Purchase this article with an account.
Gislin Dagnelie, Ava K. Bittner, IRD01 Study Group; Positive Effect of Operator Training on Goldmann Visual Field Test-retest Reliability. Invest. Ophthalmol. Vis. Sci. 2012;53(14):4600.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To examine the effects of within vs. across operator test and of operator training on the repeatability of Goldmann visual field (GVF) test results. We have previously noted that GVF results (as expressed in total seeing retinal area in mm²) can depend to an important extent on the operator, and that change of operator adds a source of variability to the data (Koenekoop ARVO e-Abstr 3323, 2011). We retrospectively investigated this same vs. different operator effect, as well as that of recent operator training efforts on pre-treatment GVF data from the QLT, Inc-sponsored IRD01 synthetic retinoid trial.
Patients with Leber Congenital Amaurosis (LCA) or retinitis pigmentosa (RP) due to an RPE65 or LRAT mutation were enrolled in the trial. Each patient underwent, as part of screening and pre-treatment visits, two GVF administrations including one or more isopters. For each patient a single isopter was selected for which the retinal area on both administrations fell between 4 and 400 mm². Prior to training in the summer of 2011, at a single site with 5 operators, GVF data had been collected in 8 patients by the same operator on both visits, while data on 4 patients had been collected by mixed operators. Since training only a small number of patients has been enrolled: Across 3 sites, 4 patients were tested twice by the same operator, with each operator testing only one patient.
Prior to training, absolute intra-operator differences in log retinal area between screening and pre-treatment visits ranged from 0.02 to 0.68 (mean±SD=0.24±0.21), inter-operator differences ranged from 0.01 to 0.80 (mean±SD=0.32±0.30). Under a multilevel statistical model accounting for testing two eyes per patient this difference was not significant (p=0.64). Following training, absolute intra-operator differences in log retinal area between screening and pre-treatment visits ranged from 0.03 to 0.40 (mean±SD=0.12±0.13). This improvement, while not statistically significantly different from either inter- (p=0.13), intra- (p=0.14) or combined inter-/intra-operator (p=0.11) results, strongly suggests a positive effect of operator training. Moreover, these findings for trained operators are consistent with the recently 20% coefficient of repeatability for GVF-derived retinal area for a single operator (Bittner et al., IOVS 52:8042-6).
These findings are preliminary, due to the small sample size of this orphan disease pilot trial. Nonetheless they confirm that both operator consistency and rigorous training have a significant beneficial effect on repeatability of the GVF. We are continuing to collect enrollment data using trained operators and expect the improvement to become significant as the trial continues.
Clinical Trial: :
This PDF is available to Subscribers Only