The aim of this study was to analyze theoretically the errors in the central corneal power calculation in eyes with keratoconus when a keratometric index (nk) is used and to clinically confirm the errors induced by this approach.
Differences (ΔPc) between central corneal power estimation with the classical nk (Pk) and with the Gaussian equation () in eyes with keratoconus were simulated and evaluated theoretically, considering the potential range of variation of the central radius of curvature of the anterior (r1c) and posterior (r2c) corneal surfaces. Further, these differences were also studied in a clinical sample including 44 keratoconic eyes (27 patients, age range: 14–73 years). The clinical agreement between Pk and (true net power) obtained with a Scheimpflug photography–based topographer was evaluated in such eyes.
For nk = 1.3375, an overestimation was observed in most cases in the theoretical simulations, with ΔPc ranging from an underestimation of −0.1 diopters (D) (r1c = 7.9 mm and r2c = 8.2 mm) to an overestimation of 4.3 D (r1c = 4.7 mm and r2c = 3.1 mm). Clinically, Pk always overestimated the given by the topography system in a range between 0.5 and 2.5 D (P < 0.01). The mean clinical ΔPc was 1.48 D, with limits of agreement of 0.71 and 2.25 D. A very strong statistically significant correlation was found between ΔPc and r2c (r = −0.93, P < 0.01).
The use of a single value for nk for the calculation of corneal power is imprecise in keratoconus and can lead to significant clinical errors.