Abstract
Purpose: :
The full-field stimulus threshold, a rod mediated psychophysical measure, has been shown useful in low-vision patients without steady fixation (Roman, Jacobson et al, Physiol Meas. 2007). We established the D-FST, a commercially available system, as a useful tool for following low-vision patients and demonstrated that it is feasible even in young patients (Klein, Birch, Doc. Ophthalmol. 2009). Here, we explored the D-FST within-visit and between-visit variability in an older population of ARMD patients.
Methods: :
Data was collected from the left eye of a sub-set of 27 patients (age 55-88, average 76) out of 56 patients enrolled in a randomized multi-center dose-escalation trial for ARMD. The sub-set was determined by the availability of the D-FST at the test sites. Patients were tested at baseline, repeatedly during 3 months of drug exposure and at an exit visit after a 1-2 week wash-out period. D-FST testing was performed three times during each visit after 30 min. of dark-adaptation as previously described (Klein, Birch, Doc.Ophthalmol. 2009). Data from baseline and exit visit 4 months later were used for repeatability measures. ANOVA testing was performed to establish repeatability of within-visit testing. Bland-Altman tests were performed, comparing the averages between visits to the difference between the visits.
Results: :
Thresholds ranged from -7.6 log cd/m2 to -1.5 log cd/m2 across all patients and all visits. No significant difference was found for within-visit testing (p=0.84, intraclass correlation coefficient = 0.89). Between-visit repeatability showed a bias of 0.1 log cd/m2 toward the exit visit (better sensitivity) and a confidence interval of +/- 0.9 log cd/m2.
Conclusions: :
Within-visit repeatability of D-FST testing in ARMD patients is high, demonstrating the test can be used with an older population. The slight bias towards the second visit could be a learning effect, but a comparison to age matched normal patients will be performed to rule out a possible treatment effect of the study drug and to determine whether ARMD affects D-FST sensitivity. The 0.9 log unit between-visit confidence interval should be considered within the context of the 9 log unit range of the D-FST and the 6 log unit spread of the dataset. It suggests, however, that the test will only detect large changes in sensitivity.
Clinical Trial: :
http://www.clinicaltrials.gov NCT01002950
Keywords: age-related macular degeneration • clinical research methodology • photoreceptors: visual performance