Impact of ophthalmic clinical service use in mitigating myopia onset and progression in preschool children: a retrospective cohort study

Abstract Background Although school screenings identify children with vision problems and issue referrals for medical treatment at an ophthalmic hospital, the effectiveness of this approach remains unverified. Objective To investigate the impact of ophthalmic clinical services on the onset and progression of myopia in preschool children identified with vision impairment. Methods Using data from the Shanghai Child and Adolescent Large-scale Eye Study (SCALE), this retrospective cohort study evaluated the visual development of children from three districts—Jing’an, Minhang, and Pudong—which are representative of geographic diversity and economic disparity in Shanghai’s 17 districts. Initially, in 2015, the study encompassed 14,572 children aged 4–6 years, of whom 5,917 needed a referral. Our cohort consisted of 5,511 children who had two or more vision screenings and complete personal information over the follow-up period from January 2015 to December 2020. We divided these children into two groups based on their initial spherical equivalent (SE): a High-risk group (SE > -0.5 D) and a Myopia group (SE ≤ -0.5 D). Within each of these groups, we further categorized children into Never, Tardily, and Timely groups based on their referral compliance to compare the differences in the occurrence and progression of myopia. Cox proportional models were applied to estimate hazard ratios (HRs) for myopia incidence per person-years of follow-up in High-risk group. Generalized additive models(GAM) was used to calculating the progression for annual spherical equivalent changes in all children. Results Of the 5,511 preschool children (mean age, 5.25 years; 52.24% male) who received a referral recommendation, 1,327 (24.08%) sought clinical services at an ophthalmic hospital. After six years of follow-up, 65.53% of children developed myopia. The six-year cumulative incidence of myopia in the Never, Tardily, and Timely groups was 64.76%, 69.31%, and 57.14%, respectively. These percentages corresponded to hazard ratios (HRs) of 1.31 (95% CI, 1.10–1.55) for the Tardily group and 0.55 (95% CI, 0.33–0.93) for the Timely group, compared with the Never group. The HRs were adjusted for age, sex, and SE at study entry. Interestingly, the Timely group showed significantly less SE progression than the other groups (P