TABLE 9

Summary of Evidence According to Key Question

No. of Studies: Overall Quality RatingsLimitationsConsistencyPrimary Care ApplicabilitySummary of Findings
KQ1: Is vision screening in children aged 1–5 y associated with improved health outcomes?
  • Screening vs no screening: 4 cohort studies

  • Intensive periodic vs 1-time screening: 1 RCT

  • Overall quality rating: fair-poor

No study evaluated school performance or other functional outcomes besides vision outcomes
3 of the 4 cohort studies were retrospective and had important methodologic shortcomings; the 1 prospective cohort study compared 1-time screening with no screening
Not applicable (not enough studies addressing the same question to judge consistency)HighNo randomized trial evaluated outcomes of preschool vision screening compared to no screening. One large, fair-quality randomized trial nested within a population-based cohort study found intensive, periodic orthoptist screening from 8 through 37 mo of age associated with reduced likelihood of amblyopia at 7.5 y of age compared with 1-time orthoptist screening at 37 mo of age by ∼1%, but the difference was only statistically significant for 1 of 2 definitions of amblyopia. A large prospective cohort study from this population found 1-time orthoptist screening at 37 mo of age associated with no significant difference in risk of amblyopia at 7.5 y compared with school-entry screening when using any of 3 prestated definitions for amblyopia. Three retrospective cohort studies found preschool screening associated with improved school-aged vision outcomes compared with no screening.
KQ1a: Does effectiveness of vision screening in children aged 1–5 y vary in different age groups?
  • Earlier vs later screening: 1 RCT, 1 cohort study

  • Overall quality rating: poor

In the RCT, it was not possible to determine whether differences in outcomes should be attributed to the earlier age at which screening was started or to the increased frequency of screening that also took place; in the retrospective cohort study, estimates were imprecise and based on a very small sample of children screenedNot applicableHighNo randomized trial directly compared outcomes of preschool vision screening in different age groups. In 1 randomized trial, screening was initiated earlier in 1 group (8 mo of age) compared with the control group (37 mo of age), but the earlier group also received periodic screening. One poor-quality retrospective cohort study found no difference between screening at 2–4 y of age vs screening before 2 y in risk of at least mild vision impairment.
    
KQ2: What is the accuracy and reliability of risk-factor assessment for identifying children aged 1–5 y at increased risk for vision impairment?
  • No studies

No study evaluated the accuracy or reliability of risk-factor assessment in preschool vision screening, and no study evaluated outcomes of targeted vs universal preschool vision screening.
KQ3: What is the accuracy of screening tests for vision impairment in children aged 1–5 y?
  • 31 diagnostic accuracy studies

  • Overall quality rating: good

Estimates of the diagnostic accuracy of different types of screening tests as well as specific screening tests within the different categories varied substantially across studies, which makes it difficult to judge comparative diagnostic utility with certaintySome inconsistency in diagnostic accuracy estimatesModerate (mostly specialty or enriched populations with high prevalence)Thirty-one studies evaluated the diagnostic accuracy of various preschool vision-screening tests. Four studies evaluated visual acuity tests (Lea symbols and HOTV tests), 3 evaluated stereoacuity tests (random dot E stereogram and Stereo Smile II), 1 evaluated the cover-uncover test, 4 evaluated some combination of clinical examination screening tests, 12 evaluated autorefractors, and 15 evaluated photoscreeners. Diagnostic accuracy estimates for all of these screening tests suggest utility for identification of children at higher risk for amblyogenic risk factors or specific visual conditions, although differences between studies in the populations evaluated, screening tests evaluated, screening thresholds applied, and target conditions sought make it difficult to reach strong conclusions about how they compare with one another. Studies that evaluated combinations of clinical tests (visual acuity, stereoacuity, and ocular alignment) generally showed superior likelihood ratios compared with studies of individual tests. In the largest study to directly compare the diagnostic accuracy of different individual screening tests (the VIP Study), differences in likelihood-ratio estimates between the various tests evaluated were generally small with overlapping 95% CIs.
KQ3a: In children aged 1–5 y, does accuracy of screening tests for vision impairment vary in different age groups?
  • 4 studies

