RESUMO
PURPOSE: In prospective no-masking, comparative, crossover monocenter clinical trial, we aimed to evaluate whether the optimal subjective refraction technique varies with the keratoconus topography and to identify relevant topographic criteria. METHOD: This study included 72 keratoconus eyes with impaired visual acuity. Each eye tested three methods of refraction (Jackson cylinder, astigmatism dial, stenopeic slit), resulting in three eyeglass lenses. Patients were assigned to the group corresponding to the eyeglass lens offering the best visual acuity. Five topographical characteristics were collected via the Pentacam: mean keratometry (Km), maximum keratometry (Kmax), distance from corneal center to Kmax (dKmax), Belin/Ambrosio Display (BAD_D), and index of surface variance (ISV). RESULTS: Forty-six eyes were included in the dial group (64.8%), 23 eyes in the cylinder group (32.4%), and only 2 eyes in the slit group (2.8%); thus, we only compared dial and cylinder groups. The main analysis retrieved a significant probability to choose dial technic for BAD_D (p = 0.024); when BAD_D is > 9.71 (ROC threshold), the positive predictive value (PPV) = 89.5%, and for ISV, p = 0.012; when ISV is > 77, PPV = 89.1%. The sub-analysis of patients with different visual acuities between cylinder and dial confirmed these results with slightly different thresholds: the probability to choose dial technic was for BAD_D, p = 0.03; when BAD_D is > 7.55, PPV = 90%, and for ISV, p = 0.0084; when ISV is > 71, PPV = 88.5%. CONCLUSION: Refraction method is linked to topographic indices ISV and BAD_D. A BAD_D > 7.55 indicates the dial method. In addition to keratoconus screening and diagnosis, this study suggests a new application of the topographer to select a suitable refraction method for eyeglass prescription. TRIAL REGISTRATION: Study registered on the ClinicalTrials.gov database under n°: NCT04174209.