Agreement of wavefront-based refraction with autorefraction and manifest refraction across refractive and astigmatic profiles in refractive surgery candidates

Abstract

Background: To assess the agreement between wavefront-based refraction (WFR) using a pyramidal aberrometer (PERAMIS) and conventional non-cycloplegic (NCR), cycloplegic (CR), and manifest refraction (MR) techniques across refractive types and astigmatic axes.

Methods: This cross-sectional study evaluated 111 right eyes of refractive surgery candidates. WFR from PERAMIS was compared with NCR, CR, and MR for sphere, cylinder, spherical equivalent (M), blur (B), and astigmatic vectors (J0, J45). Subgroup analyses examined ametropia status, myopia severity, and astigmatic axis.

Results: A consistent refractive trend: CR < MR < NCR < WFR was observed for M, even within the subgroups. WFR consistently yielded more myopic measurements than CR, MR, and NCR (M-MD: - 0.56 D (p < 0.001), - 0.39 D (p < 0.001), and - 0.10 D (p < 0.05), respectively). WFR demonstrated excellent correlation with NCR (ICC = 0.98), CR (ICC = 0.96), and MR (ICC = 0.97) for M. This correlation remained high for other refractive components (ICC > 0.90) except J45, where moderate correlation was observed (ICC: 0.75-0.82). The correlation was superior in myopic eyes (ICC values > 0.90 except for J45; MDs almost < 0.50 D) compared to hyperopic eyes (ICC values ranging from 0.71 (WFR-MR cylinder) to 0.93 (WFR-NCR sphere, MDs < 0.75 D). Furthermore, WFR showed an excellent correlation with NCR, CR, and MR across astigmatism types (M: ICCs: 0.93-0.99, MDs < 0.75 D). Despite the high correlation, the limits of agreement (LoA) were wide in the whole population, especially for WFR-CR (LoA for M: - 1.67:0.55 D) and WFR-MR (LoA for M: - 1.41:0.63 D), with similar patterns across all subgroups.

Conclusion: Pyramidal WFR may be a valuable complement to conventional refractive measurements, particularly in patients with myopia. However, its tendency toward myopic bias and wide LoAs limits its interchangeability with CR or MR. Clinicians should interpret WFR cautiously, especially in hyperopic eyes, and consider confirming measurements with subjective methods. These findings support the utility of pyramidal WFR as an efficient initial estimate in refractive evaluations, especially for surgical screening, but not as a standalone replacement for traditional refraction.

Keywords: Agreement; Consistency; Cycloplegic autorefraction; Manifest refraction; Non-cycloplegic autorefraction; PERAMIS; Wavefront refraction.