Comparison of Corneal Thickness Measurements Using Galilei, HR Pentacam, and Ultrasound

Purpose: To assess the degrees of agreement in measuring corneal thickness in normal eyes between 2 noncontact systems based on the dual Scheimpflug system (Galilei; Ziemer) and rotating Scheimpflug imaging (HR Pentacam; Oculus) and also between each of these methods and the gold-standard method of ultrasound pachymetry. Methods: In a prospective study, measurement agreement was assessed in 47 eyes of 47 healthy subjects. All eyes were examined with each of the 3 devices. Measurements made with the Galilei and HR Pentacam were compared with those made with ultrasound. The central corneal thickness (CCT) and thinnest pachymetry of the Galilei and HR Pentacam were also compared. Results: The mean values of CCT obtained from Galilei, HR Pentacam, and ultrasound were 560.57 +/- 29.10, 542.31 +/- 30.50, and 548.61 +/- 29.92 mu m, respectively. The 95% limits of agreement of each of these devices with ultrasound were -11.03 and +34.94 mu m for the Galilei and -33.67 and +20.74 mm for the Pentacam. The 95% limits of agreement of Galilei with Pentacam in measurement of central and thinnest corneal thickness were -3.39 and +33.11 and -10.40 and +27.29 mu m, respectively. Although CCT measurements made with the HR Pentacam were thinner (P < 0.001), and those obtained with Galilei were thicker (P < 0.001) than those made with ultrasound, there were significant correlation between Galilei and ultrasound (r = 0.92; P < 0.0001) and between HR Pentacam and ultrasound (r = 0.89; P < 0.0001). The central and thinnest corneal thickness measurements made with the HR Pentacam were thinner (P < 0.0001) than those made with Galilei, but there was a significant correlation between Galilei and HR Pentacam in measuring CCT (r = 0.96; P < 0.0001) and thinnest corneal thickness (r = 0.95; P < 0.0001). Bland-Altman plots showed good agreement between these methods. Conclusions: Although differences between the devices were statistically significant, there was good correlation and agreement between Galilei and Pentacam in measuring central and thinnest corneal thickness. The corneal thickness measurements made with the HR Pentacam and Galilei also showed good correlation and agreement with those made with ultrasound.