RESUMEN
PURPOSE: To detect keratoconus progression, accuracy of tomographic measurements is crucial. The impoved axial resolution of optical coherence tomography (OCT) compared to Scheimpflug photography serves as the motivation to investigate and compare the repeatability of the anterior segment OCT MS-39 (CSO) to Pentacam HR (Oculus Optikgeräte GmbH) in patients with keratoconus. METHODS: One hundred twenty-three eyes of 123 patients with keratoconus were enrolled and subdivided in four groups by maximum keratometry (Kmax): Kmax < 48.00 diopters (D), Kmax of 48.00 to 53.01 D, Kmax of 53.00 to 58.00 D, and Kmax > 58.00 D. Three consecutive measurements per eye were acquired with the MS-39 and compared to the Pentacam HR. Kmax, thinnest pachymetry, anterior asphericity, and posterior elevation data were compared. Within-subject standard deviation (Sw), coefficient of variation (CoV), test-retest repeatability (TRT), and the intra-class correlation (ICC) were calculated and evaluated. Bland-Altman plots were also analyzed. RESULTS: The Pentacam HR measures significantly higher Kmax values than the MS-39, with a more pronounced difference for severe cases of keratoconus (0.57 D for all cases; 1.88 D for cases with Kmax > 58.00 D). Thinnest pachymetry was approximately 5 µm thinner when measured by the Pentacam HR than the MS-39, independently of keratoconus stage. A further progressed keratoconus stage was significantly associated with increased measurement errors and resulted in worse repeatability (Kmax < 48.00 D: Sw = 0.18 D, TRT = 0.50 D, CoV = 0.39%, ICC = 0.989; Kmax > 58.00 D: Sw = 0.53 D, TRT = 1.48 D, CoV = 0.90%, ICC = 0.984). The behavior was similar for other tomographic parameters. CONCLUSIONS: The Pentacam HR and the MS-39 have an overall good agreement for keratoconus; however, the Pentacam HR measures steeper and thinner than the MS-39. The association between the magnitude of topographic and tomographic parameters and their measurement errors suggests that the diagnosis of keratoconus disease progression should be based on the stage and the test-retest repeatability rather than on a fixed value (eg, 1.00 D). [J Refract Surg. 2022;38(4):250-255.].
