Purchase this article with an account.
V. Kappou, R. Asaoka, N. G. Strouthidis, D. F. Garway-Heath; HRT-3 Moorfields Reference Plane (MRP): Effect on Rim Area (RA) Repeatability and Identification of Progression. Invest. Ophthalmol. Vis. Sci. 2008;49(13):3650. doi: https://doi.org/.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To assess the effect of the MRP on HRT RA repeatability and to apply it to an event analysis algorithm reported previously (Fayers, et al. Ophthalmol 2007; 114:1973-80).
MRP applies the standard RP (SRP) to baseline topographies and keeps the vertical distance between the SRP and the reference ring constant for follow-up images.Assessing Repeatability:MRP was applied to test-retest data (Strouthidis, et al. Br J Ophthalmol 2005 Nov;89(11):1427-32). Inter-observer, inter-visit RA repeatability coefficients (RCs) were calculated and stratified according to good, medium and poor image quality.Assessing progression:MRP was applied to HRT images acquired from 198 ocular hypertensive (OHT) and 21 controls examined prospectively using HRT ‘classic’ and visual fields (1993-2001). HRT ‘classic’ images were analysed using HRT 3 software. HRT progression was defined according to inter-visit sectoral RA difference exceeding inter-observer, inter-visit RA RC in 2 or more sectors, with confirmation required in at least one of two consecutive, subsequent, images.
The inter-observer, inter-visit RCs for each sector are summarised in the table below. Good quality images had standard deviation < 18, medium 18 - 25, poor quality > 25 (<21, 21 - 35 and > 35, respectively, for 320 RP using HRT ‘classic’ acquisitions).Estimated specificity of the progression algorithm was 95%, based on the number of improving subjects and number of progressing controls. Progression rate in the OHT cohort was 40% compared to 28% for the 320 RP. Likewise, there was a slight improvement in agreement with field progression (AGIS criteria) - 15.1% (MRP) compared to 12.1% (320 RP).
The MRP improves RA repeatability compared to the SRP for medium and poor quality images; the effect is equivocal for good quality images. Repeatability is similar between MRP and 320 RP, however application of the MRP to an event analysis resulted in improved detection and agreement with fields compared to the 320 RP.
This PDF is available to Subscribers Only