Combining visual acuity with refraction reduces overestimation of myopia prevalence in school screenings: an age-stratified analysis

Ningfeng Li^1,2Yu Jiang^1,2Xu Zhang^1,2Wenzhi Huang^1,2Junyan Zhang³Bozheng Zhang⁴Zongyin Gao^1,2Yunxia Leng^1,2*

• ¹Ophthalmology, Guangzhou First People's Hospital, School of Medicine, South China University of Technology, Guangzhou, China
• ²Ophthalmology, Guangzhou First People’s Hospital, Guangzhou Medical University, Guangzhou, China
• ³Clinical Epidemiology and Evidence-based Medicine, Shanxi Bethune Hospital, Shanxi Academy of Medical Sciences, Taiyuan, China
• ⁴Emory University College of Arts and Sciences, Atlanta, United States

Background: School myopia screening commonly employs non-cycloplegic spherical equivalent refraction (SER ≤-0.50 D) for its practicality. However, this SER-only approach likely overestimates the prevalence of visually significant myopia, as it cannot distinguish true myopia from accommodative pseudomyopia, especially in younger children. We quantified the disparity between this SER-only criterion and a combined criterion integrating uncorrected visual acuity (UCVA), the SER + UCVA criterion, and examined its variation across age groups. Methods: This serial cross-sectional study (2018-2021) included 20,750 students aged 5-18 years from Southern China. Myopia was defined using two criteria based on non-cycloplegic measurements at least one eye: 1) SER-only: SER ≤-0.50 D; and 2) SER+UCVA: SER ≤-0.50 D plus age-impaired UCVA (> 0.20 logMAR at age five; > 0.00 logMAR at ages ≥six) in the same eye. Age-stratified prevalence estimates were compared. Results: The SER + UCVA criterion yielded consistently lower prevalence estimates than the SER-only criterion (e.g., 48.89% vs. 59.48% in 2021), corresponding to a 15-21% annual relative overestimation. Crucially, this overestimation demonstrated a strong age gradient. It was most severe in young children, with the SER-only myopia prevalence nearly double the SER + UCVA myopia prevalence at ages 5-6 years (relative difference >50%), and progressively narrowed to approximately 10-15% in adolescents (14-18 years). Notably, a significant acceleration in SER + UCVA myopia prevalence was observed in children aged 7-10 years between 2019 and 2020. Furthermore, from age 10 onwards, SER + UCVA myopia prevalence was significantly higher in females than males (all P < 0.001). Conclusion: Sole reliance on non-cycloplegic SER substantially overestimates the burden of visually significant myopia, especially in younger children. Incorporating UCVA provides a more accurate and functionally relevant metric for public health screening. We advocate for adopting age-specific SER + UCVA criteria in school-based screenings to optimize referral efficiency, with the greatest benefit expected for younger populations where specificity gains are maximal. Incorporating UCVA provides a more accurate and functionally relevant metric for public health screening.