Abstract
Purpose: :
to estimate agreement between central corneal thickness (CCT) measurements obtained by ultrasound pachymetry (UP) or using the Pentacam® imaging system.
Methods: :
UP and Pentacam CCT measurements were made in 136 eyes of 68 glaucoma patients and 132 eyes of 66 healthy subjects. The GEE method was used to ensure independence of eye measurements in a single subject. Agreement was determined by calculating intraclass correlation coefficients (ICC) for the two sets of measurements obtained using the different instruments. By constructing Bland Altman plots, we examined whether differences between CCT measurements were affected by their size and determined the presence of systematic and/or proportional biases. Through multivariate regression analysis it was also established if the measurement differences observed were conditioned by the keratometric power of the steepest and flattest corneal axes, their orientation or by age.
Results: :
ICCs were 0.705 (p<0.0001) (95%CI 0.61-0.78) for the glaucomatous eyes and 0.578 (p<0.0001) (95%CI 0.452-0.681) for the normal eyes. Mean differences between the two sets of measurements were 1.139 for the glaucomatous eyes and 11.58 for the normal eyes. We detected a significant systematic bias only in the set of measurements recorded in the normal eyes (single sample t-test; t=-0.431, p=0.667 for the glaucomatous, and t=4.456, p<0.0001 for the normal eyes). Regression lines obtained from the Bland-Altmann plots indicated no proportional bias in the normal eyes (r=0.157; p=0.083) but identified such a bias in the glaucomatous eyes (r=0.205; p=0.016) with differences increasing with increasing CCT (B=-0.171).
Conclusions: :
CCT measurements differ when obtained by UP or using the Pentacam. When measurements are made in glaucomatous eyes, these differences are proportional while CCT readings in healthy eyes show systematic differences.
Keywords: intraocular pressure