To evaluate the accuracy of 3 mm Best Fitted Sphere (BFS) centered on the thinnest point measurement to detect keratoconus and to compare this to BFSs with 10 mm float measurement.

In this cross sectional study, 42 eyes from 29 patients with confirmed keratoconus were compared with 49 normal eyes of 27 candidates of refractive surgery. Anterior and posterior corneal surfaces were analyzed using Orbscan IIz slit-scanning topography in both groups. BFSs with 10 mm float measurement and 3mm centered on the thinnest point measurement were compared.

For detecting keratoconus, higher overall predictive accuracies (area under the curve =100%) for anterior and posterior BFS 3mm with center of the thinnest point compared to the anterior (98%) and posterior (99.5%) BFS 10 mm float measurements were achieved. There is no cutoff point using BFS 10 mm measurements for anterior float (range 42-43 diopters) and posterior float (range 52-54) for discriminating keratoconus from normal eyes. However, with a cutoff point of 45.4 diopters and 62.75 diopters for anterior and posterior BFS 3mm measurements, respectively, neither normal eyes, nor keratoconus eyes were misclassified.

Compared to the BFS 10 mm float measurement, BFS 3 mm centered on the thinnest point effectively discriminates keratoconus from normal corneas.