  • Overall quality rating: fair

Limited numbers of studies with some inconsistencySome inconstancyModerate (mostly specialty of enriched populations with high prevalence)Evidence on the comparative accuracy of preschool vision tests in different age groups among children aged 1–5 y is limited. Four studies found no clear differences in the diagnostic accuracy of various screening tests in preschool-aged children stratified according to age. Testability using common visual acuity tests, stereoacuity tests, photoscreening, and autorefractors generally exceeds 80%–90% in children ≥3 y of age, and there are small increases in testability through 5 y of age. Four studies found substantially lower testability with the random dot E stereotest, Lea symbols visual acuity testing, and the SureSight autorefractor in preschool-aged children 1–3 y of age, compared with those 3–5 y of age. One large study of statewide screening with the MTI photoscreener found that testability was 94% at 1 y of age.
KQ4: What are the harms of vision screening in children aged 1–5 y?
  • Psychosocial: 1 large cohort study, poor quality

  • False-positives: 7 studies

  • Overall quality rating: poor

Sparse evidence, except for positive predictive valuesNot applicable (not enough studies addressing the same question to judge consistency)HighEvidence on harms of preschool vision screening is limited. Although preschool vision screening is associated with potential psychosocial harms related to the treatments, 1 large cohort study found a 50% reduction in odds of being bullied at the age of 7.5 y among children offered screening compared with those who were not offered screening.
In populations with a prevalence of visual conditions of <10%, 6 of 7 studies reported false-positive rates of >70%. One large study of a statewide preschool photoscreening program found that 20% of children with positive screen results who did not meet criteria for amblyopia or amblyogenic risk factors (false-positives) were prescribed glasses. No study evaluated effects of unnecessary corrective lenses or treatment for amblyopia on long-term vision or functional outcomes.
KQ5: What is the effectiveness of treatment for vision impairment in children aged 1–5 y?
  • Treatment vs no treatment: 1 RCT

  • Patching vs no treatment (>85% received eyeglasses): 2 RCTs

  • Comparisons of treatment: 5 RCTs

  • Overall quality rating: fair

All trials evaluated older (≥3-y-old) preschool-aged children
No trial evaluated effects of treatment compared with no treatment on school performance or other measures of function besides vision outcomes
ConsistentHighIn children with unilateral refractive errors, 1 good-quality trial found patching plus eyeglasses and eyeglasses alone more effective than no treatment by an average of ∼1 line on the Snellen eye chart after 1 y. Effects were larger (1–2 lines of visual acuity improvement) in the subgroup of children with worse baseline visual impairment. One fair-quality and 1 good-quality trial found that patching resulted in a statistically significant but small (<1 line on the Snellen eye chart) average improvement in visual acuity after 5–12 wk of follow-up in children with amblyopia who were pretreated with eyeglasses if needed.
Five fair- or good-quality trials found no differences in visual acuity improvement in the amblyopic eye between shorter and longer daily patching regimens (2 trials), different atropine regimens (2 trials), or between patching and atropine (1 trial). Three trials found no interaction between age and amblyopia treatment effects among preschoolers 3–7 y old, and 1 trial found that delaying treatment for 1 y was associated with similar outcomes compared with immediate treatment in children 3–5 y old. One other trial found younger (3-y-old) preschoolers required fewer hours per day of patching to experience optimal improvements in visual acuity compared with older preschool-aged children (4–8 y old).
KQ6: What are the harms of treatment for children aged 1–5 y at increased risk for vision impairment or for vision disorders?
  • Nonamblyopic eye visual acuity loss: 5 RCTs

  • Overall quality rating for these 5 studies: fair

  • Adverse psychosocial effects: 2 RCTs

  • Overall quality rating for these 2 studies: poor

Sparse evidence on adverse psychosocial effects or effects of compliance on clinical outcomesConsistentHighAlthough 1 short-term (5-wk) trial found no increased risk of decreased nonamblyopic eye visual acuity associated with patching vs no patching, 1 trial found patching associated with increased risk of ≥2 lines of visual acuity loss compared with atropine (9% vs 1.4%; P <.001), and 1 trial found atropine plus a plano lens associated with increased risk of ≥1 line of visual acuity loss compared with atropine alone (17% vs 4%; P = .005). In both trials, nonamblyopic eye visual acuity subsequently returned to baseline in almost all children. Two other trials found no difference in risk of nonamblyopic eye visual acuity loss in direct comparisons of different patching or atropine regimens.
Evidence on adverse psychosocial effects of amblyopia treatments is limited. One fair-quality follow-up study from a randomized trial found that children were more upset by patching plus eyeglasses compared with eyeglasses alone, and 1 good-quality trial found patching associated with worse emotional well-being compared with atropine.