Comparative study of central corneal thickness measurement with slit-lamp optical coherence tomography and visante optical coherence tomography.

PURPOSE

To evaluate the repeatability and reproducibility of central corneal thickness (CCT) measurements obtained by 2 anterior segment optical coherence tomography (OCT) imaging systems and to examine their agreements with ultrasound pachymetry.

DESIGN

Observational cross-sectional study.

PARTICIPANTS

Fifty eyes from 50 healthy normal subjects were recruited.

METHODS

In one randomly selected eye in each subject, CCT was measured by slit-lamp OCT (SLOCT), Visante OCT, and ultrasound pachymetry. For anterior segment OCT measurements, both automatic and manual CCTs were obtained. Twenty-five of the 50 subjects were invited for 2 more visits within a week to evaluate repeatability and reproducibility of CCT measurement.

MAIN OUTCOME MEASURES

Central corneal thickness measurement obtained by the 3 methods and their agreements. Intrasession and intersession within-subject standard deviation (S(w)), precision (1.96xS(w)), coefficient of variation (CV(w)) (100xS(w)/overall mean), and intraclass correlation coefficient (ICC) were calculated to evaluate repeatability and reproducibility.

RESULTS

Good repeatability and reproducibility were found for both automatic and manual CCT measurements obtained by SLOCT and Visante OCT. For intrasession repeatability, CV(w) and ICC values ranged between 0.9% and 1.2% and 0.96 and 0.98, respectively. For intersession reproducibility, the respective CV(w) and ICC values ranged between 1.2% and 1.4% and 0.94 and 0.96. Although no significant difference was found between automatic/manual SLOCT measurements and ultrasound pachymetry, automatic Visante OCT CCT (535.7+/-30.2 microm) was significantly less than CCT with ultrasound pachymetry (550.3+/-31.14 microm) (P<0.001). In contrast, manual Visante OCT measurement (558.8+/-32.8 microm) was slightly higher than ultrasound pachymetry (P<0.001). Nevertheless, SLOCT and Visante OCT measurement of CCT had 95% limits of agreement comparable to that of ultrasound pachymetry. The best agreement was observed in the manual SLOCT measurement (95% limits of agreement between -15.5 and 11.7 microm).

CONCLUSIONS

Both SLOCT and Visante OCT automatic and manual CCT measurements were reliable and showed comparable agreement with ultrasound pachymetry. Although the 2 anterior segment OCT imaging systems have similar design and working principles, clinicians should be aware of the differences in CCT measurement between the 2 anterior segment OCTs.