Vision Screening in Children Aged 6 Months to 5 Years: Evidence Report and Systematic Review for the US Preventive Services Task Force

Importance  Preschool vision screening could allow detection and treatment of vision abnormalities during a critical developmental stage, preserving function and quality of life.

Objective  To review the evidence on screening for and treatment of amblyopia, its risk factors, and refractive error in children aged 6 months to 5 years to inform the US Preventive Services Task Force.

Data Sources  MEDLINE, Cochrane Library, CINAHL, and trial registries through June 2016; references; and experts, with surveillance of the literature through June 7, 2017.

Study Selection  English-language randomized clinical trials (RCTs) or prospective cohort studies that evaluated screening, studies evaluating test accuracy, RCTs of treatment vs inactive controls, and cohort studies or case-control studies assessing harms.

Data Extraction and Synthesis  Dual review of abstracts, full-text articles, and study quality; qualitative synthesis of findings. Studies were not quantitatively pooled because of clinical and methodological heterogeneity.

Main Outcomes and Measures  Visual acuity, amblyopia, school performance, functioning, quality of life, test accuracy, testability, and harms.

Results  Forty studies were included (N = 34 709); 34 evaluated test accuracy. No RCTs compared screening with no screening, and no studies evaluated school performance, function, or quality of life. Studies directly assessing earlier or more intensive screening were limited by high attrition. Positive likelihood ratios were between 5 and 10 for amblyopia risk factors or nonamblyogenic refractive error in most studies of test accuracy and were greater than 10 in most studies evaluating combinations of clinical tests. Inability to cooperate may limit use of some tests in children younger than 3 years. Studies with low prevalence (<10%) of vision abnormalities showed high false-positive rates (usually >75%). Among children with amblyopia risk factors (eg, strabismus or anisometropia), patching improved visual acuity of the amblyopic eye by a mean of less than 1 line on a standard chart after 5 to 12 weeks for children pretreated with glasses (2 RCTs, 240 participants); more children treated with patching than with no patching experienced improvement of at least 2 lines (45% vs 21%; P = .003; 1 RCT, 180 participants). Patching plus glasses improved visual acuity by about 1 line after 1 year (0.11 logMAR [95% CI, 0.05-0.17]) for children not pretreated with glasses (1 RCT, 177 participants). Glasses alone improved visual acuity by less than 1 line after 1 year (0.08 logMAR [95% CI, 0.02-0.15], 1 RCT, 177 participants).

Conclusions and Relevance  Studies directly evaluating the effectiveness of screening were limited and do not establish whether vision screening in preschool children is better than no screening. Indirect evidence supports the utility of multiple screening tests for identifying preschool children at higher risk for vision problems and the effectiveness of some treatments for improving visual acuity outcomes.

The most common causes of vision problems in children are amblyopia (a neurodevelopmental disorder that arises from abnormal processing of visual images that leads to a functional reduction of visual acuity) and its associated risk factors (Table 1), nonamblyopic strabismus and nonamblyopic refractive error.^9-12 Recent prevalence estimates of amblyopia, strabismus, and anisometropia (a difference in refractive power between the eyes, in which one foveal image is more blurred than the other) among US children younger than 6 years range from 1% to 6%.^12-16

A variety of vision screening tools are available to evaluate children (Table 2). Left untreated, vision abnormalities in young children could lead to problems at school, bullying, reduced function and quality of life, depression and anxiety, and injuries. Vision abnormalities are often treatable, but efficacy can decrease as children age, and visual loss can become irreversible.^20-24 Untreated amblyopia rarely resolves spontaneously.^25,26

In 2011, the US Preventive Services Task Force (USPSTF) recommended screening children to detect amblyopia or its risk factors at least once between the ages of 3 to 5 years (B recommendation) and concluded that the evidence was insufficient to assess the balance of benefits and harms of vision screening for children younger than 3 years (I statement). To inform an updated recommendation, a review was undertaken of the evidence on benefits and harms of vision screening in children; screening test accuracy; and benefits and harms of treatment of amblyopia, its risk factors, and refractive error.

