Published in:
01-11-2020 | Ultrasound | Original Paper
Comparison of central corneal thickness measurement by scanning slit topography, infrared, and ultrasound pachymetry in normal and post-LASIK eyes
Authors:
Wing-cheung Ho, Philip Tsze-ho Lam, Thomas Yee-hang Chiu, Mandy Ching-man Yim, Fion Tung-ching Lau
Published in:
International Ophthalmology
|
Issue 11/2020
Login to get access
Abstract
Objectives
To compare central corneal thickness (CCT) measurements by scanning slit topography (SST), infrared pachymetry (IRP), and ultrasound pachymetry (USP), and their agreement in normal and post-laser in situ keratomileusis (LASIK) eyes.
Methods
Sixty normal and 35 post-LASIK subjects were recruited. Only one eye from each subject was analyzed. Non-contact pachymetry was performed first, and the order for SST (Orbscan IIz) and IRP (Tonoref III) was randomized for each patient, to be followed by contact USP (Echoscan US-4000). Pearson’s correlation, paired t test, and Bland–Altman plots were used to investigate association, difference, and agreement among different instruments respectively.
Results
The measurements obtained with the instruments were highly correlated. Compared to CCT determined by USP (CCTUSP), CCT determined by SST (after correction with acoustic factor) (CCTSSTC) was thicker by 7 µm in normal eyes (P < 0.001). There was no significant difference between CCTSSTC and CCTUSP (P = 0.128), but a thickness-dependent deviation in post-LASIK eyes (P = 0.003). The CCT determined by IRP (CCTIRP) was thicker than CCTUSP in normal (P < 0.001) and post-LASIK eyes (P < 0.001) and demonstrated proportional overestimation with thinner corneas, with less predictable ultrasonic equivalent corneal thickness in normal eyes. Conversely, CCTIRP significantly underestimated CCT compared to CCTSST and showed increasing underestimation with thinner corneas in both normal and post-LASIK eyes (both P < 0.001).
Conclusion
Central corneal thickness determined by SST, IRP and USP were not interchangeable or interconvertible, probably attributed to difference in methodologies. Compensation with algorithms may improve agreements amongst instruments.