Abstract:
To determine the accuracy of self-refraction (SR) in myopic teenagers, we compared visual and refractive outcomes
of self-refracting spectacles (FocusSpecs) with those obtained using cycloplegic subjective refraction (CSR) as a gold
standard.
Methods. A total of 203 eligible schoolchildren (mean [TSD] age, 13.8 [T1.0] years; 59.1% were female) completed an
examination consisting of SR with FocusSpecs adjustable spectacles, visual acuity with the logMAR (logarithm of the
minimum angle of resolution) chart, cycloplegic retinoscopy, and CSR. Examiners were masked to the SR findings.
Wilcoxon signed rank test and paired Student t test were used to compare measures across refraction methods (95%
confidence intervals [CIs]).
Results. The mean (TSD) spherical equivalent refractive error measured by CSR and SR was j1.22 (T0.49) diopters (D) and
j1.66 (T0.73) D, respectively, a statistically significant difference ofj0.44D(p G 0.001, t = 15.517). The greatest proportion
of participants was correctable to visual acuity greater than or equal to 6/7.5 (logMAR 0.1) in the better eye by CSR (99.0%;
95% CI, 96.5 to 99.7%), followed by cycloplegic retinoscopy (94.1%; 95% CI, 90.0 to 96.6%) and SR (85.2%; 95% CI, 79.7
to 89.5%). These proportions differed significantly from each other (p G 0.001, Wilcoxon signed rank test). Myopic inaccuracy
of greater than 0.50 D and greater than or equal to j1.00 D was present in 29 (15.3%) and 16 (8.4%) right eyes,
respectively, with SR. In logistic regression models, failure to achieve visual acuity greater than or equal to 6/7.5 in right eyes
with SR was significantly associated with age (odds ratio, 1.92; 95% CI, 1.12 to 3.28; p = 0.017) and spherical power (odds
ratio, 0.017; 95% CI, 0.005 to 0.056; p G 0.001).
Conclusions. Self-refraction offers acceptable visual and refractive results for young people in a rural setting in Ghana,
although myopic inaccuracy in the more negative direction occurred in some children.
