Reproducibility, Repeatability, and Correlation of Central Corneal Thickness Measurement with the Pentacam Scheimpflug System and Ultrasound Pachymetry.

Abstract

The importance of an accurate determination of central corneal thickness (CCT) relies on its diagnostic and therapeutic implications in glaucoma, corneal ectasias, corneal edema and endothelial function monitoring, and refractive surgery suitability, among others. We aimed to analyze the repeatability, reproducibility, correlation, and laterality variations of CCT measurements performed with the Pentacam HR and the standard of care ultrasound A-scan (USP). A cross-sectional study including CCT measurements of healthy individuals was performed by three independent examiners with the Pentacam HR and USP. Intra-observer and inter-observer variations were calculated with intra-class correlation coefficients (ICCs). Bland-Altman plots and 95% limits of agreement (95% LoA) were used to assess the agreement between devices. Linear correlation was calculated with Pearson's coefficient. Thirty individuals (60 eyes), including 10 (33.3%) men and 20 (66.6%) women, with a mean age of 30.0 ± 9.1years, were studied. No statistical differences were found in CCT measurements between Pentacam HR (range 500 - 609 µm) and USP (range 498 - 628 µm). There was a high degree of correlation in repeatability and reproducibility of each independent device (ICC > 0.90). Pearson's correlation between 1 vs. 2, 2 vs. 3, and 3 vs. 1 Pentacam HR attempts were 0.914, 0.958, and 0.925, respectively (p < 0.001). Corresponding results for USP were 0.957, 0.957, and 0.943 (p < 0.001). The Pentacam HR tended to overestimate CCT by a mean difference of 3.77 µm (95% LoA, - 24.9 - 18.4). Right eyes were also overestimated (- 3.6 ± 14.1 µm) with the Pentacam HR device, whereas left eyes were underestimated (1.3 ± 11.1 µm). The Pentacam HR device provides reliable operator-independent estimates of CCT. Right eyes exhibited a tendency to overestimate with the Pentacam HR. We suspect this difference is due to USP underestimation related to patients' position while performing the study. In clinically relevant scenarios, performing a third measurement and cautiously measuring right eyes can provide higher accuracy.