[Skip to Content]
[Skip to Content Landing]
Figure.
Bland-Altman Plots for Pairwise Comparisons of 3 Tests of Central Corneal Thickness
Bland-Altman Plots for Pairwise Comparisons of 3 Tests of Central Corneal Thickness

The dotted line represents the mean difference between the 2 tests, and the shaded area represents the 95% limits of agreement. CCT indicates central corneal thickness; IVCM, in vivo confocal microscopy; USG, ultrasonic pachymetry.

Table.  
Intra- and Interrater Repeatability of CCT Measurements Taken With 3 Different Devices
Intra- and Interrater Repeatability of CCT Measurements Taken With 3 Different Devices
1.
Herse  P, Siu  A.  Short-term effects of proparacaine on human corneal thickness.  Acta Ophthalmol (Copenh). 1992;70(6):740-744.PubMedGoogle ScholarCrossref
2.
Salvetat  ML, Zeppieri  M, Miani  F, Parisi  L, Felletti  M, Brusini  P.  Comparison between laser scanning in vivo confocal microscopy and noncontact specular microscopy in assessing corneal endothelial cell density and central corneal thickness.  Cornea. 2011;30(7):754-759.PubMedGoogle ScholarCrossref
3.
McLaren  JW, Nau  CB, Erie  JC, Bourne  WM.  Corneal thickness measurement by confocal microscopy, ultrasound, and scanning slit methods.  Am J Ophthalmol. 2004;137(6):1011-1020.PubMedGoogle ScholarCrossref
Brief Report
September 2016

Comparison of In Vivo Confocal Microscopy, Ultrasonic Pachymetry, and Scheimpflug Topography for Measuring Central Corneal Thickness

Author Affiliations
  • 1Department of Cornea and External Diseases, Aravind Eye Hospital, Madurai, India
  • 2Francis I. Proctor Foundation, University of California, San Francisco
  • 3Department of Ophthalmology, University of California, San Francisco
JAMA Ophthalmol. 2016;134(9):1057-1059. doi:10.1001/jamaophthalmol.2016.2183
Abstract

Importance  In vivo confocal microscopy could be useful in cases of fungal or acanthamoeba keratitis to determine the depth of infectious elements, but its accuracy in determining corneal thickness relative to more traditional techniques has not been well characterized.

Observations  In this cross-sectional observational case series, central corneal thickness was determined by in vivo confocal microscopy, ultrasonic pachymetry, and Scheimpflug topography in 47 normal corneas and 23 keratoconic corneas from November 2014 to July 2015. Analyses undertaken from July 2015 to August 2015 showed that confocal microscopy overestimated the central corneal thickness in nonkeratoconic eyes, measuring on average 5 μm thicker (95% limits of agreement [LoA], 54-64) than ultrasonic pachymetry and 14 μm thicker (95% LoA, 47-76) than Scheimpflug topography. The bias was more pronounced in keratoconic eyes, where confocal microscopy overestimated central corneal thickness by 50 μm relative to ultrasonic pachymetry (95% LoA, 77-178) and 52 μm relative to Scheimpflug topography (95% LoA, 69-174).

Conclusions and Relevance  Confocal microscopy overestimated central corneal thickness relative to the other instruments, but the magnitude of the differences were small, especially in nonkeratoconic eyes. This level of measurement bias is acceptable to us for determining the depth of fungal filaments and acanthamoeba cysts in infectious keratitis.

Introduction

In vivo confocal microscopy could be useful in cases of fungal or acanthamoeba keratitis to determine the depth of infectious elements. However, little is known about the relationship of corneal thickness measurements performed with in vivo confocal microscopy relative to more traditional techniques. In this study, we measured central corneal thickness with in vivo confocal microscopy, Scheimpflug topography, and ultrasonic pachymetry to determine the agreement between these methods.

Methods

In this prospective series, patients 18 years or older presenting to the Aravind Eye Hospital in Madurai, India, between November 2014 and July 2015 were enrolled into 2 convenience groups: (1) patients with keratoconus and (2) patients without corneal disease. Individuals with dry eye syndrome and contact lens users were excluded. A single eye was randomly selected from each study participant for central corneal thickness (CCT) measurements with 3 devices in the following order: Pentacam (Oculus Optikgeräte GmbH); followed by the application of a topical anesthetic to perform ultrasonic pachymetry (Pacscan 300P; Sonomed); and finally in vivo laser scanning confocal microscopy (Heidelberg Retinal Tomograph III equipped with Rostock Corneal Module; Heidelberg Engineering GmbH). Two examiners took 3 separate measurements with each test; the mean value was used for the present analysis. For Pentacam measurements, scans were repeated until 3 tests with a quality score of 90% or greater were obtained. For confocal microscopy, the digital micrometer gauge was set to 0 on visualization of the most superficial epithelial cells, and volume scans were taken until the endothelium was visualized; the clearest image of endothelial cells was used for the CCT estimate. This study received ethical approval from the Aravind Eye Care System Institutional Review Board and adhered to the tenets of Declaration of Helsinki. Written informed consent was obtained from each patient before examination.