Detailed methods and additional details of results and analyses are reported in the full evidence report available at https://www.uspreventiveservicestaskforce.org/Page/Document/final-evidence-review/vision-in-children-ages-6-months-to-5-years-screening. Figure 1 shows the analytic framework and key questions (KQs) that guided the review.

PubMed/MEDLINE, the Cochrane Library, and the Cumulative Index to Nursing and Allied Health Literature (CINAHL) were searched for English-language articles published from January 2009 through June 2016. Search strategies are listed in the eMethods in the Supplement. To identify relevant studies published before 2009, all articles included in the 2011 systematic review for the USPSTF were assessed.^28-30 ClinicalTrials.gov and the World Health Organization International Clinical Trials Registry platform were searched for unpublished literature. To supplement electronic searches, the reference lists of pertinent articles, all studies suggested by reviewers, and comments received during public commenting periods were reviewed. Since June 2016, ongoing surveillance was conducted through article alerts and targeted searches of high-impact journals to identify major studies published in the interim that may affect the conclusions or understanding of the evidence and therefore the related USPSTF recommendation. The last surveillance was conducted on June 7, 2017.

Two investigators independently reviewed titles, abstracts, and full-text articles to determine eligibility using prespecified criteria for each KQ (eTable 1 in the Supplement). Disagreements were resolved by discussion. The review included English-language studies of children aged 6 months to 5 years conducted in countries categorized as “very high” on the United Nations Human Development Index. Only studies rated as good or fair quality were included.

For each included study, 1 investigator extracted pertinent information about the populations, tests or treatments, comparators, outcomes, settings, and designs, and a second investigator reviewed for completeness and accuracy. To provide a consistent metric for visual acuity outcome measures, results were converted to logarithm of the minimal angle of resolution (logMAR) measurements using established conversion charts.³¹ Measures of visual acuity are generally reported as Snellen (eg, 20/20, 20/25, 20/30, 20/40, 20/50) or logMAR scales (eg, 0.00, 0.09, 0.18, 0.30, 0.40). Two independent investigators assessed the quality of studies as good, fair, or poor, using predefined criteria developed by the USPSTF and adapted for this topic (eTables 2-5 in the Supplement).³² Disagreements were resolved by discussion. Individual study quality ratings are reported in the Supplement (eTables 2-5).

Findings for each question were summarized in tabular and narrative format. Results of test accuracy studies were not quantitatively pooled because of considerable clinical and methodological heterogeneity (eg, different tests, target condition definitions, populations, and results), and there were too few treatment trials making similar comparisons to attempt quantitative synthesis.

For KQ2, sensitivities, specificities, likelihood ratios (LRs), and predictive values were calculated when articles reported sufficient data. When qualitatively evaluating LRs, positive LRs indicated a minimal (>1-2), small (>2-5), moderate (>5-10), or large (>10) increase in the risk of the condition of interest (eg, amblyopia or its risk factors). Negative LRs indicated a minimal (0.5-<1), small (0.2-<0.5), moderate (0.1-<0.2), or large (<0.1) decrease in the risk of the condition of interest. Likelihood ratios less than 0.1 or greater than 10 provide strong evidence for ruling out (negative LR <0.1) or ruling in (positive LR >10) diagnoses.^33,34

Definitions for what constitutes a minimal clinically important change in visual acuity in young children vary across studies. Recent studies consider a change of 0.2 logMAR (about 2 lines on the Snellen chart) the minimal clinically important change.^35-39 Others consider smaller changes clinically meaningful, generally between 0.10 logMAR (about 1 line on the Snellen chart) and 0.15 logMAR (between 1 and 2 lines).^40-42 Large treatment studies have calculated sample size requirements based on the ability to detect a change of at least 0.1 logMAR between treatment groups.^43-46 When assessing whether improvement in visual acuity represents a clinically meaningful change, practitioners may also consider that visual impairment associated with amblyopia can become permanent and may limit function for the child’s lifetime.^23,47

The overall strength of the body of evidence was assessed for each KQ as high, moderate, low, or insufficient using methods developed for the USPSTF (and the Evidence-based Practice Center program^29,30), based on the overall quality of studies, consistency of results between studies, precision of findings, and risk of reporting bias.

