Repeatability and Measurement Error in Optical Coherence Tomography Derived Retinal Thickness Measurements in Neovascular Age-Related Macular Degeneration

Abstract

Purpose: : Optical coherence tomography (OCT) has assumed a central role in determining retreatment with intravitreal anti-VEGF agents for neovascular age-related macular degeneration (nAMD). However, no studies have investigated the repeatability of automated retinal thickness measurement in these patients to distinguish clinical change from measurement error. The aim of this study was to determine the repeatability of automated StratusOCT^® retinal thickness measurements in patients with nAMD and to quantify errors in automated measurements.

Methods: : This is a retrospective analysis of OCT images from 50 consecutive patients attending for treatment of nAMD. In line with departmental protocols, two optimized fast macular thickness map protocol scan sets were acquired in a single imaging session. An ophthalmologist experienced in clinical trials work analyzed scans for the presence of segmentation error and manually measured the center-point thickness allowing calculation of the error in automated measurement. The patients were at different stages of treatment and disease activity.

Results: : The coefficient of repeatability was 74 µm for the central 1mm subfield for all 50 scan sets with 45 affected by segmentation error (90%). We calculated a revised coefficient of repeatability of 50 µm for the central 1mm subfield excluding scans in which all 6 line scans were affected by segmentation error. The median error of automated central retinal thickness measurements was 12% when segmentation error affected 2 or more line scans compared to an error of 1.5% in scan sets in which segmentation error affected less than 2 line scans (P < 0.0001, rank-sum test)

Conclusions: : This is the first study to report the repeatability of retinal thickness measurements in nAMD. The results of this study are important to both clinicians and clinical trial investigators when using quantitative OCT measures to determine retreatment decisions with anti-VEGF agents.

Keywords: imaging/image analysis: clinical • age-related macular degeneration