A_max to scotopic I_max diagnoses feline hereditary rod cone degeneration more efficiently than any other combination of long protocol electroretinogram parameters

Aim To evaluate two recent methods for detecting feline hereditary rod cone degeneration with maximum efficiency from a long full-field flash ERG protocol. One combines 12 of these measures in an equation that is derived from iterative principal components factor analysis. The other uses the amplitude of the a-wave to the brightest available flash alone. Methods We tested the original 12-parameter equation, by applying it to 50 new ERG series in 23 backcrossed cats. They were necessarily either heterozygous or homozygous for hereditary rod cone degeneration. A masked observer compared the ERG score and fundus examinations. We reanalyzed the old, new and combined data sets. Data sets with only one session per animal were analysed to avoid problems from non-random sampling. A two factor linear model of the a-wave was evaluated. Results The prior equation, applied to the new data, discriminated the groups as well as it had initially. In the reanalysis, group separation continued to increase with even fewer measures compared to the previously reported study. Eventually, one measure, the amplitude of the a-wave (a_max) to the brightest scotopic flash (I_max) discriminated the groups better than any other measure or combination of measures in all analyses, including data sets using only one session for each animal and in a two factor linear model of the a-wave. Conclusion A_max to I_max alone proved to be the best diagnostic criterion in all analyses. No linear model is likely to discriminate affected from unaffected animals more effectively because additional variables increased variance more rapidly than they increased discrimination. A_max to I_max may detect other rod and rod/cone dystrophies equally efficiently.