Analysis of Four Aberrometers for Evaluating Lower and Higher Order Aberrations

PURPOSE: To compare the measurements of lower and higher order aberrations (HOA) of 4 commonly used aberrometers. SETTING: Massachusetts Eye & Ear Infirmary, Boston, USA. DESIGN: Prospective, cross-sectional study, in a controlled, single-blinded fashion. METHODS: Multiple readings were obtained in 42 eyes of 21 healthy volunteers, at a single visit, with each of the following aberrometers: Alcon LADARWave®, Visx WaveScan®, B & L Zywave®, and Wavelight Allegro Analyzer®. Results were compared and analyzed in regards to the lower and HOA, to the different wavefront sensing devices and software, Tscherning and Hartmann–Shack and between the Fourier and Zernike algorithms. Statistical analysis included Bland-Altman plots, Intraclass Correlation Coefficient (ICC), multiple comparison tests with Analysis of Variance and Kruskal-Wallis. Significant level was set to p<0.05 and alpha level correction was adjusted under the Bonferroni criteria. RESULTS: Most measurements of all 4 aberrometers were comparable. However, statistically significant differences were found between the aberrometers in total HOA (tHOA), spherical aberration (SA), horizontal coma and astigmatism (2,2). LADARwave and Wavescan showed significant differences in tHOA (P<0.001, ICC = 0.549, LoA = 0.19±0.5) and in SA (P<0.001, ICC = 0.733, LoA = 0.16±0.37). Wavescan showed a significant difference compared to Zywave (p<0.001, ICC = 0.920, LoA = 0.09±0.13) in SA. Comparisons between Allegro Analyzer and Zywave demonstrated significant differences in both Horizontal Coma (3,1) (p<0.001, ICC = −0.207, LoA = −0.15±0.48) and Astigmatism (2,2) (P = 0.003, ICC = −0.965, LoA = 0.2±2.5). Allegro Analyzer also differed from Wavescan in Horizontal Coma (3,1) (P<0.001, ICC = 0.725, LoA = −0.07±0.25). CONCLUSIONS: Although some measurements were comparable predominately in the lower order aberrations, significant differences were found in the tHOA, SA, horizontal coma and astigmatism. Our analysis suggests that sensor design contributes to agreement in lower order aberrations, and Fourier and Zernike expansion might disagree in higher order aberrations. Therefore, comparison between aberrometers was generally possible with some exceptions in higher order measurements.