Comparison of central corneal thickness in healthy eyes using ultrasound pachymetry, non-contact specular microscopy and a high-resolution Scheimpflug camera

Purpose To compare and analyze the interchangeability in measuring central corneal thickness (CCT) using ultrasonic pachymetry (USP, PACHMATE 2, DGH, Inc, Exton, PA, USA), non-contact specular microscopy (NCSM, CEM-530, Nidek CO, LTD, Gamagori, Japan) and a high-resolution Scheimpflug Camera (Pentacam HR, OCULUS, Wetzlar, Germany). Methods An observational, cross-sectional study was performed recruiting 216 volunteers, for a sample size of 216 eyes with no ocular abnormalities other than refractive errors. All subjects underwent pachymetric measurements obtained by USP, NCSM and Pentacam HR. Examinations were performed by the same examiner with USP always following the noncontact examinations. Results The mean CCT (+/- SD) was 560.30 +/- 38.80 mu m, 556.76 +/- 36.83 mu m and 547.31 +/- 35.28 mu m for USP, NCSM and Pentacam HR, respectively. The Bland-Altman analysis showed that the highest concordance was found between USP and NCSM, with differences between - 13.18 mu m and 20.26 mu m. For the differences between measurements obtained with Pentacam HR and USP, the differences at Bland-Altman plot were between - 28.25 and 13.57 mu m. The lowest concordance was found for the CCT values measured with Pentacam HR and NCSM, with differences between - 25.67 and 6.86 mu m. The intraclass correlation coefficient (ICC) between all pairs of measurements was between 0.979 and 0.987, suggesting that the mean measurements were strongly correlated. Conclusion Measurements obtained with all three devices had high correlation. USP and NCSM were found in good agreement and high concordance, too. The above results indicate that these two devices are interchangeable in clinical practice. Pentacam HR may be a useful alternative for measuring CCT; however, it significantly underestimates CCT and cannot be used interchangeably with the other devices that we studied.