Accuracy of Corneal Thickness by Swept-Source Optical Coherence Tomography and Scheimpflug Camera in Virgin and...

To assess intraobserver repeatability, intersession reproducibility, and agreement of swept-source Fourier-domain optical coherence tomography (SS-OCT) and the Scheimpflug camera in measuring corneal thickness in virgin and grafted eyes with Fuchs endothelial corneal dystrophy (FECD).

Thirty-six control eyes, 35 FECD eyes, 30 FECD with corneal edema eyes, 25 Descemet stripping automated endothelial keratoplasty (DSAEK) eyes, and 29 Descemet membrane endothelial keratoplasty (DMEK) eyes were included. The apical center, pupillary center, and thinnest corneal thickness were determined in 3 consecutive images and repeated 2 weeks later. Repeatability and reproducibility coefficients, intraclass correlation coefficients, and 95% limits of agreement (LOA) between measurements were calculated. Agreement between devices was assessed using Bland-Altman analysis.

Corneal thickness measurements were highly reproducible and repeatable with both systems. SS-OCT showed better repeatability in all corneal locations in the normal, FECD, FECD with edema, DSAEK, and DMEK groups (coefficient of variation ≤0.60%, ≤0.36%, ≤0.43%, ≤1.09%, and ≤0.48%, respectively) than the Scheimpflug (coefficient of variation ≤1.15%, ≤0.92%, ≤1.10%, ≤1.25%, and ≤1.14%, respectively). Between-session 95% LOA for SS-OCT was less than 3% for all groups except for the FECD with edema group, being almost double using the Scheimpflug camera. Differences between instruments were statistically significant in all groups and locations (P < 0.01) except in the DSAEK group (P ≤ 0.51); however, SS-OCT underestimated all measurements.

SS-OCT provides better reproducible and repeatable measurements of corneal thickness than those obtained with the Scheimpflug camera in patients with FECD or an endothelial transplant. Variations between examinations higher than the 95% LOA observed in our study should raise awareness of changes in the endothelial function.