Axial length measurement in eyes with diabetic macular edema - A-scan ultrasound versus IOLMaster

Objective: To compare axial length (AL) measurements obtained with ultrasound (US) and IOLMaster (software version 3.01.0294; Carl Zeiss Meditec, Jena, Germany) in eyes with diabetic clinically significant macular edema (CSME) and to examine the correlation between measurement difference and foveal thickness. Design: Prospective nonrandomized clinical study. Participants: Twenty-three healthy eyes of 23 participants and 21 eyes with CSME of 21 diabetic outpatients. Methods: The authors prospectively measured AL using US and IOLMaster in eyes of both groups. Foveal thickness was measured with optical coherence tomography. The difference between both methods was assessed using the t test, and its correlation to foveal thickness was evaluated by Pearson coefficient. Main Outcome Measure: Difference in AL measurements between IOLMaster and US. Results: Mean +/- standard deviation foveal thickness in healthy and CSME eyes was 201 +/- 26 mu m (range, 156-240) and 475 +/- 106 mu m (range, 331-758), respectively. Mean AL by US and IOLMaster in healthy eyes was 23.17 +/- 1.07 mm (range, 21.17-25.18) and 23.25 +/- 1.10 mm (range, 21.17-25.26), respectively (P = 0.51); in CSME eyes, it was 23.02 +/- 0.97 mm (range, 21.53-24.81) and 23.27 +/- 0.924 mm (range, 21.9-25.13), respectively (P<0.001). Correlation between AL difference and foveal thickness was poor in both healthy (r = 0.18, P = 0.40) and CSME (r = 0.02, P = 0.93) eyes. Conclusions: Axial length measurements using applanation A-scan US and IOLMaster in eyes with CSME differ statistically and clinically significantly. This is likely a result of the fundamentally different methodology of the 2 methods in measuring eyes with a pathologically thickened retina. A conversion formula based on foveal thickness can not be provided at present.