The influence of the learning effect on automated perimetry in a Turkish population

Abstract

To evaluate the influence of learning effect on the outcome of automated perimetry and to assess the factors associated with the learning effect in a Turkish population. The Swedish Interactive Threshold Algorithm (SITA) standard 30-2 test was performed on a Humphrey Visual Field Analyzer in 102 normal Turkish subjects who had not previously undertaken any form of perimetry. Each subject completed three testing sessions, each separated by at least one day. The SITA standard 30-2 testing sessions included both eyes. The right eye was always tested before left eye. The reliability parameters, test duration and visual field (VF) global indices (mean deviation [MD] and pattern standard deviation [PSD]) obtained from right eyes in the first and third sessions were compared to assess the learning effect. Change in these parameters between the first and third sessions was calculated, and the effect of gender, age and educational level on this change was evaluated. The reliability parameters, test duration and VF global indices each improved between the first and third sessions (P ≤ 0.001). The change in the parameters was greater in subjects over 50 years of age and with an educational level less than high school (P ≤ 0.01). No correlation was found between learning effect and gender. In normal Turkish subjects, a significant learning effect was observed during VF testing with the SITA standard 30-2. The learning effect was correlated with age and educational level.