Abstract
Purpose :
To evaluate validity of the SVOne handheld, smartphone-based autorefractor (AR) compared to a standard tabletop AR (Topcon KR-800S) in identifying refractive errors in a community screening setting.
Methods :
Participants, 50 years of age and older in community eye screenings underwent visual acuity (VA) testing. Those with VA worse than 20/40 in either eye underwent autorefraction with the conventional tabletop AR followed by 2 trials with the handheld AR. Statistical analysis of power vectors [M (spherical equivalent SE), J0, and J45] were conducted using the right eye. Paired t-test conducted between tabletop and handheld AR. Intraclass Correlation Coefficient (ICC) calculated between handheld and tabletop AR. Test-retest Pearson correlation calculated between handheld AR trials. Lastly, modified Bland Altman plots were constructed between tabletop and handheld AR.
Results :
Out of the 47 subjects who underwent successful tabletop AR and agreed to participate in our study, 35 (74%) were successfully autorefracted via the handheld device. On average, subjects were 63.6 (±9.7) years of age, 21 (60.0%) were female, and 33 (94.3%) were African American. Tabletop AR mean M was -0.26D (95% CI: -1.00, 0.48) and the handheld AR mean M was -0.27D (95% CI: -1.02, 0.49) (P>0.05). Mean astigmatism with the tabletop and handheld ARs were +0.91D (95% CI: 0.71, 1.12) and +1.18D (0.91 to 1.44) respectively (p<0.05). Intraclass correlation coefficient between tabletop and handheld AR for M was 0.96 (95% CI: 0.91, 0.98) (p<0.05), for J0 was 0.35 (0.02, 0.61) (p<0.05), and 0.23 (-0.10, 0.52) (p>0.05) for J45. Good correlation was noted between handheld test-retest for M (Pearson correlation (r)=0.96 (p<0.05), but poor test-retest correlation noted for J0 and J45 vectors [r=0.48 (p<0.05), r=0.61 (p<0.05), respectively].
Conclusions :
The SVOne handheld AR provides a similar measurement of M (SE) to the Topcon KR-800S tabletop AR in subjects 50 years of age and older. Repeated measures using the handheld device shows strong correlation between trials for M. However, when considering vectors J0 and J45, the device shows poor repeatability as well as poor correlation with the tabletop AR. In elderly populations, The SVOne handheld AR may be a useful community-screening tool for quickly assessing SE, but will most likely provide less consistent measures of astigmatism than tabletop AR.
This is an abstract that was submitted for the 2017 ARVO Annual Meeting, held in Baltimore, MD, May 7-11, 2017.