Repeatability, Reproducibility and Agreement of Central Corneal Thickness Measurements by Two Noncontact Pachymetry Devices

This study was designed to assess the repeatability, reproducibility, and agreement of Noncontact Pachymetry (NPC) (Topcan TRK-2P) and the non-contact optical low coherence reflectometer (Lenstar LS 900) devices for measuring Central Corneal Thickness (CCT) of healthy corneas. A total of 82 healthy volunteers were evaluated. The first observer used both the TopconTRK-2P and Lenstar 900 devices while the second observer only used the TopconTRK-2P for the measurements. The measurements with either device were repeated three times for each patient, consecutively. The central corneal thickness measurements with the Topcon TRK-2P revealed mean ± Standard Deviation (SD) values of 553.1 ± 36.1 micrometer (µm) for the first observer and 552.3 ± 35.9µm for the second observer and the mean ± SD of CCT was 537.3 ± 34.8µm with the Lenstar 900. The difference between the CCT measurements of the observers using the Topcon TRK-2P (P = 0.142) was insignificant. However, significantly lower measurements were found with the Lenstar 900 compared with the Topcon TRK-2P (P 0.001). The central corneal thickness measurements obtained by the Topcon TRK-2P were found to have high repeatability for both observers with a lower SD, less than 1% Coefficient of Variation (CV) and higher than 0.99 Intra-Class Correlation Coefficient (ICC) (Observer 1: 3.77 SD, 0.68 CV and 0.995 ICC; the second observer: 3.58 SD, 0.65 CV and 0.995 ICC). There was an excellent inter-observer reproducibility between the two observers for Topcon TRK-2P with 2.71 SD, 0.49 CV, and 0.994 ICC. The Bland-Altman plot showed high agreement between the two devices. These results suggest that the TopconTRK-2P is a reliable device for evaluating CCT in healthy corneas compared with Lenstar 900.