Repeatability of tear meniscus evaluation using spectral-domain Cirrus® HD-OCT and time-domain Visante® OCT.

Abstract

To investigate the intra-rater, inter-rater and inter-device repeatability of a spectral-domain OCT (Cirrus) and a time-domain OCT (Visante) for tear meniscus height (TMH) and area (TMA) measurements. 20 participants with no known eye disease were recruited. Both eyes of participants were imaged with both OCTs under the similar conditions. The inferior tear meniscus was imaged at 6 o'clock position whereas the superior meniscus was imaged at 12 o'clock position. Data from the right eyes was analyzed. Two raters independently measured TMH and TMA using the OCT images, and one rater repeated the measurements. Intra-rater, inter-rater and inter-device repeatability of measurements were assessed using Bland-Altman plots and pooled standard deviation. For intra-rater repeatability, TMH and TMA measurements were more repeatable in Cirrus than Visante (95% limits of agreement (LOA): TMH (μm), -22 to 66 (Cirrus), -125 to 45 (Visante); TMA (μm(2)), -1632 to 5331 (Cirrus), -38,050 to 21,874 (Visante)). For inter-rater agreement, TMH and TMA were also more repeatable in Cirrus than Visante (95%LOA: TMH (μm), -29 to 107 (Cirrus), -215 to 252 (Visante); TMA (μm(2)), -6650 to 9567 (Cirrus), -33,119 to 39,272 (Visante)). Inter-device agreement was poor (95%LOA: TMH (μm), -158 to 150; TMA (μm(2)), -32,903 to 14,076). There was no significant difference in inferior TMH between Cirrus and Visante (p>0.05). Inferior TMA was significantly lower in Cirrus by a mean difference of 10,223μm(2) (95% confidence interval, 5479, 14,966) (p=0.0002). Spectral-domain OCT is superior to time-domain OCT for intra-rater and inter-rater repeatability of TMH and TMA measurements.