Objectives: To determine whether indices derived from corneal topography could be used to differentiate between three populations at varying degrees of risk for presenting keratoconus, but otherwise unremarkable clinical appearance.
Methods: Eighty subjects lacking clinical signs of keratoconus were divided into three groups according to their background. Fifty refractive surgery candidates were used as controls, to be compared with 15 clinically normal fellow eyes of subjects with unilateral keratoconus, and 15 eyes from first and second degree relatives of the former keratoconic patients. Corneal topography indices (curvature, wavefront, and thickness) were obtained from Placido disc (Keratron) and scanning-slit (Orbscan) devices. Receiver operator characteristic curve analysis (sensitivity, specificity, and area under the curve) was performed to highlight the ability of any single index to discriminate between groups, and cutoff values for a fixed 0.80 specificity were determined.
Results: Corneal wavefront indices (root mean square and vertical coma) exhibited the best performance for discriminating between controls and the fellow eye of unilateral keratoconus patients. Corneal curvature and thickness indices showed poor capability to differentiate between the groups. Relatives and controls were distinguishable by a single thickness parameter.
Conclusions: The proposed cutoff values between controls and patients at risk for keratoconus may prove useful for screening purposes.