Purpose: To evaluate the agreement of corneal curvature measured by a newly introduced swept-source anterior segment optical coherence tomography (CASIA2) and Pentacam (Oculus, Wetzlar, Germany).
Methods: Corneal curvature was measured in 49 healthy subjects (age: 24.8 ± 4.36) by using CASIA2 and Pentacam. Repeatability was evaluated for CASIA2, and agreement between the CASIA2 and Pentacam was assessed. The coefficient of repeatability, the relative coefficient of repeatability, the range of the limits of agreement were analyzed, and Bland-Altman plots were performed to assess the interdevice agreement of measurement.
Results: For the intradevice reliability of the CASIA2, the mean difference between the 2 examinations was from 0.02 to 0.14 D for the corneal curvatures and the limits of agreement range was from 0.45 to 2.42 D for the corneal curvatures; there was no statistically significant difference found between the first and second examinations obtained from CASIA2 (P > 0.05). For agreement between the 2 devices, the Bland-Altman analysis of the posterior corneal curvature showed a mean difference of 0.16 D (P = 0.047) in the flat posterior corneal curvature and 0.16 D (P = 0.046) in the steep posterior corneal curvature. For the measurement of anterior corneal curvature and true net power (TNP) with the Pentacam and CASIA2, there was no significant difference in the flat anterior corneal curvature (P = 0.585), steep anterior corneal curvature (P = 0.274), flat TNP (P = 0.685), and steep TNP (P = 0.447).
Conclusions: Our study indicated generally good agreement of corneal curvature measurement by using Pentacam and CASIA2. We consequently expect no difference in clinical outcome when measurements for corneal curvature, especially anterior corneal curvature and TNP, are performed with either of the devices.