Submicrovolt full-field cone electroretinograms: artifacts and reproducibility

The clinical utility of submicrovolt full-field 30-Hz (cone) electroretinograms was assessed by quantifying their contamination by electrical and photoelectric artifacts from xenon-flash stimulators and their test-retest variation in patients with retinitis pigmentosa. Artifacts obtained in saline with four commonly used electrodes varied with electrode type and consisted of an early, brief electrical component and a superimposed, extended photoelectric component. Techniques for minimizing these artifacts are described. Electroretinogram recordings from patients with advanced retinitis pigmentosa or congenital rod monochromatism indicate that these artifacts can be virtually eliminated with bipolar lenses. To assess test-retest variation, narrow-band-filtered responses were obtained twice during 6 weeks from patients with amplitudes less than 1 μV; threshold criteria for signficant (p<0.05) change in amplitude with this technique were approximately 0.25 log unit for each of two different systems.