Accuracy of self-refraction using a portable device based on badal optics, iREF, in children in southern China: A cross-sectional study

OBJECTIVE: To compare the refractive and visual accuracy between a portable device based on Badal optical system (iREF), noncycloplegic autorefraction, and cycloplegic subjective refraction. METHODS: This cross-sectional study enrolled school-children aged 8-18 years in southern China. The operation procedure of self-refraction (SR) using iREF was taught by one researcher, and then the participants used iREF to test the spherical equivalent refraction (SE_SR) by themselves. Then, non-cycloplegic autorefraction (NCAR) was conducted and spherical equivalent (SE_NCAR) was calculated by another researcher. The third researcher, blind to the procedure and results above, performed subjective cycloplegic refraction (CR) and calculated SE_CR. RESULTS: We enrolled 209 participants, with a mean age of 12.13 ± 2.14 years old (males accounting 50.72 %) and a mean SE of -2.91 ± 1.94 D. Bland-Altman plots showed 10 (4.8 %) of the right eye fell outside the 95 % consistency limit of SE_CR and SE_SR. The SE of the right eye by NCAR and SR was -3.03 ± 1.93D and -2.94 ± 1.89D, respectively. The error of SR (-0.03 ± 0.36D) was significantly lower than that of NACR (-0.12 ± 0.31D) (P < 0.001). The absolute error of SR was less than 0.50D in 185 cases (88.5 %), which was not significantly different from that of NCAR (190 cases, 90.9 %) (P = 0.487). High myopia (SE < -6.00 D) and high astigmatism (over 2.00 D) had more obvious errors than other groups. CONCLUSION: There is high consistency of the spherical equivalent measured by iREF self-refraction and cycloplegic refraction, while the accuracy diminishes in patients with high refractive errors.