PURPOSE: To compare whether free spectacles or only a prescription for spectacles influences wearing rates among Tanzanian students with un/undercorrected refractive error (RE). DESIGN: Cluster randomised trial. SETTING: 37 secondary schools in Dar es Salaam, Tanzania. PARTICIPANTS: Distance visual acuity was measured in 6,904 year-1 students (90.2% response rate; median age 14 years; range 11-25 years) using a Snellen E-chart. 135 had RE requiring correction. INTERVENTIONS: Schools were randomly allocated to free spectacles (arm A) or prescription only (arm B). PRIMARY OUTCOME: Spectacle use at 3 months. RESULTS: The prevalence of un/undercorrected RE was 1.8% (95% CI: 1.5 to 2.2%). At 3 months, 27/58 (47%) students in arm A were wearing spectacles or had them at school compared with 13/50 (26%) in arm B (adjusted OR 2.4, 95% CI 1.0 to 6.7). Free spectacles and myopia were independently associated with spectacle use. CONCLUSIONS: The low prevalence of un/undercorrected RE and poor uptake of spectacles, even when provided free, raises doubts about the value of vision-screening programmes in Tanzanian secondary schools. Policy decisions on school vision screening in middle- and low-income countries should take account of the cost-effectiveness as well as competing demands for scarce resources.
RCT Entities:
PURPOSE: To compare whether free spectacles or only a prescription for spectacles influences wearing rates among Tanzanian students with un/undercorrected refractive error (RE). DESIGN: Cluster randomised trial. SETTING: 37 secondary schools in Dar es Salaam, Tanzania. PARTICIPANTS: Distance visual acuity was measured in 6,904 year-1 students (90.2% response rate; median age 14 years; range 11-25 years) using a Snellen E-chart. 135 had RE requiring correction. INTERVENTIONS: Schools were randomly allocated to free spectacles (arm A) or prescription only (arm B). PRIMARY OUTCOME: Spectacle use at 3 months. RESULTS: The prevalence of un/undercorrected RE was 1.8% (95% CI: 1.5 to 2.2%). At 3 months, 27/58 (47%) students in arm A were wearing spectacles or had them at school compared with 13/50 (26%) in arm B (adjusted OR 2.4, 95% CI 1.0 to 6.7). Free spectacles and myopia were independently associated with spectacle use. CONCLUSIONS: The low prevalence of un/undercorrected RE and poor uptake of spectacles, even when provided free, raises doubts about the value of vision-screening programmes in Tanzanian secondary schools. Policy decisions on school vision screening in middle- and low-income countries should take account of the cost-effectiveness as well as competing demands for scarce resources.