To flag a patient’s ERG result as normal or abnormal, the API needs something to measure it against — a set of reference values collected from people with healthy retinal function. That reference set is what “normative data” means here, and the API’s primary source for it is a 2025 dataset published by Baker and colleagues.
What the dataset is
The Baker et al. (2025) dataset reports ISCEV-standard full-field ERG reference limits drawn from 407 healthy adult volunteers, recorded at multiple UK clinical electrophysiology centres — including UCL’s Institute of Ophthalmology, Moorfields Eye Hospital, St Thomas’ Hospital, and King’s College London. It was published in Documenta Ophthalmologica.
Why 407 healthy volunteers, and why three age groups
Retinal responses change gradually with age, so a single reference range for every adult would be inaccurate at both ends of the age spectrum. The dataset instead splits its 407 subjects into three age bands — 18–35, 36–59, and 60–80 — and provides separate reference ranges for each of the five standard ISCEV recording protocols, and for each of four common electrode types. A result is only compared against the range that actually matches the patient’s age group and the equipment used to record them.
How the API uses it
When a recording is processed, the API’s NormativeReference component loads the relevant Baker et al. (2025) reference range for the patient’s age band, protocol, and electrode type, and uses it to calculate a Z-score — a statistical measure of how far the patient’s result sits from the healthy-population average. That Z-score is what ultimately drives the Green/Amber/Red traffic-light signal in the Layer 1 and Layer 2 report.
Why this particular dataset
Reference ranges collected at one clinic don’t automatically transfer to another — differences in equipment calibration, room lighting, and electrode technique can shift what counts as “normal.” A separate multi-centre study (Hamilton et al., 2015) tested exactly this, recording the same two volunteers at fifteen UK centres, and found the results were consistent enough across sites to be clinically interchangeable. That consistency check is part of why the Baker et al. (2025) dataset was chosen as the API’s primary reference standard, rather than a smaller, single-site alternative.
Full citation
Baker et al. (2025), ISCEV standard full-field ERG reference limits from 407 healthy subjects, Documenta Ophthalmologica. Full paper: https://link.springer.com/article/10.1007/s10633-025-10009-2