Purchase this article with an account.
Ou Tan, Xinbo Zhang, David Huang; Three Ways to Reduce False Positive Glaucoma Diagnosis in OCT Measurement of Retinal Nerve Fiber Layer Thickness. Invest. Ophthalmol. Vis. Sci. 2018;59(9):2119.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To reduce the dreaded “red disease” (false positive glaucoma diagnosis in high myopes) by the use of axial length, disc size and spherical refractive error information in glaucoma diagnosis based on retinal nerve fiber layer (RNFL) thickness as measured by OCT.
Two hundreds and forty six normal eyes and 298 peri-metric glaucoma eyes from the Advanced Imaging for Glaucoma study were analyzed. The RNFL thickness and disc area (DA) were obtained using the RTVue spectral OCT (Optovue). Axial length (AL) was obtained by IOLMaster (Zeiss). Eyes were grouped according to spherical equivalent (SE) refractive error: hyperope (>1D), emmetrope (+1 to -1D), low myope (-1 to -4D), and high myope (<-4D). The 1 percentile normative cutoff for overall, superior and inferior quadrant RNFL thickness average was adjusted using regression analysis against independent variables: AL, DA, or SE. Glaucoma is diagnosed if any of the 3 RNFL parameters was abnormal (below 1 percentile).
Axial length (p<0.001) and DA (p=0.06) were larger in the glaucoma group, compared to normal. Axial length, disc area and spherical error were significantly (p<0.001) correlated with RNFL thickness (Pearson R=-0.40, 0.28 and 0.24 respectively). The specificities of glaucoma diagnosis were 1.00, 0.98, 0.99, and 0.83 in the hyperope, emmetrope, low myope, and high myope groups, respectively; the sensitivities were 0.54, 0.57, 0.52, and 0.60. In high myopes, specificities were improved to 0.94, 0.94, and 1.00 using AL, DA, and SE adjustment, respectively; however, sensitivities were reduced to 0.40, 0.60, and 0.37 .
The specificity of glaucoma diagnosis based on RNFL thickness is poor for high myopes. It can be improved by the adjustment of diagnostic cutoffs using AE, DA, or SE information.
This is an abstract that was submitted for the 2018 ARVO Annual Meeting, held in Honolulu, Hawaii, April 29 - May 3, 2018.
This PDF is available to Subscribers Only