Keratometric measurements with the IOLMaster™

Findings obtained in a project for a thesis included the following:

Reproducibility measurements of highly precise toroidal test spheres using keratometers from different manufacturers resulted in the following standard deviations of the flat and steep radii (in mm) when focusing precisely:

CL 110 (Carl Zeiss Jena)	0,015	0,010
CL 150 (Carl Zeiss Jena)	0,007	0,008
IOLMaster (Carl Zeiss Jena)	0,000	0,005
Javal (Haag-Streit)	0,037	0,046

The IOLMaster showed the lowest deviations in reproducibility measurements of the test sphere.

Reproducibility on the living eye (test subjects up to 60 years old with healthy eyes) was in the same order of magnitude for the instruments tested. The measurements of the radii showed the following standard deviations (in mm):

CL 150 (Carl Zeiss Jena)	0,017	0,025
IOLMaster (Carl Zeiss Jena)	0,018	0,031
Javal (Haag-Streit)	0,031	0,034

Measuring errors of 20 µm lead to a deviation from the target refraction of 0.1 dpt in the average eye.

The measurements with the IOLMaster resulted in slightly steeper radii than those with the CL 150, while those with the Javal showed slightly flatter radii. A torn tear film was not found to affect the keratometric measurement in test subjects with healthy eyes.

Under normal conditions, keratometric measurements are unlikely to take long enough for the tear film to be torn (the mean time in the group studied was 13 seconds, maximum 23 seconds, minimum 8 seconds).

The following standard deviations (in mm) were measured in patients at the Jena University Eye Hospital (80% of them over 61 years of age).

CL 110 (Carl Zeiss Jena)	0,020	0,030
IOLMaster (Carl Zeiss Jena)	0,022	0,026

More accurate measuring results were obtained with test subjects with eye diseases when they were asked to blink.
In test subjects with eye diseases, better results were obtained in the measurements using tear replacement fluid, but there was no difference compared with measurements taken after the test subject was asked to blink. However, the group of 15 patients was too small to allow general conclusions to be drawn.

Conclusions:

Keratometric measurements with the IOLMaster show very good reproducibility, at least as good as with conventional manual keratometers.
Typical standard deviations on the living eye (patients) were 22/26 µm.
There was less scatter in the measuring values if the test subject was asked to blink immediately prior to the measurement.
Measurements using tear replacement fluid (Corneregel Fluid) showed no improvement compared with those obtained after the test subject blinked.

Abstract

from the thesis of
Matthias Rütz
Jena College of Applied Sciences, Faculty of Ophthalmic Optics