Intraobserver variability in optical coherence tomography.

PURPOSE: To assess the intraobserver variability of optical coherence tomography (OCT) in patients with clinically stable maculas. DESIGN: Retrospective, observational case series using a diagnostic instrument. METHODS: Retrospective study. SETTING. Private retina practice. PATIENT POPULATION. Twenty-two eyes of 19 patients, each studied twice within 4 months. OBSERVATION PROCEDURE. Optical coherence tomograph measurements of the macula obtained by two photographers. MAIN OUTCOME MEASURES. Ordinary least products (OLP) and Bland-Altman analysis of OCT data. RESULTS: No fixed or proportional bias was detected in foveal zone thickness and total macular volume between measurements of either of the OCT operators. The coefficient of repeatability for foveal zone thickness was 37.0 mum for photographer 1 and 34.8 mum for photographer 2. The coefficient of repeatability for total macular volume was 0.29 mm(3) for photographer 1 and 0.10 mm(3) for photographer 2. CONCLUSION: Foveal zone thickness and total macular volume measurements show low intraobserver variability when analyzed by OLP and Bland-Altman techniques. Bounds are given for foveal zone thickness and total macular volume for which longitudinal OCT measurements by the same operator can be considered to reflect natural history or intervention effects rather than in. traobserver variability. (C) 2004 by Elsevier Inc. All rights reserved.