Purpose: To investigate the repeatability of three anterior segment clinical grading scales: 1) verbal descriptors scale (VDS), 2) photographic matching scale (PS), and 3) continuous matching scale (CS).
Methods: Five optometrists graded 30 slides each of 3-9-o'clock staining, bulbar redness, and palpebral conjunctival roughness twice, separated by at least a day. VDS and PS were five-point scales (0-4) with half grades permitted. The CS was a 5-second, 240-frame video movie generated using morphing software. PS and CS grading was done with references presented on a computer screen.
Results: Averaged across observers, the test-retest intraclass correlation, correlation coefficient of concordance, and Pearson's r ranged from 0.95 to 0.99 (all p < 0.001). Coefficients of repeatability using CS to grade all three ocular conditions ranged between 0.31 and 0.49. The corresponding PS and VDS coefficients of repeatability ranged between 0.37 and 0.49; PS generally had better repeatability than VDS.
Conclusions: Each of the clinical grading scales was reliable. The coefficients of repeatability showed that bulbar redness and palpebral conjunctival roughness were graded with higher precision using CS.