Repeatability and Between-Instrument Agreement of the Discovery System

Abstract

Purpose : To assess the between-visit repeatability of refractive error (lower-order) and higher-order aberration measurements made with a new commercially-available aberrometer, the Discovery System, and to examine the between-instrument agreement of refractive error measurements between the Discovery System and the Grand Seiko WAM-5500 open-field autorefractor.

Methods : Five cycloplegic aberrometry measurements of the right eye were made on 25 young adults using the Discovery at two separate visits by the same examiner. Ten cycloplegic measurements were made at visit 2 with the Grand Seiko. Refractive error values (sphere, cylinder, and axis) and Zernike coefficients calculated over a 6-mm pupil through the 6th order were obtained from the Discovery System. Refractive error measurements from both instruments were converted to vector space (M, J₀, and J₄₅) and averaged. Higher-order RMS (HORMS) was calculated (3^rd through 6^th order). Between-visit repeatability and between-instrument agreement of refractive error was assessed using Bland-Altman difference versus mean plots. HORMS and spherical aberration (C4,0) repeatability was also evaluated. A t-test was used to compare the mean difference to zero (bias), and the 95% limits of agreement (LoA) were calculated.

Results : Mean (±SD) age and spherical-equivalent refractive error at visit 1 were 23.4 ± 1.7 years and -2.91 ± 1.85 D, respectively. The median interval between the two visits was 5 days (range: 1-14 days). When evaluating between-visit repeatability of the Discovery System, there was no between-visit bias for M, J₀, J₄₅, HORMS, or spherical aberration (all p>0.70), and the 95% LoA were ±0.34 D, ±0.14 D, ±0.15 D, ±0.095 µm, and ±0.068 µm, respectively. There was no bias between instruments for defocus (p>0.23), but for J₀ and J₄₅, the Discovery on average measured 0.19 D and 0.12 D more positive than the Grand Seiko, respectively (both p<0.01). The 95% LoA for M, J₀, and J₄₅ between the two instruments were ±0.82 D, ±0.32 D, and ±0.27 D, respectively.

Conclusions : The Discovery System is repeatable and appropriate for measuring lower- and higher-order aberrations longitudinally. Minimal differences in astigmatic measurements between the Discovery and Grand Seiko indicate that refractive error measurements between the instruments should not be used interchangeably.

This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.