Purpose: To assess intraobserver and interobserver reliability of central corneal thickness (CCT) measurements using optical low-coherence reflectometry (OLCR) technology and its agreement with spectral-domain optical coherence tomography (SD-OCT).
Setting: Rio Hortega University Hospital, Valladolid, Spain.
Design: Evaluation of diagnostic technology.
Methods: To analyze OLCR intraobserver repeatability, 1 examiner obtained 4 successive measurements. To study interobserver reproducibility, a different examiner obtained another CCT measurement. To determine agreement with SD-OCT, the first examiner also obtained CCTs. Intraobserver and interobserver within-subject standard deviation (S(w)), coefficient of variation (CV(w)), and limits of agreement (LoA) were obtained for OLCR reliability analysis; for study agreement, data were analyzed using the paired-sample t test and the LoA were calculated.
Results: For OLCR intraobserver repeatability, the S(w) and precision (1.96×S(w)) were 2.33 and 4.56 μm, respectively. The intraobserver CV(w) was 0.42%. For interobserver reproducibility, the S(w) and precision were 11.59 and 22.71 μm, respectively; the CV(w) was 2.10%. The mean difference between observers was -1.35 μm (95% confidence interval [CI], -3.97 to 1.26). The width of the LoA was 45.27 μm. The mean CCT difference between OLCR and SD-OCT was 5.68±11.46 μm (95% CI, 8.29-3.08 μm; P=.0001), and the width of the LoA was 44.93 μm.
Conclusions: Optical low-coherence reflectometry technology provided reliable intraobserver and interobserver CCT measurements. Although OLCR underestimated the pachymetry by less than 6 μm compared with SD-OCT, its interchangeability fell within the range of interobserver reproducibility. Both noncontact pachymetry measurements seem to be clinically useful and may be used interchangeably with minimum calibration adjustment.
Financial disclosure: No author has a financial or proprietary interest in any material or method mentioned.
Copyright © 2012 ASCRS and ESCRS. Published by Elsevier Inc. All rights reserved.