Test-retest reliability of scotopic full-field electroretinograms in rabbits

Abstract

Purpose: To explore test-retest reliability of standard scotopic full-field ERG measurements in New Zealand White rabbits. The ERG is widely used for testing of retinal integrity after any ocular treatment. We here present detailed stimulus-response dependencies for single healthy and untreated animals, concentrating on test-retest reproducibility.

Materials and methods: Five New Zealand White rabbits (aged 8-10 weeks, weight about 2.0-2.5 kg) underwent binocular ERG measurements after intramuscular anesthesia and pharmaceutical pupillary dilatation at a baseline day and 10 days later. Eleven scotopic flash strengths (0.0001-10 cd s/m²) were presented. Variability was quantified via the 95% limits of agreement (LOA).

Results: The a-waves displayed the typical monotonic sigmoid amplitude increase with flash strength, and the b-waves peaked at 0.01 cd s/m², followed by a marked dip at 0.1-0.3 cd s/m². LOA of both waves went through a maximum in the dip region. LOA divided by mean amplitudes (relative LOA) was fairly flat over flash strength, around 20% beyond the dip. Intraindividual interocular variability was markedly lower, around 10%.

Conclusions: Scotopic ERG responses in rabbits display a region of high variability at 0.1-0.3 cd s/m²; beyond that region the amplitude-LOA is 20%, the interocular LOA being half that value. The use of intraindividual control eyes for testing any toxicity of ocular agents thus appears markedly more sensitive. As a rule of thumb, we found the relative 95% LOA as 33% between individuals, 20% across sessions and 10% between eyes.

Keywords: ERG; Rabbits; Test–retest reliability.