ISCEV standard full-field ERG reference limits from 407 healthy subjects, derived from transference and validation of reference data between electrode types and centres.

Purpose: To establish age-adjusted reference intervals for the ISCEV standard full-field electroretinogram (ERG) recorded with silver thread electrodes in the lower fornix, based on a combined reference sample involving recordings from reference subjects and transference of data between two centres and two types of electrode.

Methods: Silver thread lower fornix ERG reference data from two centres underwent verification for inclusion in the reference sample (n = 251). Comparison analysis was performed to determine whether gold foil reference data could be included in the silver thread reference range, directly or with adjustment. Reference subjects and patients underwent ERG testing with both silver thread and gold foil electrodes (n = 53) and skin electrodes (n = 41). A linear model, fitted to the electrode comparison data, was used to transform gold foil ERG reference data for inclusion in the reference sample (n = 156). The combined sample of 407 reference individuals was used to derive age-adjusted reference limits for the main DA 0.01, DA 3, DA 10, LA 30 Hz and LA 3 ERG components.

Results: Silver thread ERG reference data was sufficiently similar across two centres to justify combination into a single reference sample. Peak times for gold foil and silver thread ERGs were closely comparable (r² 0.75-0.98, Bland-Altman bias ≤ 1.6 ms for all ERG components), with LA 30 Hz peak time showing the highest agreement (bias: - 0.2 ms, 95% limits of agreement (LOA): - 1.1 to 0.7 ms, 'silver thread-gold foil'). There was a clinically significant amplitude difference between electrode types: silver thread ERGs were 55-65% of the amplitude of gold foil ERGs (LOA ranged from 29 to 90%) and skin ERGs were 35-38% of the amplitude of silver thread ERGs (LOA ranged from 18 to 54%). Pooled reference data formed a sufficient sample covering 8 decades, from which age-adjusted parametric and nonparametric reference limits were calculated with reference to current guidelines.

Conclusions: ISCEV standard silver thread ERG data were consistent across the two centres, allowing transference of reference data. Reference data recorded with gold foil electrodes can be transformed for inclusion in a silver thread ERG reference range. The study highlights methods of pooling multiple sources of reference data into a larger, more robust sample, pertinent to standardization, clinical management, and multi-centre studies. These reference data could be adopted by other centres or combined with other datasets, following suitable verification.