Comparison of retinal nerve fiber layer thickness values using Stratus Optical Coherence Tomography and Heidelberg Retina Tomograph-III.

PURPOSE: To evaluate the usefulness of retinal nerve fiber layer (RNFL) thickness measurements using the Heidelberg Retina Tomograph-III (HRT-III) in normal, ocular hypertensive, and glaucomatous eyes and compare the thickness measurements using HRT-III and Stratus Optical Coherence Tomography-3 (OCT-3).

METHODS: Sixty-nine normal eyes, 60 eyes ocular hypertensive, and 111 glaucomatous were included. All participants underwent visual field, HRT-III, and OCT-3 examinations on the same day. Patients were classified into 3 groups according to intraocular pressure and visual field damage. The sensitivity/specificity of RNFL thickness measurements and RNFL thickness classifications using HRT-III and OCT-3 were calculated. The sensitivity/specificity of the height variation contour (HVC) from the HRT-III were calculated. The receiver operating characteristic curves (ROC) and areas under the ROC were plotted. Agreement was calculated using Bland-Altman method and the kappa coefficient.

RESULTS: The RNFL thickness sensitivity/specificity were 32.4%/87%, for the HRT-III and 72.97%/81.15% for the OCT-3 in relation to the glaucoma diagnosis (least specific criteria). The RNFL thickness sensitivities/specificities were lower in early glaucoma. The areas under the ROC for RNFL measurements were 0.72 using HRT-III, 0.86 with OCT-3 (P=0.001), and 0.54 for the HVC. The RNFL classification kappa coefficient was 0.36. Bland-Altman analysis confirmed that the RNFL measurements were not interchangeable.

CONCLUSIONS: The sensitivity of RNFL damage detection using HRT-III was lower compared with OCT-3, especially in early glaucoma. RNFL thickness agreement between HRT-III and OCT-3 was only fair. HVC was not useful for glaucoma detection.