Ocular component measurement using the Zeiss IOLMaster

Abstract

Purpose: Axial length is traditionally measured using A-scan ultrasound. The IOLMaster is a new instrument that uses partial coherence interferometry to measure axial length. We compared the repeatability of these techniques for both an experienced and an inexperienced observer, the agreement between the two techniques, and the effect of cycloplegia on IOLMaster measurements.

Methods: Five measurements of axial length and three measurements of anterior chamber depth were taken with the IOLMaster in two sessions separated by 1 to 12 days in 20 young adults. The two examiners each took measurements, and the subject was then cyclopleged with 1% tropicamide. The IOLMaster readings were then repeated by both examiners, followed by five ultrasound readings. Repeatability was evaluated by calculating the difference between measurements from the two sessions. The mean and standard deviation of these differences was then used to determine the 95% limits of agreement (LoA) for each technique. In addition, the agreement between the IOLMaster and ultrasound was assessed, along with the effect of cycloplegia on IOLMaster readings.

Results: The IOLMaster was more repeatable than ultrasound. For axial length, the 95% LoA were -0.11 to +0.07 mm, -0.06 to +0.05 mm, and -0.25 to +0.35 mm, for noncycloplegic IOLMaster, cycloplegic IOLMaster, and ultrasound, respectively. The two instruments showed modest agreement with each other (mean difference, +0.12 mm; 95% LoA, -0.39 to +0.64 mm; p > 0.0125). Cycloplegia had no significant effect on IOLMaster axial length measurements. The 95% LoA for anterior chamber depth measurement were -0.11 to +0.18 mm, -0.06 to +0.04 mm, and -0.19 to +0.21 mm, for noncycloplegic IOLMaster, cycloplegic IOLMaster, and ultrasound, respectively. The IOLMaster gave significantly longer anterior chamber depths than ultrasound (mean, +0.18 mm; 95% LoA, -0.02 to +0.37 mm; p < 0.0125), and cycloplegia produced significantly deeper anterior chamber depths using the IOLMaster (mean, +0.12 +/- 0.09 mm; 95% LoA, -0.05 to +0.29 mm; t = 6.17; p < 0.001). The experienced observer's measurements were more repeatable than the inexperienced observer's for ultrasound, but not for the IOLMaster.

Conclusions: The superior repeatability of the IOLMaster suggests that it should become the standard for axial length measurement. The 95% limits of agreement for the cycloplegic measurements correspond to a change in refractive error of +/-0.12 D.