Comparison of the reliability and repeatability of central corneal thickness measurements from four different non-contact devices

PURPOSE: To compare the reliability and repeatability of central corneal thickness (CCT) values from four different non-con-tact measurement devices.
METHODS: The study was conducted in 130 right eyes of 130 subjects with no ophthalmological pathology other than re-fractive errors. For each eye, data were recorded by making three consecutive measurements with a Scheimpflug camera (Pentacam, Oculus Optical gerate GmbH, Wetzlar, Germany), specular microscope (SM) (Cellchek XL; Konan Medical USA, Torrance, CA, USA), Lenstar LS 900® (Haag-Streit AG, Switzerland), and anterior segment optical coherence tomography (AS-OCT) (Carl Zeiss Meditec, Inc. Dublin, CA, USA). All measurements were analyzed using intraclass correlation coefficients (ICC), ANOVA or Friedman test, and Bland-Altman plots.
RESULTS: There were no statistically significant differences among the three consecutive measurements made with four de-vices (p=0.449, p=0.270, p=0.540, p=0.881, respectively). ICC values were 0.972, 0.997, 0.998, and 0.998, respectively. The closest agreement between measurements was a difference of 12.87 μm (95% limits of agreement [LoA]: −5.41, 20.33 μm) between AS-OCT and Pentacam, while the lowest agreement was between SM and Lenstar measurements, which had a difference of 31.92 μm (95% LoA: −21.80, 42.04 μm). This difference was 14.66 μm (95% LoA: −19.18, 10.14 μm) for AS-OCT and Lenstar, 31.86 μm (95% LoA: −17.22, 46.50 μm) for AS-OCT and SM, 30.22 μm (95% LoA: −23.04, 37.40 μm) for SM and Pentacam, and 17.28 μm (95% LoA: −14.34, 20.22 μm) for Pentacam and Lenstar.
CONCLUSION: The CCT measurements of the four different devices are highly consistent and have high repeatability. The highest ICC values were obtained with the SM, while the lowest ICC value was obtained with AS-OCT. Differences in average CCT values were similar between the AS-OCT, Lenstar, and Pentacam devices, while the difference was greater with SM. In clinical practice, CCT measurements obtained with SM should not be used interchangeably with measurements obtained with the other three devices.