Comparison of peak times of photopic flicker electroretinogram responses recorded using conventional and portable systems in Birdshot chorioretinopathy

Purpose Assessing disease activity in Birdshot uveitis can be challenging. The peak time of the 30 Hz flicker photopic electroretinogram (ERG) was shown to be sensitive for detecting retinal dysfunction. ERG testing is not readily available in many eye units. We investigated use of a portable device to obtain recordings in the clinic setting and compared these with conventional recordings. Methods Patients underwent testing with the portable device (RETeval, LKC Technologies) and conventional equipment (Espion ColorDome, Diagnosys). For portable recordings pupils were undilated and skin electrodes used; for conventional recordings a conductive fibre electrode was used with mydriasis. Flicker ERG peak times were recorded and were deemed abnormal if they exceeded the 95th centile from a healthy cohort (>190 healthy volunteers). Results 18 patients with Birdshot uveitis underwent recordings. Recordings with the portable device took c. 3 min and were well‐tolerated. Mean (SD) age was 56.3 (12.8) years. For portable recordings, mean (SD) peak times were 30.1 (3.7) and 30.2 (3.9) ms for right and left eyes respectively. For conventional recordings mean (SD) peak times were 31.4 (3.9) and 31.8 (4.0) ms respectively. Peak times were significantly shorter with the portable device (p = 0.026 and 0.002 for right and left eyes). Correlation between devices was high (0.83 and 0.89 for right and left eyes) and highly significant (p < 0.0001). Strength of agreement between the two methods was good (kappa statistic 0.64; 95% CI, 0.36‐0.92). Conclusions Correlation between the portable and conventional ERG recordings was high with a good level of agreement. Recordings were rapid and well‐tolerated. There were significant differences in peak times between the two methods highlighting the importance of normative data specific to each device.