Abstract
Purpose:
Keratoconus (KCN), a corneal ectasia, is diagnosed by abnormal corneal topography or tomography (Orbscan). The Ocular Response Analyzer (ORA) is a newer tool that measures corneal biomechanics and uses its non-FDA approved KCN Match Index software (KMI) to compare 37 biomechanical variables to a patient database. This study aims to evaluate the agreement between Orbscan indices and KMI values.
Methods:
This retrospective study included patients (>18y) who underwent both Orbscan and ORA KMI testing. The ORA KMI software assigns probability percentages in 5 classes from normal to severe KCN. The Orbscan measures properties of the anterior and posterior corneal surfaces. Using the KMI value as ‘truth’, we analyzed its agreement with Orbscan indices commonly used to diagnose KCN. Raw agreement scores were calculated, and those greater than .70 were considered significant. Cohen’s kappa agreement coefficients were determined for the significant indices, and those that were >.61 were considered to have substantial agreement with KMI scoring.
Results:
Among 5 patients (10 eyes), the ORA KMI values ranged from .017 to 1.214. KMI software classified 4 eyes as normal, 3 as KCN suspect, 3 as mild KCN, and none as moderate or severe KCN. Indices with significant raw agreement rates (>.70) were normal band scale (NBS, .7), thinnest corneal thickness (TCT, .9), irregularity index (II) at 3mm (.8) and at 5mm (.9), posterior float elevation above BFS (.9), and astigmatism difference between eyes (.7). Of these positive raw agreement rates, the following kappa coefficients were calculated between KMI and each index: NBS (.83), TCT (.78), II at 3mm (.6) and at 5mm (.78), posterior float elevation above BFS (.78), and astigmatism difference between eyes (.39). Corneal steepness, inferior steepening at 3mm and at 5mm, posterior BFS, and ratio of BFS radii had raw agreement rates ≤.7 and therefore they did not sufficiently agree with the KMI.
Conclusions:
Compared to KMI, the NBS, TCT, II at 3mm and 5mm, posterior float elevation above BFS, and astigmatism difference between eyes had the highest agreement rates (>70%). Of these, NBS, TCT, II at 5mm, and posterior float elevation above BFS had “substantial” agreement by kappa scoring and thus may be more sensitive for diagnosing KCN than other indices. By measuring corneal biomechanics the ORA KMI may be a useful adjunct in diagnosing KCN, especially in early subtle stages of the disease.
Keywords: 574 keratoconus