A total of 40 published studies (described in 46 articles^40,44,48-91) with 34 709 participants were included (Figure 2). The main results for each KQ are summarized below.

Key Question 1. Does screening for amblyopia, its risk factors, and refractive error in children aged 6 months to 5 years reduce long-term amblyopia or improve visual acuity, school performance, functioning, and/or quality of life?

Key Question 1a. Does the effectiveness of screening in children aged 6 months to 5 years vary among different age groups?

One randomized clinical trial (RCT)^86,89 and 1 cohort study⁹⁰ enrolling children from the Avon Longitudinal Study of Parents and Children (ALSPAC) project were included (Table 3). The ALSPAC project is a geographically defined birth cohort study enrolling 14 000 children born in southwest England between April 1991 and December 1992.⁸⁶ Both studies reported prevalence of amblyopia at age 7.5 years; neither evaluated school performance, function, or quality of life outcomes. The major methodological shortcoming in both studies was high attrition; around half of children did not have results and were excluded from analyses. In addition, the method of randomization in the (self-described) RCT was inadequate (based on last digit of the mother’s day of birth).^86,89

The RCT (n = 3490) compared intensive orthoptist screening (clinical examination, age-specific visual acuity testing, and cover-uncover testing) before age 3 years (at 8, 12, 18, 25, 31, and 37 months) with 1-time orthoptist screening at age 37 months.^86,89 Baseline data for amblyopia or amblyopia risk factors were not reported. Children in both groups were offered “usual care” in terms of surveillance for visual problems: (1) examination at ages 8 and 18 months by a health visitor (community pediatric nurse), with referrals if a visual problem was suspected⁸⁶; and (2) visual screening at school entry (ages 4-5 years) by a school nurse.⁸⁹ The prevalence of amblyopia at 7.5 years was approximately 1% lower in the intensive screening group than in the control group, but the difference was statistically significant for only 1 of their 2 definitions of amblyopia (Table 3).⁸⁹ Among those who received patching treatment (n = 40 in each group), presence of residual amblyopia at 7.5 years was more likely in the 1-time screening group than in the intensive-screening group, but the difference was statistically significant for only 1 of the 2 amblyopia definitions, and estimates were imprecise; visual acuity at 7.5 years in the worse eye was better in the intensive-screening group than in the 1-time screening group (Table 3).⁸⁹

The prospective cohort study (n = 6081 completers) compared orthoptist screening at age 3 years in 1 health district with no preschool screening in 2 other health districts.⁹⁰ Screening examinations by the orthoptist consisted of a monocular vision test, a cover test, and an assessment of binocularity; failure of any part of the examination resulted in referral for further evaluation. All children in the study area were offered vision screening at school entry (ages 4-5 years).⁹⁰ Among participants who attended the examination at age 7.5 years and were not part of the ALSPAC RCT, no statistically significant differences in amblyopia were apparent between groups based on any of the studies’ 3 definitions of amblyopia (Table 3).⁹⁰

Key Question 2. What is the accuracy and reliability of screening tests for amblyopia, its risk factors, and refractive error in children aged 6 months to 5 years?

Thirty-four fair-quality studies (described in 38 articles) were included (eTables 6-17 in the Supplement).^{48,49,51-62,64-85,88,91} The studies evaluated a variety of test types, including visual acuity tests, stereoacuity tests, ocular alignment tests, autorefractors, photoscreeners, and retinal birefringence scanning. Screening was administered by a variety of personnel across studies (eg, pediatricians, ophthalmologists, nurses, research staff). Sample sizes ranged from 63⁷⁵ to 4040.^70,91

About one-third of the studies included participants younger than 3 years.^{48,56,65,66,71,72,75-77,80,81,85,88,91} The included studies reported accuracy of tests for a variety of target conditions, ranging from very specific (eg, astigmatism) to broad (eg, amblyopia risk factors). The prevalence of target conditions was generally much higher in samples from ophthalmology clinics^{53,56,57,62,66,68,69,72,75,77,80,81,85,88} than in those from primary care, community, Head Start, or school settings.

