Purpose: To analyze and define the possible errors that may be introduced in keratoconus classification when the keratometric corneal power is used in such classification.
Materials and methods:Retrospective study including a total of 44 keratoconus eyes. A comprehensive ophthalmologic examination was performed in all cases, which included a corneal analysis with the Pentacam system (Oculus). Classical keratometric corneal power (P k ), Gaussian corneal power (P c Gauss ), True Net Power (TNP) (Gaussian power neglecting the corneal thickness effect), and an adjusted keratometric corneal power (P kadj ) (keratometric power considering a variable keratometric index) were calculated. All cases included in the study were classified according to five different classification systems: Alió-Shabayek, Amsler-Krumeich, RabinowitzMcDonnell, collaborative longitudinal evaluation of keratoconus (CLEK), and McMahon.Results: When P k and P kadj were compared, differences in the type of grading of keratoconus cases was found in 13.6% of eyes when the Alió-Shabayek or the Amsler-Krumeich systems were used. Likewise, grading differences were observed in 22.7% of eyes with the Rabinowitz-McDonnell and McMahon classification systems and in 31.8% of eyes with the CLEK classification system. All reclassified cases using P kadj were done in a less severe stage, indicating that the use of P k may lead to the classification of a cornea as keratoconus, being normal. In general, the results obtained using P kadj , P c Gauss or the TNP were equivalent. Differences between P kadj and P c Gauss were within ± 0.7D.
Conclusion:The use of classical keratometric corneal power may lead to incorrect grading of the severity of keratoconus, with a trend to a more severe grading.