Abstract
Purpose: :
Optical coherence tomography-based assessment of macular morphology and thickness is used to determine the need for retreatment with ranibizumab in the management of diabetic macular oedema (DMO). The test-retest variability of Spectralis OCT measurements in this condition has not previously been reported. The aim of this work was to determine the repeatability of retinal thickness measurements with the Spectralis OCT in patients with DMO. This will lead to a more evidence-based approach to retreatment of DMO with ranibizumab, by distinguishing clinical change from measurement variability.
Methods: :
43 eyes of 43 patients with centre-involving diabetic macular oedema underwent four consecutive scans on a single Spectralis OCT machine (Heidelberg Engineering) using the preset "fast macular volume" protocol, with the retinal tracking function enabled. Coefficients of repeatability were determined for thickness and volume measurements for the nine ETDRS subfields using methods described by Bland and Altman.
Results: :
The coefficient of repeatability for central subfield thickness was 8.04µm (95% CI 7.68-8.39μm) and for total macular volume was 0.20mm3 (0.19-0.21mm3).Repeatability coefficients for other subfields ranged from 5.66μm in the outer temporal subfield to 19.52μm in the inner superior subfield, with some of this variation explained by segmentation error. There was no correlation between mean retinal thickness and standard deviation in any subfield.
Conclusions: :
Our results suggest that a change of ≥ 9μm in the central retinal subfield is highly indicative of true clinical change rather than measurement variability. This repeatability estimate of Spectralis OCT based macular thickness change is of value when developing retreatment criteria for the use of ranibizumab in the treatment of DMO.
Keywords: imaging methods (CT, FA, ICG, MRI, OCT, RTA, SLO, ultrasound) • macula/fovea • diabetic retinopathy