Agreement among 3 optical imaging methods for the assessment of optic disc topography

Abstract

Purpose: To assess the agreement of disc topography measurements between the Heidelberg Retina Tomograph (HRT II), Retinal Thickness Analyzer (RTA), and Optical Coherence Tomograph (StratusOCT).

Design: Observational cross-sectional study.

Participants: Forty-two randomly chosen eyes of 42 subjects.

Methods: Each subject underwent HRT II, RTA, and StratusOCT examination. Two experienced examiners drew the contour lines for the HRT II and RTA. Bland and Altman plots were used to evaluate agreement for each topographic parameter among the instruments. The Spearman coefficient of rank correlation was evaluated for each topographic parameter.

Main outcome measures: Agreement in the measurement of optic disc topography among 3 imaging instruments, as evaluated by regression-based 95% limits of agreement.

Results: For optic disc area, the agreement between HRT II-RTA and StratusOCT-RTA revealed the existence of proportional bias, indicated by significant slopes of the regression lines (P = 0.01 and P = 0.02, respectively). The 95% limits of agreement between instruments varied with the actual optic disc size measurement. Heidelberg Retina Tomograph disc area measurements tended to be consistently lower than StratusOCT disc area measurements (fixed bias). The Spearman correlation coefficient between the instruments ranged from r = 0.35 (rim area, HRT II-StratusOCT) to r = 0.91 (cup area, HRT II-RTA).

Conclusions: Moderate to high correlation was found in measurements of optic disc topography among different instruments. However, the analysis of agreement indicated important discrepancies among instruments. Therefore, these instruments should not be used interchangeably to obtain measurements of the optic disc for glaucoma diagnosis.