Effect of keratoconus grades on repeatability of keratometry readings: Comparison of 5 devices

PURPOSE: To determine the repeatability of keratometry (K) measurements of a Scheimpflug pachymeter (Pentacam), Placido topographer (Eyesys), scanning-slit corneal topographer (Orbscan), partial coherence interferometry (PCI) device (IOLMaster), and Javal manual keratometer with different grades of keratoconus. SETTING: Noor Eye Hospital, Tehran, Iran. DESIGN: Retrospective case series. METHODS: Keratometry was performed first with Scheimpflug pachymetry followed, in order, by Placido topography, scanning-slit corneal topography, PCI, and manual keratometry. Repeatability was examined in groups with a maximum K of less than 50.0 diopters (D), 50.0 to 55.0 D, and more than 55.0 D. RESULTS: Seventy-eight eyes of 45 keratoconus patients were assessed. In Group 1, repeatability was highest with Scheimpflug pachymetry and lowest with scanning-slit corneal topography (0.36 to 1.24). In Group 2, the intraclass correlation coefficients (ICCs) for maximum K ranged from 0.823 with the scanning-slit corneal topography to 0.974 with Scheimpflug pachymetry. The repeatability index for minimum K (0.53 to 2.11) and maximum K (0.60 to 1.92) in this group was highest with Scheimpflug pachymetry and with for scanning-slit corneal topography. In Group 3, the ICCs for minimum K and maximum K ranged from 0.890 to 0.990, and the repeatability index for minimum K varied between 1.66 with Scheimpflug pachymetry to 2.98 with Placido topography; for maximum K, the index was from 2.15 with PCI to 2.81 with the manual keratometer. CONCLUSIONS: In mild keratoconus, the 5 devices had acceptable repeatability in K readings. In cases with a maximum K reading greater than 55.0 D, all devices had reduced repeatability as a result of measurement errors; thus, measurements might not be so reliable.