Findings from the Vision In Preschoolers (VIP) study, the largest study for this KQ, were reported in multiple manuscripts (up to 4040 participants).^{55,60,64,70,78,82,84,91} Phase 1 of the VIP study enrolled 3- to 5-year-olds and compared the accuracy of 11 screening tests.⁷⁸ Phase 2 compared the performance of nurse screeners with that of lay screeners for 4 tests.⁸⁴ Unlike many of the included studies, the VIP study evaluated accuracy for a broad range of conditions, including significant nonamblyogenic refractive error. The applicability of the VIP study may be limited because it did not enroll a representative spectrum of patients (as demonstrated by the high prevalence of target conditions, ranging from 21%-36%), study participants may have experienced fatigue from the number of tests, and testing was conducted by skilled personnel in a controlled environment (in phase 1).

Detailed results of studies evaluating test accuracy are provided in the eResults and eTables 6 through 17 in the Supplement. Six publications evaluated visual acuity tests,^{53,73,74,78,82,84} including 3 from the VIP Study Group.^78,82,84 When screening test cutoffs were set to achieve specificities of 90%, phase 1 of the VIP study found that an abnormal result moderately increased the likelihood of amblyopia, amblyopia risk factors (strabismus, astigmatism, hyperopia, myopia, anisometropia), or significant nonamblyogenic refractive error (positive LR, 6.1 [95% CI, 4.8-7.6]).⁷⁸ A normal result indicated a small decrease in the likelihood (negative LR, 0.43 [95% CI, 0.38-0.50]).

Four fair-quality studies (total, 1854 participants) evaluated a combination of clinical tests, including visual acuity tests, stereoacuity tests, and ocular alignment tests (eTables 6-7 in the Supplement).^52,54,67,80 Three of the 4 found that abnormal results indicated a large increase in the likelihood of amblyopia or its risk factors (positive LRs ranged from 12-17).^52,67,80 The 4 studies found more variability for negative LRs (range, 0.10-0.91).

Sixteen fair-quality studies (16 712 observations) evaluated autorefractors (eTables 6-17 in the Supplement).^{49,51,55,57,58,65,66,70,72-74,77,78,84,88,91} Overall, most studies found moderate positive LRs and small negative LRs, although some found large positive and negative LRs.

Eleven fair-quality studies (12 publications, 6187 observations) evaluated photoscreeners (eTables 6-15 in the Supplement).^{56,59,67-69,73,75-78,81,85} Overall, most studies found moderate positive LRs and small negative LRs, although some found larger or smaller LRs.

Key Question 2a. Does the accuracy or reliability of screening tests for amblyopia, its risk factors, and refractive error vary among different age groups?

Five studies evaluated whether accuracy varies by age (eTable 13 in the Supplement).^{54,66,69,81,82} All 5 evaluated different screening tests and assessed different age stratifications/comparisons. Overall, data were limited and estimates were somewhat imprecise, but studies did not find any clear differences in test accuracy when results were stratified by age.

Many included studies reported testability information, although few reported data stratified by age or for children younger than 3 years. eTable 14 in the Supplement details the proportion unexaminable reported by studies. Overall, testability exceeded 90% in the majority of studies. Few studies reported testability rates less than 80%, but all that did included children younger than 3 years.^48,66,71,77 Some studies demonstrated that testability rates improved somewhat as children age.^{53,66,71,77,78,80} One study (n = 1170) found that testability rates were 10% for a visual acuity test at ages 24 months to younger than 30 months and steadily improved to 80% by ages 36 months to younger than 42 months and to 95% by ages 48 months to younger than 54 months.⁷¹

For autorefractors and photoscreeners, the VIP study found testability rates close to 100% (all participants were 3 years or older).⁷⁸ Two studies from ophthalmology clinics and 1 from a primary care practice reported better testability for older preschool children than for younger ones (eResults in the Supplement).^65,66,77

Key Question 3. What are the harms of screening for amblyopia, its risk factors, and refractive error in children aged 6 months to 5 years?

