Abstract
Purpose :
To compare identification of high risk glaucoma suspects based on either PERG or OCT assessment.
Methods :
The StopRGCD trial is a prospective randomized controlled trial assessing the usefulness of PERG in identifying high risk glaucoma patients candidates for randomization to treatment. We asked the question of whether high risk patients identified based on significant (p<0.05) PERG amplitude loss in at least one eye also displayed structural abnormalities consistent with glaucoma in corresponding eyes. Twenty glaucoma suspects recruited into the StopRGCD trial underwent baseline steady-state PERG (Jorvec Corp) and Cirrus OCT testing that included RNFL and GCIPL scans. PERG was recorded simultaneously from each eye from surface electrodes in response to horizontal square-wave gratings presented on a LED visual display (1.6 cycles/deg, 98% contrast, reversal rate 15.63 Hz, 25 deg square field; 800 cd/sqm mean luminance). Three expert clinicians reviewed baseline RNFL and GCIPL OCT scans and characterized OCT scans as either normal or consistent with early pre-perimetric glaucomatous changes. All clinicians were masked to PERG status and each other’s evaluations. Agreement was assessed with the kappa coefficient (agreement ≤0.4=poor, ≥0.75=excellent) and by computing sensitivity and specificity of OCT findings as a predictor of PERG abnormalities (PERG considered gold standard).
Results :
Kappa coefficients of agreement between PERG and OCT determinations differed for the 3 clinicians, ranging from poor (0.33, p=0.14), to fair (0.44, p=0.02), to moderate (0.57, p=0.01). One clinician had 100% specificity and 50% sensitivity in identifying patients with PERG abnormalities based on OCT scans, another had 92% sensitivity and 63% specificity, while the 3rd clinician had results intermediate between the two.
Conclusions :
There was agreement beyond chance between PERG determinations and clinician OCT assessments, which were statistically significant for 2 of 3 clinicians, but the agreement was no better than moderate as assessed with kappa. Clinicians varied in their interpretations of OCT abnormalities as judged against PERG, one with high specificity at the expense of sensitivity, another with high sensitivity at the expense of specificity, and a 3rd intermediate between the two. PERG assessment may complement OCT assessment in the identification of patients with pre-perimetric glaucoma.
This is an abstract that was submitted for the 2017 ARVO Annual Meeting, held in Baltimore, MD, May 7-11, 2017.