RESUMO
SIGNIFICANCE: The Lanthony D15 has been reported to have poorer repeatability than the Farnsworth D15. This study found that two trials of the test provide high short-term repeatability and can be administered this way for occupational testing. PURPOSE: This study aimed to determine the short-term repeatability of the Lanthony D15 in patients with color vision deficiency. Repeated trials were used to examine if learning effects occur and to determine how many trials would be necessary to ensure the highest short-term repeatability for occupational testing. METHODS: Twenty male subjects (mean [standard deviation] age, 27.2 [4.3] years) with congenital color vision deficiency, ranging from mild to severe, participated in this single-visit study. Visual acuity, color vision book screening, Farnsworth D15, and anomaloscope testing were performed for classification purposes. Ten trials of the Lanthony D15 were performed. Color confusion index scores from each trial were determined, and a repeated-measures analysis of variance was used to compare the scores across trials. Orthogonal polynomial analysis was performed to detect any trends across trials through the third order. The intraclass correlation coefficient was calculated. RESULTS: No differences in color confusion index (mean [standard error of the mean], 3.57 [0.04]) were found across the 10 trials ( P = .18). Legendre polynomials showed no statistical significance (all P > .39). The intraclass correlation coefficient was 0.81 (95% confidence interval, 0.70 to 0.90). Based on the method of Shrout and Fleiss, intraclass correlation coefficients of 0.7, 0.8, and 0.9 could be achieved with an average of one, two, and four trials of the test, respectively. However, empirically, 0.9 was not achievable. CONCLUSIONS: The Lanthony D15 test has fairly high short-term repeatability. Thus, although more trials would likely improve clinical certainty, the mean result of two trials appears sufficient for occupational testing.