Repeatability of corneal measurements from the Casia2 anterior segment tomographer in a cataractous population.

Abstract

The purpose of this study was to investigate the repeatability of corneal power and thickness measures in a large patient cohort. In this retrospective non-randomised cross sectional single-centre study we evaluated a dataset containing 900 Casia2 anterior segment tomography measurements from 300 eyes from 300 patients (3 repeat measurements each) taken prior to cataract surgery. Only complete measurements marked as 'Successful' and with a sequence of 3 measurements for each eye performed on the same day were considered. Keratometry (K), corneal back surface (P) and RealPower (R) were each decomposed into power vector components SEQ/C0/C45. The mean (MEAN) and standard deviations (SD) for each sequence of measurements were derived and analysed. For K the MEAN and SD SEQ/C0/C45 were 43.77/0.17/-0.02 D and 0.12/0.18/0.18 D, for P MEAN and SD were -6.13/-0.25/0.00 D and 0.01/0.03/0.02 D, and for R MEAN and SD were 42.77/-0.06/-0.03 D and 0.13/0.19/0.19 D. For K and R 20%/42%/41% and 24%/45%/44% of repeat measurements respectively fell outside limits of ±1/8 D. The cornea was slightly thicker at the apex than at the thinnest point (544.6 vs. 536.2 µm) and SD was 4.3 vs. 3.9 µm. Keratometry underestimates against-the-rule corneal astigmatism by around 0.23 D. The variation of corneal power with repeat measurements is in a range of ±1/8 D for keratometric and RealPower SEQ. Where the Casia2 is used for lens power calculation, repeat measurements could help to improve the refractive outcome.