Measurements of optic disk size with HRT II, Stratus OCT, and funduscopy are not interchangeable

Abstract

Purpose: To assess the interchangeability of optic disk size measurements using slit-lamp funduscopy, optical coherence tomography (OCT-3), and confocal scanning laser ophthalmoscopy (HRT-II) in clinical practice.

Design: Prospective nonrandomized clinical study.

Methods: Measurements of vertical disk diameter (VDD) were obtained with the three methods. Disk area was obtained from OCT and HRT printouts. True agreement between methods in measuring VDD was assessed using Bland-Altman graphs and 95% limits of agreement (LoA). Disks were classified as small, average, or large, and agreement between methods in this classification was assessed using kappa statistics.

Results: Forty-eight patients were enrolled (mean age 53.4 +/- 14.3 years). VDD (mean +/- SD) was 1.58 +/- 0.15, 1.70 +/- 0.22, and 1.90 +/- 0.24 mm with funduscopy, HRT, and OCT, respectively. Very large LoA were observed: -0.29 to 0.70 mm for OCT and HRT, -0.07 to 0.71 mm for OCT and funduscopy, and -0.29 to 0.53 mm for HRT and funduscopy. There was poor agreement (kappa < 0.4) in classification of disk size as small, average, or large whether disk diameter or area was compared and using two definitions of disk size.

Conclusions: We observed a large range of differences in estimating disk size with HRT, OCT, and funduscopy. This precludes interchangeable use of these measurements in clinical practice, and does not allow simple conversion formulas to be proposed. In addition, there is poor agreement between these methods in classifying disk size as small, average, or large. At present, estimation of both absolute and relative disk size can only be defined separately for each measurement modality.