One controlled study that evaluated potential psychosocial effects was included,⁸⁷ and 16 studies of test accuracy described in KQ2 were used to calculate false-positive rates.^{48,51,52,55,62,65,67,69,75-78,80-82,85} The controlled study used the ALSPAC population-based cohort (n = 4473) to assess bullying by age 8 years.⁸⁷ It prospectively compared children who had been offered state-provided preschool screening for amblyopia (at 37 months) with those who had not. Children were asked whether they had repeatedly (≥4 times a month) been bullied. Among the subgroup of patched children, the study showed a lower likelihood of being bullied for children offered screening than for those not offered early screening (25.7% vs 47.1%, P = .033; adjusted odds ratio, 0.39 [95% CI, 0.16-0.92], adjusted for sex, paternal socioeconomic class, highest level of maternal education, type of housing).

The most frequently assessed potential harms of screening were false-positive findings (which would lead to unnecessary referrals). In general, studies with a lower prevalence (<10%) of vision abnormalities showed much higher false-positive rates (usually >75%), while those with a high prevalence had lower false-positive rates (usually <35%) (eFigure in the Supplement).

Key Question 4a. Does treatment of amblyopia, its risk factors, and refractive error in children ages 6 months to 5 years improve visual acuity?

Key Question 4b. Does treatment of amblyopia, its risk factors, and refractive error in children ages 6 months to 5 years reduce long-term amblyopia or improve school performance, functioning, and/or quality of life?

Three trials were included (Table 4)^40,44,50; all evaluated patching for amblyopia or amblyopic risk factors. Two compared patching with no patching (children were pretreated with eyeglasses if indicated in both groups),^44,50 and 1 compared patching plus eyeglasses vs eyeglasses alone vs no treatment.⁴⁰ One of the patching vs no patching trials included a run-in phase, during which all participants wore updated eyeglass prescriptions until visual acuity in the amblyopic eye stopped improving⁴⁴; another trial treated children with refractive error with 6 weeks of corrective lenses before allocation.⁵⁰ All 3 studies included children based on visual acuity criteria. One of the 3 trials reported enrolling screen-detected children.⁴⁰ Two of the trials reported best corrected visual acuity,^40,44 and 1 measured improvement in visual acuity as a secondary outcome (the trial focused primarily on assessing adherence).⁵⁰

Overall, the trials indicated that treatments for amblyopia or its risk factors resulted in small mean improvements in visual acuity (Table 4). In the study with the run-in phase,⁴⁴ patching improved visual acuity by a mean of 0.7 of a line on a standard visual acuity chart, and more children treated with patching than with no patching had at least 2 lines of improvement in acuity (45% vs 21%).

Two included trials^40,44 examined treatment outcomes for subgroups defined by baseline visual acuity. First, 1 trial (n = 180) assessed subgroups with either moderate (20/40 to 20/100) or severe (20/125 to 20/400) amblyopia at baseline.⁴⁴ Findings for these subgroups were similar to the overall trial results for the primary outcome, visual acuity in the amblyopic eye. Second, the trial that compared patching plus eyeglasses, eyeglasses alone, and no treatment among preschoolers (n = 177) assessed subgroups defined by baseline visual acuity abnormalities.⁴⁰ The authors assessed children with mild (0.18-0.30 logMAR) and moderate or worse (≥0.48 logMAR) refractive error at baseline and examined differences between treatment groups. For children with moderate refractive error at baseline, patching plus eyeglasses resulted in much greater improvement than no treatment at 1 year (0.27 logMAR [95% CI, 0.14-0.39], compared with improvement for all participants of 0.11 logMAR [95% CI, 0.05-0.17]); the difference between eyeglasses alone and no treatment did not reach statistical significance. For children with mild refractive error at baseline, neither treatment was significantly different than no treatment at the end of the trial.

