Abstract
Purpose : The rising trend in myopia prevalence has been a recent global concern. Timely detection is important for early intervention. We have recently demonstrated that smartphone photorefraction is an affordable promising screening tool for myopia. As with all photosceener technology, there could be variance in accuracy across different age groups and refractive error ranges. This study aimed to explore the performance of smartphone photorefraction across different refractive errors in children and adults.
Methods : Participants (N = 471) between 7 to 42 years old (Group A: N = 238; mean age = 10.4 ± 1.4 years and Group B: N = 233; mean age = 21.5 ± 6.5 years) were recruited. Smartphone photorefraction measurements taken at 1m were compared with open-field autorefraction at 3m and 1m testing distance. Refractive errors were classified into 4 categories (low: –0.75 D to –2.00 D, intermediate: < –2.00 D to –4.00 D, moderate: < –4.00 D to –6.00 D, and high: < –6.00 D) for comparison.
Results : The refractive error range were similar between Groups A and B (range; mean ± SD: -0.75 to -12.50 D; -3.94 ± 1.67 D versus -0.75 to -12.70 D; -4.10 ± 2.32D). The effects of age and refractive error range on accommodation and photorefraction were evaluated using two-way ANOVA. Our results showed no significant difference in accommodation between the two age groups across all four refractive error categories (p > 0.05). Smartphone photorefraction achieved comparable accuracy detecting myopia within the low to moderate ranges for both groups [Mean Absolute Error (MAE) of Groups A versus B: MAELow = 0.74 ± 0.59 D versus 0.82 ± 0.50 D, p = 0.17; MAEIntermediate = 0.50 ± 0.44 D versus 0.59 ± 0.56 D, p = 0.02; MAEModerate = 0.98 ± 0.70 D versus 0.99 ± 0.68 D, p = 0.90]. The variation was higher at high myopia range (MAEHigh = 1.33 ± 0.96 D versus 1.83 ± 1.21 D, p < 0.001). Sensitivity and specificity of detecting visually significant myopia of < -2.00 D in Groups A and B were 96%/70% and 96%/83% respectively.
Conclusions : In line with AAPOS guidelines, the smartphone photorefraction demonstrated reliable results comparable to autorefraction in detecting low to moderate magnitudes of myopia. The difference in accommodation between these two groups was not significant for myopia. Our results demonstrates its potential as a useful myopia screening tool across a broad age and refractive error range.
Methods : Participants (N = 471) between 7 to 42 years old (Group A: N = 238; mean age = 10.4 ± 1.4 years and Group B: N = 233; mean age = 21.5 ± 6.5 years) were recruited. Smartphone photorefraction measurements taken at 1m were compared with open-field autorefraction at 3m and 1m testing distance. Refractive errors were classified into 4 categories (low: –0.75 D to –2.00 D, intermediate: < –2.00 D to –4.00 D, moderate: < –4.00 D to –6.00 D, and high: < –6.00 D) for comparison.
Results : The refractive error range were similar between Groups A and B (range; mean ± SD: -0.75 to -12.50 D; -3.94 ± 1.67 D versus -0.75 to -12.70 D; -4.10 ± 2.32D). The effects of age and refractive error range on accommodation and photorefraction were evaluated using two-way ANOVA. Our results showed no significant difference in accommodation between the two age groups across all four refractive error categories (p > 0.05). Smartphone photorefraction achieved comparable accuracy detecting myopia within the low to moderate ranges for both groups [Mean Absolute Error (MAE) of Groups A versus B: MAELow = 0.74 ± 0.59 D versus 0.82 ± 0.50 D, p = 0.17; MAEIntermediate = 0.50 ± 0.44 D versus 0.59 ± 0.56 D, p = 0.02; MAEModerate = 0.98 ± 0.70 D versus 0.99 ± 0.68 D, p = 0.90]. The variation was higher at high myopia range (MAEHigh = 1.33 ± 0.96 D versus 1.83 ± 1.21 D, p < 0.001). Sensitivity and specificity of detecting visually significant myopia of < -2.00 D in Groups A and B were 96%/70% and 96%/83% respectively.
Conclusions : In line with AAPOS guidelines, the smartphone photorefraction demonstrated reliable results comparable to autorefraction in detecting low to moderate magnitudes of myopia. The difference in accommodation between these two groups was not significant for myopia. Our results demonstrates its potential as a useful myopia screening tool across a broad age and refractive error range.
| Original language | English |
|---|---|
| Title of host publication | Investigative Ophthalmology & Visual Science |
| Publication status | Published - 1 Jun 2024 |
| Event | ARVO Annual Meeting 2024 - Seattle, United States Duration: 5 May 2024 → 9 May 2024 |
Conference
| Conference | ARVO Annual Meeting 2024 |
|---|---|
| Country/Territory | United States |
| City | Seattle |
| Period | 5/05/24 → 9/05/24 |