Impact of Free Glasses and a Teacher Incentive on Children's Use of Eyeglasses - A Cluster-Randomized Controlled Trial

728 of 4,376 screened children randomized by school (358 in 47 intervention schools, 370 in 47 control schools); 693 (95.2%) analysed

Urban migrant children, mean age 10.9, in 94 randomly chosen primary schools in China

Cluster RCT whose outcome is spectacle WEAR, not achievement - included because wear is the binding constraint on every achievement result in this topic. Intervention teachers received a tablet computer if at least 80% of children given glasses were wearing them at unannounced visits. Primary outcome was observed (not self-reported) wear at 6 months, which is the honest measure.

Observed spectacle wear at 6 months was 68.3% in the incentive arm vs 23.9% in control (adjusted OR 11.5, 95% CI 5.91-22.5, P < .001); self-reported 90.6% vs 32.1%. The authors explicitly benchmark against their own earlier trial: observed wear was only 41% among similar-aged children given free glasses WITHOUT teacher incentives. Free provision alone gets roughly 4 in 10 children wearing glasses; free provision plus a teacher accountability incentive gets roughly 7 in 10. Other predictors of wear were baseline spectacle wear, worse uncorrected acuity, and parental spectacle wear.

Low for the treatment contrast (school randomization). Note that parental spectacle wear predicted child wear - a familial correlation that is simultaneously genetic (myopia is highly heritable, so myopic parents have myopic children) and behavioural (normalization of wearing glasses), and the design cannot separate the two.