Purchase this article with an account.
Scott Schwartz, Milton M Hom; Effects of Dry Eye on Ocular Coherence Tomography in Glaucoma. Invest. Ophthalmol. Vis. Sci. 2019;60(9):5562.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
The effects of corneal drying over 15 seconds on temporal domain ocular coherence tomography(TD-OCT), were shown to affect scan quality. Glaucoma medications, especially prostaglandin analogues, have been shown to adversely affect tear film stability and osmolarity. Since modern day spectral domain (SD) OCT instruments scan more quickly (approximately 2 seconds), this study set out to evaluate the effects of dry eye on the reliability and results of SD-OCT.
This was a multicenter study with data collection from 2 sites. 52 consecutive patients with glaucoma and glaucoma suspects (104 eyes) were enrolled. Patients completed the Ocular Surface Disease Index (OSDI) Non-invasive tear breakup time (NIBUT) (Oculus Keratograph 5M) was performed. Eyes were classified as dry if they followed TFOS DEWSII criteria of NIBUT < 10 sec and OSDI ≥ 13.0. (Figure 1) 3 optic nerve head scans were averaged as were 3 ganglion cell complex scans. Signal strength was the primary outcome measure.
Patients in the dry eye cohort showed a directional trend in better signal strength on both the optic nerve head scan (ONH) (Mean(x)=57.14, Standard Deviation(σ)9.24,) and ganglion cell complex scan (GCC) (x=59.65, σ=8.46) compared with the non-dry eye ONH (x=53.08, σ=9.73 p=0.14) and GCC (x=55.01, σ=9.32 p=.07)
Dry eye, and similar inflammation seen from glaucoma medications has no statisticaly significant negative effect on SD-OCT and may even improve signal strength during testing. With the fast acquisition times of SD-OCT, previous concerns are mitigated. Clinically this is important, in our previous American Academy of Optometry Poster, it was shown that artificial tear use in SD-OCT testing reduces signal strength. Therefore, we believe there is no indication to compensate for ocular surface disease at the time of testing in most cases.
This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.
This PDF is available to Subscribers Only