Reliability testing of the Dutch version of the Radner Reading Charts

Abstract

Purpose: To statistically analyze the test-retest and inter-chart reliability of the newly developed actual logarithmically scaled Dutch Radner Reading Charts in an older population affected by macular disease. These Dutch Radner Reading Charts are developed according to the strict principles of sentence construction of the originally German language Radner Reading Charts.

Methods: Thirty-six subjects aged 50 years or older and affected with a macular disease monocularly read the three charts of the Radner Reading Charts in a randomized order twice with 1 month in between. The subjects were divided into three groups according to their distance logMAR visual acuity (group 1, > or =0.1; group 2, 0.12 to 0.4; and group 3, 0.42 to 0.8). Reading acuity (logRAD, the reading equivalent of logMAR), logRAD score, maximum reading speed, critical print size, and logRAD/logMAR ratio were measured. Variance component analysis was used to determine the sources of variability.

Results: The test-retest and inter-chart reliability was high for all visual acuity groups and variables. For all groups together the chart accounted for maximal 5% of the total variability for all measurements. The individual subject did have the largest influence on the measurements (88 to 98% of the variability).

Conclusions: The Dutch version of the Radner Reading Charts provided high reliable test-retest and inter-chart measurements of reading performance in a heterogeneous group of subjects with subnormal to low vision. This study also showed that the strict principles of sentence construction of the originally German Radner Reading Charts may also be successfully used for other languages.