Abstract
Purpose: :
1) To determine agreement between Fast and Slow Stratus Optical Coherence Tomography (OCT) macular measurements in patients with diabetic macular oedema 2) To use a grading system to assess the degree of macula oedema and evaluate the effect on method agreement of the Slow and Fast scanning protocols of the OCT.
Methods: :
50 eyes of 50 patients with diabetic macular oedema were scanned with Fast and Slow OCT Macular protocols, centred on the fovea. The imaging process was peformed in a random sequence between the two protocols and by two experienced operators. All patients underwent full clinical examination, refraction and pupil dilation prior to imaging. The degree of oedema was classified by 2 independent observers experienced in OCT interpretation. Using a grading system derived from histological descriptions of diabetic macular oedema, scans were graded on a scale from 1 (increased retinal thickness) to 4 (gross cystoid spaces) and then divided into two core groups: mild and gross. Bland Altman and Paired T Tests were used to calculate the data. A difference between protocol measurements of 20% or more was deemed clinically relevant.
Results: :
Of the 50 patients imaged, 8 were excluded due to poor image quality and inability of OCT software to correctly map the layers of the retina. Bland Altman showed greater method disagreement as the amount of oedema increased. The mean difference between foveal thickness measurements of the two scanning protocols (95% limits of agreement) was 0.2 (range –27.9 to 28.3 microns), P= 0.9148. 22 patients were graded as mild, 20 graded as gross. Analysis of the mean difference between foveal thicknesses classified as mild was 2.0 (range –15.5 to 19.6 microns). Gross oedema mean difference was 1.4 (range –36.6 to 33.8 microns).
Conclusions: :
Overall, we found no evidence of systematic bias between the two scanning protocols. Better agreement was found when evaluating thickness measurements in instances of mild oedema, whilst less agreement was found when comparing gross oedema. Although differences were observed, no cases were found to be of clinical significance, suggesting that both protocols are interchangeable. Future studies will explore longitudinal and test–retest variability.
Keywords: retina • diabetic retinopathy • imaging methods (CT, FA, ICG, MRI, OCT, RTA, SLO, ultrasound)