Repeatability of Zone Averages Compared to Single-Point Measurements of Maximal Curvature in Keratoconus

To evaluate the repeatability of curvature zone averages centered on the point of maximum curvature (Kmax) compared to that of the single-point Kmax.

Comparative reliability analysis.

Setting: American University of Beirut Medical Center, Beirut, Lebanon.

Sixty-five eyes of 65 adult keratoconus patients. Patients with other ocular disease, history of ocular surgery or trauma, and contact lens wear within 2 weeks of image acquisition were excluded.

Eyes were evaluated with 3 consecutive scans using the Galilei dual Scheimpflug-Placido system.

Repeatability of axial and instantaneous Kmax single points, and zone averages with radii of 0.1-2.0 mm, centered on them. Repeatability was assessed by within-subject standard deviations, repeatability limits (r), and intraclass correlation coefficients.

Axial curvature zone averaging yielded clinically acceptable repeatability only in eyes with Kmax ≤50 diopters (D), for radii of 1.5 mm and 2.0 mm (r = 0.87 D and r = 0.76, respectively, vs r = 0.91 for the single-point axial Kmax). Compared to instantaneous Kmax, clinically acceptable repeatability was achieved with instantaneous zone averages of at least 1.5 mm radius in eyes with Kmax ≤50 D (r = 0.99 D and r = 0.70 D, respectively) and 2.0 mm radius in eyes with Kmax >50 D (r = 2.28 D and r = 0.87 D, respectively). For all eyes, the repeatability limit of the location of Kmax was 0.82 mm and 0.80 mm for axial and instantaneous curvature, respectively.

Instantaneous curvature zone averages centered on Kmax yielded a greater improvement in repeatability than axial zone averages and reached clinical adequacy with radii of at least 1.5 mm, for eyes with Kmax ≤50 D, and with a 2.0 mm radius for eyes with Kmax >50 D.