Abstract
Purpose: :
To compare false positive rates (FPR) of peripapillary and macular scans acquired with Cirrus and RTVue in healthy myopic eyes and to asssess the agreement between the two devices and between RNFL and ganglion cell scans in classifying subjects as normal or abnormal.
Methods: :
Forty-one healthy myopic eyes of 41 subjects were scanned with Cirrus to measure peripapillary RNFL and macular ganglion cell-inner plexiform layer (GCIPL) thicknesses thickness, and with RTVue to measure peripapillary RNFL and macular ganglion cell complex (GCC). FPRs were calculated using the devices’ predefined criteria of abnormality and compared between devices. Agreement between instruments and between RNFL and ganglion cell scans in classifying subjects as normal or abnormal was assessed with kappa (Κ) statistic.
Results: :
All eyes had a 20/20 visual acuity and normal VF. For RNFL, the FPR was 4.8% for average and ranged from 2.4% to 7.3% for quadrants on Cirrus, but was 0% for all parameters on RTVue. The differences in FPR between devices were insignificant (all P>0.05). The FPR based on the presence of any abnormal parameter was 19.2% on Cirrus and 0% on RTVue, P=0.002. Cirrus average, minimum and hemispheric GCIPL were falsely positive in 12.2% to 17%. FPRs ranged from 9.7% to 14.6% for RTVue average, focal loss volume and hemispheric GCC, with no significant differences compared to Cirrus GCIPL rates, all P>0.05. The overall Cirrus FPR (36.5%) was significantly higher than that of RTVue (14.6%), P=0.023. On Cirrus 33.3% of eyes with abnormal average GCIPL had abnormal average RNFL; no eye with abnormal average GCC had abnormal average RNFL on RTVue. The agreement between devices was fair (Κ=0.34) for ganglion cell and poor (Κ=0) for RNFL scans. The agreement between peripapillary and macular scans in classifying subjects as false positive was poor both on Cirrus (Κ=0.13) and RTVue (Κ=0).
Conclusions: :
There are no differences in FPRs of average and individual quadrant RNFL, and in FPRs of average and individual hemispheric macular ganglion cell between Cirrus and RTVue. Overall FPRs of peripapillary and macular scans seem higher with Cirrus than RTVue. The agreement between RTVue and Cirrus for classifying subjects as false positive based on RNFL was poor. The classification agreement between macular ganglion cell and peripapillary RNFL scans was also poor on both devices. The high FPR of OCT results in healthy myopic subjects may be a matter of concern during longitudinal monitoring of glaucoma.
Keywords: nerve fiber layer • macula/fovea • myopia