Ocular response analyzer (ORA) derived parameters compared to Sirius corneal topography in analyzing corneal pathology

Abstract
Purpose To evaluate the ability of Ocular Response Analyzer (ORA) to differentiate between normal and abnormal corneas as compared to Sirius corneal topography. Methods This retrospective study included 302 eyes of 151 patients. All patients underwent evaluation with ORA and Sirius corneal topography. Parameters included disease classification results on both instruments (device software classification), Surface asymmetry index (SAI) on Sirius, Corneal Hysteresis (CH), Corneal Resistant Factor (CRF), Keratoconus Match Index (KMI), Goldmann-correlated intraocular pressure (IOPg), Corneal compensated intraocular pressure (IOPcc) and waveform score (WS) on ORA. Results On Sirius, 198 eyes (65.6%) were classified as normal. On ORA, 121 eyes (40.1%) were documented as normal. Overall, 105 eyes (34.8%) were classified as normal and 88 eyes (29.1%) with non-normal classification on both Sirius and ORA. Of the 198 eyes classified as normal on Sirius, 53% were classified as normal, 39% as suspect and 8% as mild keratoconus on ORA (47% non-normal). Of the 121 eyes classified as normal on ORA, 87% were classified as normal, 6% as suspect, and 2% as keratoconus compatible on Sirius (13% non-normal). Four percent of the eyes classified as keratoconus compatible on Sirius were classified as normal on ORA. There was a significant difference when comparing normal and non-normal classifications between ORA and Sirius (p < 0.001) with poor agreement (Kappa=0.32). When including only normal and Keratoconus eyes in the analysis, good agreement was found between the two machines (Kappa=0.75). Conclusion According to our results there seems to be a significant difference between ORA and Sirius in their ability to differentiate between normal and non-normal eyes. As such, we recommend that these devices not to be used interchangeably for assessing patients prior to refractive surgery.