In vivo and in vitro repeatability of Hartmann-Shack aberrometry

To assess the in vivo and in vitro repeatability of objective refraction and higher-order aberrations (HOAs) measured by a commercially available Hartmann-Shack wavefront sensor.

Department of Ophthalmology, Johann-Wolfgang-Goethe University, Frankfurt am Main, Germany.

After pupil dilation of 40 myopic or myopic, astigmatic eyes of 20 patients, wavefront measurements were performed 6 times in each eye and in a test object provided by the manufacturer by 2 experienced examiners using a Hartmann-Shack wavefront sensor (Zywave, software version 3.21, Bausch & Lomb). The mean standard deviation (SD) and the coefficient of variation (CV) for sphere, cylinder, and each Zernike polynomial were computed for a 7.0 mm pupil diameter. Vector analysis was performed for the astigmatism. After the data were subdivided into 2 groups with 3 measurements in each, one measurement that best matched the subjective manifest refraction was chosen in each group and the difference between them was calculated.

The mean SD (CV) was 0.15 diopter (D) (7%) for the sphere value of the predicted phoropter refraction and 0.16 D (22%) for astigmatism. Thirty-two eyes had an axis deviation of at least 10 degrees. Vector analysis revealed a mean SD of 0.24@109.8. Other results for mean SD and mean CV were as follows: total in vivo higher-order RMS, 0.097 microm, 13.4%; sphere in myopic test device, 0.034 D, 0.65%; sphere in hyperopic test object, 0.035 D, 0.72%. The difference between the 2 best-matched refractions was significantly different from zero (0.11 D, P<.001). The CV was significantly higher for HOAs than for the 2nd-order aberrations (defocus and astigmatism).

Repeatability of Hartmann-Shack aberrometry by the Zywave wavefront sensor was not satisfactory, particularly for small amounts of HOAs. Under these conditions, aberrometry measurements should be repeated several times and outliers should be excluded in calculating the means.