Longitudinal variability outcomes of frontloaded visual field testing

Abstract

Aim To assess if performing two visual field (VF) tests per eye on the same visit (frontloading) can increase perimetric data and decrease longitudinal variability compared with one test per eye. Methods A prospective longitudinal study of 498 healthy, glaucoma suspects and glaucomatous eyes of 333 subjects. Two intra-visit SITA-Faster VF tests (T1 and T2) per eye were performed on each visit for five consecutive visits. Rates of change and longitudinal variability of global (mean deviation) and pointwise sensitivity were compared between two approaches using linear mixed-effects models: use of the mean of T1 and T2 sensitivity values (frontloaded approach) versus the use of T1 sensitivity values alone (non-frontloaded approach). Results The mean duration of follow-up was 2.0 (SD 0.5) years. The mean rate of global sensitivity progression was similar in both frontloaded and non-frontloaded approaches (−0.12 vs −0.07 dB/year, respectively, p=0.3). The residual SD (1.19 vs 1.36, p<0.001) and mean absolute residuals (MARs) (0.69 vs 0.78, p<0.001) were significantly smaller in the frontloaded approach, indicating less variability. The mean rate of pointwise sensitivity change was significantly more negative in the frontloaded approach (−0.18 vs −0.14 dB/year, p<0.001), with significantly smaller MAR (1.24 vs 1.53, p<0.001). The frontloaded approach still displayed significantly lower variability even when unreliable tests (false positives >15%) were excluded. Conclusion Frontloading VF tests on the same visit may help clinicians meet the recommendations of minimum test frequency in glaucoma and can decrease the longitudinal variability of global and pointwise sensitivity changes. Data are available upon reasonable request. Not applicable.