Agreement and repeatability of central corneal thickness measurements by four different optical devices and an...

To compare the repeatability and agreement of central corneal thickness (CCT) measurements by spectral-domain optical coherence tomography (OCT), corneal topography (CT) with a combined Scheimpflug-Placido system, optical biometry (OB), specular microscopy (SM), and ultrasound pachymetry (UP).

A single observer measured CCT twice in 150 eyes of 150 subjects with each of five devices: Nidek RS-3000 Advance OCT, CSO Sirius combined Scheimpflug-Placido disc system CT, Nidek AL-Scan partial coherence interferometry-based OB, Tomey EM-3000 SM, and Reichert iPac ultrasonic pachymeter. Pachymetry values corrected by the SM device software were also recorded. Levels of agreement between devices were evaluated by Bland-Altman plots with 95% limits of agreement, and repeatability for each device was analysed with intraclass correlation coefficients.

The mean CCTs measured by OCT, CT, OB, SM, corrected SM, and UP were 544.60 ± 29.56, 536.19 ± 32.14, 528.29 ± 29.45, 524.88 ± 32.38, 537.88 ± 32.38, and 545.29 ± 30.75 μm, respectively. Mean CCT differed significantly between the devices (p < 0.05) apart from between OCT and UP, and between CT and corrected SM. Mean paired differences between devices ranged between 0.68 and 20.41 μm. Repeatability with all devices was excellent (> 0.99). The range of limits of agreement was the least between OCT and UP.

Different CCT measurement techniques produce quite different results, so CCT evaluation and follow-up should be performed using the same device or devices with close compatibility.