Comparison of Central and Peripheral Corneal Thickness Measurements With Scanning-Slit, Scheimpflug, and Fourier-Domain Ocular Coherence Tomography

Saturday, April 18, 2015: 1:56 PM
Room 1B (San Diego Convention Center)
Claudia E. Perez-Straziota, MD
J. Bradley Randleman, MD
Heather M. Weissman, MD

Purpose
To compare central, regional, and relational corneal thickness values obtained with multiple technologies in normal patients and to determine their equivalence and interchangeability.

Methods
Retrospective analysis of 100 eyes from 50 patients evaluated by ultrasound pachymetry (Pachette II) (U), scanning-slit (SS) (Orbscan II), Scheimpflug imaging (Sch)(Pentacam high resolution device) (P), and spectral-domain OCT (RTVue-100) obtained as average values (OCT-A) and point measurements (OCT-P).   Measurements included central corneal thickness (CCT) for all technologies, and thinnest corneal thickness for Scanning-slit, Scheimpflug, and OCT.  Peripheral thickness measurements were obtained at the 3mm radius in the Superior (S), Nasal (N), Inferior (I) and Temporal (T) regions.

Results
Central corneal thickness (CCT) values were 563.9 ± 36.1 (U), 570.9 ± 36.1(SS), 552.8 ± 33.8 (Sch), 550.5 ± 32.7 (OCT-A), and 549.4 ± 32.7 (OCT-P). Ultrasound and (SS) were significantly different between each other (p<0.0001), and both were significantly different than all other devices (p<0.0001), while Sch was similar to OCT-A and OCT-P (p=0.4).  Differences between CCT and thinnest corneal thickness were significantly different between all technologies except SS and OCT-A.  For peripheral values, almost all locations measurements were significantly different from one another (p<0.0001). Superior-inferior values and ratios were also significantly different form one another for almost all devices, with no consistent patterns detectible.

Conclusion
There are significant, clinically relevant differences between regional and relational thickness measurements obtained with Ultrasound, scanning-slit, Scheimpflug, and OCT devices.  Screening metrics devised for one system do not appear directly applicable to other measurement systems.