[Comparison of the specificity and sensitivity of various instrument-guided keratoconus indices and classifiers]

Are the classifications achieved by the Belin-Ambrósio (BA) enhanced ectasia module, the keratoconus indices of the Pentacam HR and the ectasia screening index (ESI) using CASIA anterior segment optical coherence tomography (OCT) comparable to the topographical keratoconus classification (TKC) of the Pentacam HR? Can the indices be used interchangeably to assist in the diagnosis of keratoconus?

This retrospective study examined 228 datasets (eyes) of patients with a mean age of 36.6 ± 13.6 years which were grouped into a non-keratoconus group (group I, n = 59) and a keratoconus group (group II, n = 169) according to the clinical assessment. From the data sets of these eyes the sensitivity and specificity of various ectasia and keratoconus indices (KI) were calculated with the help of receiver operating characteristics (ROC). Groups were compared using the Mann-Whitney and χ(2)-tests.

Groups I and II had a mean KI of 1.04 and 1.26 and a mean ESI of 3.0 and 66.6, respectively. The χ(2)-test showed no significant coincidence of the distance of the thinnest point from the apex with TKC (χ(2) > 2, p > 0.35). All other BA parameters as well as the ESI showed a significant coincidence with the keratoconus diagnosis and the classification of TKC (p < 0.001). The ESI and KI along with some BA parameters (Df, Db, Dp and D) showed a good separation between groups I and II with an area under the ROC curves of > 0.93.

The enhanced indices and classifiers, such as the BA module or the ESI were comparable with the purely anterior corneal surface based TKC.