Key Question 5. What are the harms of treatment of amblyopia, its risk factors, and refractive error in children aged 6 months to 5 years?

Three trials (described in 4 articles) were included (Table 4).^40,44,50,63 Overall, the trials provided limited evidence but suggest that patching may have some psychological harms.

One trial comparing patching with no patching (n = 180) found that worsening visual acuity in the nonamblyopic eye was not significantly different between groups at 5 weeks (2.4% vs 6.8%, respectively; P = .28).⁴⁴ Among children with no ocular deviation at baseline (n = 118), 5 patients in the patching group and 3 patients in the no-patching group were noted to have a new small-angle strabismus, and 1 patient in the no-patching group was noted to have a new large-angle strabismus.

The trial comparing patching plus eyeglasses, eyeglasses alone, and no treatment found no statistically significant difference between treatment groups at 1-year follow-up in the proportion of children whose uncorrected visual acuity in the amblyopic eye worsened (change >0.1 logMAR) for those with baseline mild acuity loss (9.7% vs 6.5% vs 13.3%, respectively, P = .28) or for those with baseline moderate acuity loss (15.0% vs 11.1% vs 23.8%, P = .13).⁴⁰

A substudy⁶³ of the trial⁴⁰ that compared patching plus eyeglasses, eyeglasses alone, and no treatment examined the emotional status of children undergoing treatment; in the substudy, 144 of 177 parents of participants completed questionnaires at baseline (all participants), 3 months after beginning treatment (participants in active treatment only), and 2 years after recruitment (all participants). They found no significant differences between treatment groups with regard to being happy, cooperative, or good tempered; teasing; problems at preschool; or in emotional and behavioral problems, but found that children were more upset by patching plus eyeglasses than by eyeglasses alone (85% vs 29% at age 4 years, P = .03; 62% vs 26% at age 5 years, P = .005). Although the study reported some negative effects of glasses or patching for the child (difficulty wearing patch or glasses, upset, coping with treatment) and parent (worry about treatment, upset by treatments, arguments about treatment), it did not report a comparison with the no-treatment group for these outcomes.

One trial (n = 60) comparing no treatment, patching for 3 hours daily, or patching for 6 hours daily reported that no patients experienced an adverse event, such as inverse amblyopia or patch allergy.⁵⁰

The summary of findings is presented in Table 5. No eligible RCTs directly compared screening with no screening. For the overarching question (KQ1), the strength of evidence was graded as low because of unknown consistency (with a single study making each comparison), imprecision, and methodological limitations. One cohort study showed a reduction in harm (ie, less school-aged bullying) among patched children screened in preschool compared with patched children not screened in preschool.⁸⁷ In theory, although both glasses and patching have been reported to increase the risk of being bullied,⁹² preschool screening may allow for treatment before school starts, thus avoiding potential bullying and psychosocial distress.

Harms of preschool vision screening might include unnecessary referrals from false-positive screens, overdiagnosis, and unnecessary treatment. Studies of test accuracy show that screening tests are associated with high false-positive rates among populations with a low prevalence of vision abnormalities. A large (n = 102 508) retrospective study from a statewide photoscreening program found that 19.5% (174/890) of those with false-positive test results were prescribed glasses (ie, unnecessary treatments).⁹³ The study was not eligible for this systematic review because it did not attempt to perform the reference standard in all participants or a random sample of participants.

Regarding test accuracy, estimates for all tests suggest utility for identifying children at higher risk for amblyopia risk factors or other visual conditions. Positive LRs were in the moderate range (>5-10) for most studies, and most studies that evaluated combinations of clinical tests found high (>10) positive LRs. The VIP study, the largest to directly compare multiple tests, generally found similar accuracy across tests. The strength of evidence was graded as low, because of imprecision and methodological limitations of the individual studies. Findings are applicable to a variety of settings and screening personnel.