Results

We measured 70 eyes from 70 patients (47 normal corneas and 23 with keratoconus). The mean (SD) age of the patients was 21.8 (5.5) years, and 47 of the patients were women. The mean CCT in normal and keratoconic eyes for each of the 3 tests is shown in the Table. On average, in vivo confocal microscopy overestimated CCT relative to both Pentacam and ultrasonic pachymetry (Figure). In normal corneas, in vivo confocal microscopy measurements were an average of 5 μm thicker than ultrasonic pachymetry measurements (95% limits of agreement [LoA], 54-64) and 14 μm thicker than Pentacam measurements (95% LoA, 47-76). The bias was more pronounced in keratoconic corneas, where confocal microscopy estimates were on average 50 μm thicker than ultrasonic pachymetry estimates (95% LoA, 77-178) and 52 μm thicker than Pentacam estimates (95% LoA, 69-174).

Discussion

We found that laser scanning in vivo confocal microscopy overestimated CCT relative to ultrasonic pachymetry and Scheimpflug topography, especially in keratoconic corneas. The underlying cause of the measurement bias is not clear, although possibilities include corneal compression, which is inherent in the confocal technique and may induce measurement error, as well as epithelial edema from topical anesthetic use.1

Our findings are consistent with a previous study2 comparing in vivo confocal microscopy with specular microscopy that found confocal microscopy to slightly overestimate corneal thickness. In contrast, a study3 of a clinical tandem scanning confocal microscope found measurements to be slightly lower than those recorded by ultrasonic pachymetry.

Conclusions

This study was performed on a convenience sample of patients with and without keratoconus, and therefore, its generalizability to other diseased corneas is unclear. It might be useful to repeat the study in eyes with infectious keratitis, although such a study could be difficult because ultrasonic pachymetry and Scheimpflug topography may not be as accurate in eyes with active corneal infiltrates and edema.

The 95% LoA for the in vivo confocal microscopy comparisons indicate that 95% of individuals would have a corneal thickness within approximately 50 to 75 μm of what would be measured on ultrasonic pachymetry or Pentacam. The primary advantage of using in vivo confocal microscopy to measure corneal thickness would be to identify the depth of involvement for fungal and acanthamoeba corneal infections. For this purpose, the observed level of variability is acceptable to us, although its utility should be confirmed in future studies.

Back to top
Article Information

Corresponding Author: Soundaram MeenakshiSundaram, MD, Department of Cornea and External Diseases, Aravind Eye Hospital, 1, Anna Nagar, Madurai 625 020, India (drsoundaram123@gmail.com).

Accepted for Publication: May 19, 2016.

Published Online: July 14, 2016. doi:10.1001/jamaophthalmol.2016.2183.

Author Contributions: Drs MeenakshiSundaram and Keenan had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: All authors.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: MeenakshiSundaram, Prajna.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Keenan.

Administrative, technical, or material support: Sufi, Prajna, Keenan.

Study supervision: Prajna.

Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported.

Disclaimer: None of the instruments used in the study were donated to Aravind Eye Hospital, and none of the instrument manufacturers were involved with the study.

Additional Contributions: Sumithra Rengasamy, DOT (Department of Cornea, Aravind Eye Hospital, Madurai), coordinated the study and assisted with study participant enrollment, data collection, and data management. Vijayakumar Balakrishnan, MSc (Department of Biostatistics, Aravind Eye Hospital, Madurai), provided statistical analysis for the study as part of his routine employment at Aravind Eye Hospital. Both contributors were compensated for their work.

References
1.
Herse  P, Siu  A.  Short-term effects of proparacaine on human corneal thickness.  Acta Ophthalmol (Copenh). 1992;70(6):740-744.PubMedGoogle ScholarCrossref
2.
Salvetat  ML, Zeppieri  M, Miani  F, Parisi  L, Felletti  M, Brusini  P.  Comparison between laser scanning in vivo confocal microscopy and noncontact specular microscopy in assessing corneal endothelial cell density and central corneal thickness.  Cornea. 2011;30(7):754-759.PubMedGoogle ScholarCrossref
3.
McLaren  JW, Nau  CB, Erie  JC, Bourne  WM.  Corneal thickness measurement by confocal microscopy, ultrasound, and scanning slit methods.  Am J Ophthalmol. 2004;137(6):1011-1020.PubMedGoogle ScholarCrossref
×