Sanjay V Patel1, David O Hodge2, Emily J Treichel3, Matthew R Spiegel2, Keith H Baratz3. 1. Department of Ophthalmology, Mayo Clinic, Rochester, Minnesota, USA. Electronic address: Patel.Sanjay@mayo.edu. 2. Department of Health Sciences Research, Mayo Clinic, Jacksonville, Florida, USA. 3. Department of Ophthalmology, Mayo Clinic, Rochester, Minnesota, USA.
Abstract
PURPOSE: To determine if agreement between subjective interpretations of Scheimpflug tomography maps of corneas with Fuchs endothelial corneal dystrophy (FECD) is affected by daily and hourly changes in corneal edema. DESIGN: Reliability analysis. METHODS: Scheimpflug imaging pachymetry and posterior elevation maps of corneas with a range of severity of FECD were evaluated in a randomized manner for the presence of 3 tomographic features of edema. Agreement between interpretations of 1 masked observer was assessed (percentage, and κ-statistic with 95% confidence interval) for images taken within minutes on the same day, for images taken at a similar time on a different day, and for images taken over the course of a morning. Intra- and interobserver agreement was also assessed. RESULTS: Interpretations for individual tomographic features agreed for ≥88% of images (κ ≥ 0.75) taken within minutes on the same day; complete disagreement (ie, disagreement for all 3 tomographic features in an image) occurred in ≤3% of images. Interpretations agreed for ≥77% of images (κ ≥ 0.52) taken at a similar time on a different day; complete disagreement did not occur. Interpretations agreed for ≥81% of images (κ ≥ 0.61) taken over the course of a morning; complete disagreement occurred in ≤6% of images. Intraobserver agreement was ≥93% (κ ≥ 0.83) and interobserver agreement was ≥93% (κ ≥ 0.66); complete disagreement did not occur. CONCLUSIONS: Subjective interpretation of Scheimpflug images in FECD is highly repeatable for disease classification. Although small variations in interpretations resulted from pathophysiologic changes in corneal hydration and other factors, clinically significant disagreements in interpretation were uncommon and therefore unlikely to affect clinical decision-making.
PURPOSE: To determine if agreement between subjective interpretations of Scheimpflug tomography maps of corneas with Fuchs endothelial corneal dystrophy (FECD) is affected by daily and hourly changes in corneal edema. DESIGN: Reliability analysis. METHODS: Scheimpflug imaging pachymetry and posterior elevation maps of corneas with a range of severity of FECD were evaluated in a randomized manner for the presence of 3 tomographic features of edema. Agreement between interpretations of 1 masked observer was assessed (percentage, and κ-statistic with 95% confidence interval) for images taken within minutes on the same day, for images taken at a similar time on a different day, and for images taken over the course of a morning. Intra- and interobserver agreement was also assessed. RESULTS: Interpretations for individual tomographic features agreed for ≥88% of images (κ ≥ 0.75) taken within minutes on the same day; complete disagreement (ie, disagreement for all 3 tomographic features in an image) occurred in ≤3% of images. Interpretations agreed for ≥77% of images (κ ≥ 0.52) taken at a similar time on a different day; complete disagreement did not occur. Interpretations agreed for ≥81% of images (κ ≥ 0.61) taken over the course of a morning; complete disagreement occurred in ≤6% of images. Intraobserver agreement was ≥93% (κ ≥ 0.83) and interobserver agreement was ≥93% (κ ≥ 0.66); complete disagreement did not occur. CONCLUSIONS: Subjective interpretation of Scheimpflug images in FECD is highly repeatable for disease classification. Although small variations in interpretations resulted from pathophysiologic changes in corneal hydration and other factors, clinically significant disagreements in interpretation were uncommon and therefore unlikely to affect clinical decision-making.
Authors: Michael P Fautsch; Eric D Wieben; Keith H Baratz; Nihar Bhattacharyya; Amanda N Sadan; Nathaniel J Hafford-Tear; Stephen J Tuft; Alice E Davidson Journal: Prog Retin Eye Res Date: 2020-07-28 Impact factor: 21.198