Design and Clinical Evaluation of a Handheld Wavefront Autorefractor.

ABSTRACT

PURPOSE: To introduce a novel autorefractor design that is intended to be manufacturable at low cost and evaluate its performance in measuring refractive errors. METHODS: We developed a handheld, open-view autorefractor (the "QuickSee" [QS]) that uses a simplified approach to wavefront sensing that forgoes moving parts and expensive components. Adult subjects (n = 41) were recruited to undergo noncycloplegic refraction with three methods: (1) a QS prototype, (2) a Grand Seiko WR-5100K (GS) autorefractor, and (3) subjective refraction (SR). Agreements between the QS and GS were evaluated using a Bland-Altman analysis. The accuracy of both autorefractors was evaluated using SR as the clinical gold standard. RESULTS: The spherical equivalent powers measured from both autorefractors correlate well with SR, with identical correlation coefficients of r = 0.97. Both autorefractors also agree well with each other, with a spherical equivalent power 95% confidence interval of ±0.84 diopters (D). The difference between the accuracy of each objective device is not statistically significant for any component of the power vector (p = 0.55, 0.41, and 0.18, for M, J0, and J45, respectively). The spherical and cylindrical powers measured by the GS agree within 0.25 D of the SR in 49 and 82% of the eyes, respectively, whereas the spherical and cylindrical powers measured by the QS agree within 0.25 D of the SR in 74 and 87% of the eyes, respectively. CONCLUSIONS: The prototype autorefractor exhibits equivalent performance to the GS autorefractor in matching power vectors measured by SR.