Measuring the agreement of keratometry readings of four devices in eyes with keratoconus

To determine the keratometry measurement agreement using Sirius corneal topography and Scheimpflug camera, Tomey corneal topography, Topcon autokeratorefractometer, and Tomey OA-2000 optical biometry in eyes with different severity of keratoconus.

In this retrospective study 115 eyes in different stages of keratoconus were divided into 2 groups of mild (stage 1), and moderate to severe keratoconus (stages 2, 3), according to the Amsler-Krumeich classification. Keratometry measurements were obtained using Sirius corneal topography and Scheimpflug camera (phoenix V3.7.01.08), Tomey corneal topography (Tms SW22C-200S-200), Topcon autokeratorefractometer (KR8900), and Tomey optical biometry (OA-2000 Opt-Meas V.4E).

In group 1 All devices demonstrated fair agreement in average keratometry values (95% LoA range > 1 D). However, it was poorer for group 2 (95% LOA range > 3 D). In group 1 Bonferroni test revealed statistically significant difference in average K readings among (Topcon autokeratorefractometer 8900 and Tomey OA-2000 biometry, p < 0.01), and between (Tomey topography and Topcon autokeratorefractometer, p < 0.05). Also in group 2 significant difference was observed in the average keratometry of the most instruments (p < 0.01) except for the (Tomey topography and Topcon autokeratorefractometer) and (Tomey topography and Tomey OA-2000 biometry.

According to our investigation in mild, moderate and severe keratoconus the agreement in K reading between Topcon autokeratorefractometer, OA-2000 optical biometry, Sirius topography and Tomey topography was poor. The agreement declines especially in the steep meridian and it was not acceptable clinically. These devices should not be applied interchangeably.