Association analyses of the measurements of the photopic negative response evoked by two ISCEV protocols

Abstract

Purpose: To perform association analyses between the measurements of photopic negative response (PhNR) evoked by two ISCEV protocols.

Methods: A total of 172 eyes from 72 post-operative pediatric cataract patients and 24 healthy children were enrolled. The amplitude and peak time of PhNR were analyzed in three eye groups, 1. healthy controls; 2. fellow eyes of unilaterally affected patients; 3. affected eyes. PhNR responses were measured with skin-electrodes and evoked by the ISCEV standard protocols of PhNR and light-adapted 3.0, referred to as PhNR1 and PhNR2. The correlation coefficients between PhNR1 and PhNR2 measurements were calculated. The generalized estimating equation (GEE) model of PhNR1, with PhNR2 as a predictor, was evaluated after adjusting for correlation between paired eyes.

Results: Both the amplitude (P = 0.025) and the peak time (P = 0.036) of PhNR1 showed a significant difference among the three eye groups, which was not observed in PhNR2. The four correlation coefficients (Pearson, Intraclass, Lin's and Kendall's) between z-score transformed PhNR1 and PhNR2 measurements were generally moderate: 0.52, 0.52, 0.52, 0.36 for amplitude (P < 0.001), and 0.57, 0.57, 0.57, 0.36 for peak time (P < 0.001). The amplitude of PhNR1 cannot be precisely predicted by PhNR2, with a mean absolute percentage error (MAPE) of 36.7%, while the peak time of PhNR1 can be precisely predicted with a MAPE of 3.9%.

Conclusions: PhNR1 appears to be a more sensitive measure than PhNR2 for detecting eye group differences. Further research is needed to confirm this and explore its clinical applications. PhNR1 may not be entirely replaced by PhNR2 due to moderate correlation and low prediction precision in amplitude.

Keywords: Association analysis; Children; Electroretinogram; Photopic negative response.