement error in assessing the size of posterior subcapsular cataracts from retroillumination photographs.

This content is PDF only. Please click on the PDF icon to access.

Abstract

Twenty-six eyes with posterior subcapsular opacities of various sizes were photographed with the Neitz-Kawara Retroillumination camera. The outline of the opacity in a single photograph of each opacity was traced onto a transparent plastic overlay twice by two independent outliners. Two methods were used to estimate the area within the outlines of the opacities. In the first, a transparent overlay with a standard grid was used to count the number of boxes within the outlines. The second method used computer planimetry to estimate the area within the tracings. We estimated the measurement error associated with a single outlining of an opacity and the contribution of the measurement error to overall sample size requirements in studies comparing the mean areas of posterior subcapsular opacities. Variability in the measurement techniques contributed fewer than 20 additional subjects to overall sample size estimates, a small contribution to total sample size requirements in most studies. An outliner's inherent variability in outlining an opacity was a much larger contributor to the measurement error than was variability in assessing the area of the outline of the opacity. While within outliner variability was similar for the two persons outlining the opacities, there were systematic differences in the way the two traced the outlines. Variability from the use of separate photographs of the same opacity taken by different photographers was minimal.