TABLE 5

RCTs and Controlled Observational Studies of Preschool Vision Screening

Study (Year); DesignNo. of Treatment and Control Subjects (No. Approached, No. Eligible, No. Enrolled, No. Lost to Follow-up)Subject Age, Gender, and DiagnosisCountrySettingScreening InterventionResultsQuality Rating
Williams et al20,21 (2002 and 2003); RCTNumber approached and eligible not reported: 3490 enrolled (2029 intensive screening, 1490 1-time screening); 1929 analyzed at 7.5 yAge: cohort initially tested at age 8–37 mo and followed to age 7.5 y; female: 48% (in those who attended final outcome assessment); diagnosis: baseline amblyopia or amblyogenic risk factors not reportedUnited KingdomHospital eye services clinicScreening at 8, 12, 18, 25, 31, and 37 mo: cover testing, Cardiff cards at 8 and 12 mo, Cardiff and Kays pictures test at 18, 25, and 31 mo, Kays picture test and HOTV test at 37 mo, noncycloplegic autorefraction (performed at all visits, but only used for referral at 37 mo)
Screening at 37 mo: cover testing, Kays picture test and HOTV test, noncycloplegic autorefraction
Screening at 8, 12, 18, 25, 31, and 37 mo vs screening at 37 mo only; amblyopia A at 7.5 y of age: 1.4% (16/1088) vs 2.7% (22/826), RR: 0.55 (95% CI: 0.29–1.04); amblyopia B at 7.5 y of age: 0.6% (69/1088) vs 1.8% (15/876), RR: 0.35 (95% CI: 0.15–0.86)
Residual amblyopia A among children treated with occlusion: 25% (10/40) vs 8% (3/40); OR: 1.56 (95% CI: 0.62–3.92); residual amblyopia B among children treated with occlusion: OR: 4.11 (95% CI: 1.04–16.29)
Mean visual acuity in the worse eye after patching treatment (adjusted for confounding variables): 0.15 (95% CI: 0.083–0.22) vs 0.26 (0.17–0.35), P < .001
Amblyopia A: interocular difference in acuity ≥0.2 logMAR (2 lines on the chart); amblyopia B: interocular difference in acuity ≥0.3 logMAR
Fair
Williams et al21 (2003); prospective cohort study8042 evaluated for inclusion; 1917 excluded because of inclusion in quasi-randomized trial and 44 because of developmental delay or organic eye disease; 6081 included (1516 offered screening at 37 mo, 4565 not offered screening at 37 mo); loss to follow-up not describedAge: cohort tested at 7.5 y; were offered or not offered screening at 37 mo; female: 49%; diagnosis: baseline amblyopia or amblyogenic risk factors not reportedUnited KingdomHospital eye services clinicScreening at 37 mo: Kay's pictures or Sheridan Gardiner singles visual acuity test, cover test, and 20 D prism or test of stereopsis (or both)
No screening at 37 mo
Offered and received screening at 37 mo vs school-entry screening; amblyopia A at 7.5 y of age: 1.1% (11/1019) vs 2.0% (100/5062), adjusted OR: 0.63 (95% CI: 0.32–1.23); amblyopia B at 7.5 y of age: 0.7% (7/1019) vs 1.3% (65/5062), adjusted OR: 0.72 (95% CI: 0.32–1.60); amblyopia C at 7.5 y of age: 1.9% (19/1019) vs 3.4% (171/5062), adjusted OR: 0.65 (0.38–1.10); mean visual acuity in the worse eye after patching treatment (adjusted for confounding variables): 0.14 (95% CI: 0.11–0.18; n = 25) vs 0.22 (95% CI: 0.20–0.23; n = 166), P < .0001
Amblyopia A: interocular difference in acuity ≥0.2 logMAR (2 lines on the chart); amblyopia B: visual acuity in amblyopic eye 0.3 logMAR or worse (6/12 or worse); amblyopia C: visual acuity in amblyopic eye 0.18 logMAR or worse (6/9 or worse)
Fair
Eibschitz-Tsimhoni et al22 (2000); retrospective cohort study988 examined in “screening” city, 808 had attended screening at 1–2.5 y of age and included in analyses; 782 children examined in “nonscreening” city; loss to follow-up not describedAge: 8 y; gender not reported; diagnosis: 1% vs 2.6 amblyopiaIsraelPreschool screeningOphthalmologic examination by an ophthalmologist or orthoptist, including Hirschberg corneal reflex text, monocular fixation and following test, ductions and versions examination, cover-uncover test, alternative cover test, and retinoscopy without cycloplegiaScreening vs no screening at 1–2.5 y
Amblyopia at 8 y of age: 1.0% (8/808) vs 2.6% (20/782); RR: 0.39 (95% CI: 0.17–0.87); amblyopia with visual acuity worse than 20/60 at 8 y of age: 0.1% (1/808) vs 1.7% (13/782); RR: 0.07 (95% CI: 0.01–0.57)
Poor
Feldman et al23 (1980); retrospective cohort studyNumber approached and eligible not reported; 1508 enrolled (745 screening 6–12 mo before school entry and 763 no screening before school entry); loss to follow-up not describedAge: mean 6 y; gender not reported; diagnosis: 13% had at least mild (visual acuity 20/40 or worse) best-corrected vision impairmentCanadaPreschool and school screeningIlliterate E visual acuity test, administered by school nurseRR for at least mild vision impairment (visual acuity 20/40 or worse) upon school entry, screened 6–12 mo previously vs not screened: 10% (78/763) vs 15% (112/745); RR: 0.68 (95% CI: 0.52–0.89)Poor
Kohler et al24 (1978); retrospective cohort studyNumber approached and eligible not reported; 2178 enrolled (619 screened at 4 y of age and 1519 not screened); loss to follow-up not describedAge: 7 y; gender not reported; diagnosis: 49% had vision disorders classified as requiring treatment, functional amblyopia, or strabismusSwedenPreschool and school screeningLinear E-chart, administered by school nurseRR for newly diagnosed vision disorder, amblyopia, or strabismus at 7 y of age, not screened at age 4 vs screened: 0.7% (11/1519) vs 5% (29/619); RR: 0.15 (95% CI: 0.08–0.31)Poor