Luisa Pierro1, Lorenzo Iuliano, Marco Gagliardi, Alessandro Ambrosi, Paolo Rama, Francesco Bandello. 1. *MD †PhD Department of Ophthalmology, Vita-Salute University, San Raffaele Scientific Institute, Milan, Italy (LP, LI, MG, FB); Vita-Salute San Raffaele University, Milan, Italy (AA); and Cornea and Ocular Surface Unit, San Raffaele Scientific Institute, Milan, Italy (PR).
Abstract
PURPOSE: To assess agreement between one ultrasonic (US) and nine optical instruments for the measurement of central corneal thickness (CCT), and to evaluate intra- and inter-operator reproducibility. METHODS: In this observational cross-sectional study, two masked operators measured CCT thickness twice in 28 healthy eyes. We used seven spectral-domain optical coherence tomography (SD-OCT) devices, one time-domain OCT, one Scheimpflug camera, and one US-based instrument. Inter- and intra-operator reproducibility was evaluated by intraclass correlation coefficient (ICC), coefficient of variation (CV), and Bland-Altman test analysis. Instrument-to-instrument reproducibility was determined by ANOVA for repeated measurements. We also tested how the devices disagreed regarding systemic bias and random error using a structural equation model. RESULTS: Mean CCT of all instruments ranged from 536 ± 42 μm to 577 ± 40 μm. An instrument-to-instrument correlation test showed high values among the 10 investigated devices (correlation coefficient range 0.852-0.995; p values <0.0001 in all cases). The highest correlation coefficient values were registered between 3D OCT-2000 Topcon-Spectral OCT/SLO Opko (0.995) and Cirrus HD-OCT Zeiss-RS-3000 Nidek (0.995), whereas the lowest were seen between SS-1000 CASIA and Spectral OCT/SLO Opko (0.852). ICC and CV showed excellent inter- and intra-operator reproducibility for all optic-based devices, except for the US-based device. Bland-Altman analysis demonstrated low mean biases between operators. CONCLUSIONS: Despite highlighting good intra- and inter-operator reproducibility, we found that a scale bias between instruments might interfere with thorough CCT monitoring. We suggest that optimal monitoring is achieved with the same operator and the same device.
PURPOSE: To assess agreement between one ultrasonic (US) and nine optical instruments for the measurement of central corneal thickness (CCT), and to evaluate intra- and inter-operator reproducibility. METHODS: In this observational cross-sectional study, two masked operators measured CCT thickness twice in 28 healthy eyes. We used seven spectral-domain optical coherence tomography (SD-OCT) devices, one time-domain OCT, one Scheimpflug camera, and one US-based instrument. Inter- and intra-operator reproducibility was evaluated by intraclass correlation coefficient (ICC), coefficient of variation (CV), and Bland-Altman test analysis. Instrument-to-instrument reproducibility was determined by ANOVA for repeated measurements. We also tested how the devices disagreed regarding systemic bias and random error using a structural equation model. RESULTS: Mean CCT of all instruments ranged from 536 ± 42 μm to 577 ± 40 μm. An instrument-to-instrument correlation test showed high values among the 10 investigated devices (correlation coefficient range 0.852-0.995; p values <0.0001 in all cases). The highest correlation coefficient values were registered between 3D OCT-2000 Topcon-Spectral OCT/SLO Opko (0.995) and Cirrus HD-OCT Zeiss-RS-3000 Nidek (0.995), whereas the lowest were seen between SS-1000 CASIA and Spectral OCT/SLO Opko (0.852). ICC and CV showed excellent inter- and intra-operator reproducibility for all optic-based devices, except for the US-based device. Bland-Altman analysis demonstrated low mean biases between operators. CONCLUSIONS: Despite highlighting good intra- and inter-operator reproducibility, we found that a scale bias between instruments might interfere with thorough CCT monitoring. We suggest that optimal monitoring is achieved with the same operator and the same device.