Purchase this article with an account.
Paul C Knox, Anna O'Connor, Ashli F Milling; Effects of age and blur on, and test-retest variability of, a handheld radial shape deformation test. Invest. Ophthalmol. Vis. Sci. 2014;55(13):5605. doi: https://doi.org/.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
A new radial shape deformation test, presented on an Apple iPod Touch (handheld Radial Shape Deformation test; hRSD), has been reported to be a cheap, effective means of monitoring macular disease (Wang et al 2013, IOVS 54:5497). We investigated the performance of the hRSD test in healthy participants.
Resolution acuity at near and distance and contrast sensitivity were measured in 84 healthy participants (52 female) aged 16-80 years (mean: 44years±SD16years) with normal or corrected to normal visual acuity (Median: 0.00, IQR: 0.20LogMAR, near and distance). The hRSD test was used monocularly in each participant, twice under normal viewing conditions (to measure intra-sessional variability), and once with induced optical blur equivalent to three LogMAR lines. 18 participants were tested on a third occasion (median 72 days between sessions) to assess longer term variability.
Mean thresholds for detecting global radial shape deformation, using the hRSD test, for participants aged 16-54 years (-0.74±0.13 LogMAR) and 55 years and over (-0.69±0.13 LogMAR) was statistically significantly different (t = 2.5; p=0.02). A significant difference was also found between normal and blurred conditions (paired samples t test; Normal: -0.70±0.23 LogMAR; Blurred: -0.62±0.25 LogMAR; t=-6.4; p<0.0001) with a mean difference of 0.09 LogMAR. For the intrasessional test/retest data, Bland-Altman 95% limits of agreement were -0.25 to 0.23 LogMAR; mean difference (bias) was -0.01. For the 18 participants (36 eyes) who were tested on a third occasion, there was no statistically significant difference between tests 1 (-0.80±0.15 LogMAR) and 3 (-0.79±0.13LogMAR; t = 1.05; p= 0.30) with Bland-Altman 95% limits of agreement of -0.30 to 0.24LogMAR .
Measured thresholds for the hRSD test were slightly higher (ie worse) compared to published values (Wang et al 2009 OVS 86:695), but followed the same pattern, consistent with stable visual performance to the fifth decade with a slight increase thereafter in the absence of manifest pathology. They remained well below the cut-off value of -0.37LogMAR suggested as being indicative of disease. Test-retest results showed little variability and bias, and the effect of blur, while statistically significant, was of little clinical significance. We are now investigating the hRSD test as a potential screening test for detecting new neovascular AMD.
This PDF is available to Subscribers Only