Accuracy did not clearly differ for preschool children in different age groups. However, unlike studies of photoscreeners, most studies of clinical test accuracy did not enroll children younger than 3 years. Data were relatively limited and estimates were somewhat imprecise, but studies did not find any clear differences in accuracy of tests when results were stratified according to age. Testability may limit the utility of some screening tests, especially clinical tests, in children younger than 3 years. Although relatively few studies assessed changes in testability by age, those that did generally found better testability in children 3 years or older, and some reported low testability rates for visual acuity and stereoacuity tests for those younger than 3 years. In contrast, some data suggest that photoscreeners have high testability rates for children as young as 1 year.⁹⁴

The review found evidence of moderate strength supporting the effectiveness of some treatments for improving visual acuity outcomes, although mean improvements were small. No studies evaluated potential effectiveness of treatments for reducing long-term amblyopia or for improving school performance, functioning, or quality of life, and no eligible studies evaluated atropine or vision therapy. The included trials all enrolled children 3 years or older, and applicability to those younger than 3 years is unclear. The trials varied somewhat in the populations (with amblyopic risk factors and pretreated with glasses or with amblyopic risk factors but not pretreated with glasses) and interventions compared (2 evaluated patching vs no patching; 1 compared patching plus glasses vs glasses alone vs no treatment). The trial that compared patching plus eyeglasses, eyeglasses alone, and no treatment enrolled screen-detected children, demonstrating the applicability of findings to the main population of interest for this review.⁴⁰

Taken together, the treatment trials provide evidence of moderate strength that (1) patching improves visual acuity of the amblyopic eye by a mean of less than 1 line on an eye chart after 5 to 12 weeks compared with no patching for children with amblyopic risk factors pretreated with glasses, (2) patching plus glasses improves visual acuity by about 1 line after 1 year compared with no treatment for children with amblyopic risk factors not pretreated with glasses, and (3) glasses alone improve visual acuity by less than 1 line after 1 year compared with no treatment for children with amblyopic risk factors. The magnitude of improvement for patching plus glasses or glasses alone was greater for those with worse baseline visual acuity. Few of the trials reported binary outcomes that may help determine how many participants achieved a clinically meaningful change, although 1 trial reported that more children treated with patching than with no patching experienced improvement of at least 2 lines.⁴⁴

The review has several limitations. First, for studies of test accuracy conducted in ophthalmology settings, details about the study participants were sometimes limited, making it difficult to determine whether participants had known impaired visual acuity or obvious symptoms of impaired visual acuity. Thus, the review may have included some studies that would not meet eligibility criteria if additional description of the study populations was available. Second, the review did not include comparative effectiveness (ie, head-to-head) studies, such as those comparing atropine with patching. The previous review for the USPSTF described head-to-head trials that compared different patching regimens (eg, 2-hour vs 6-hour patching), different atropine regimens (daily atropine vs weekend atropine), and patching with atropine.^{43,45,46,95-97} It concluded that the trials found no differences in visual acuity improvement in the amblyopic eye between the treatments. Third, studies published in languages other than English and those conducted in countries not categorized as very high on the Human Development Index were excluded.

Studies directly evaluating the effectiveness of screening were limited and do not establish whether vision screening in preschool children is better than no screening. Indirect evidence supports the utility of multiple screening tests for identifying preschool children at higher risk for vision problems and the effectiveness of some treatments for improving visual acuity outcomes.

Corresponding Author: Daniel E. Jonas, MD, MPH, University of North Carolina at Chapel Hill, 5034 Old Clinic Bldg, Chapel Hill, NC 27599 (daniel_jonas@med.unc.edu).

Accepted for Publication: July 13, 2017.

Author Contributions: Dr Jonas had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Jonas, Amick, Wallace, Feltner.

Acquisition, analysis, or interpretation of data: Jonas, Amick, Wallace, Feltner, Vander Schaaf, Brown, Baker.

Drafting of the manuscript: Jonas, Amick, Wallace, Vander Schaaf, Brown, Baker.

Critical revision of the manuscript for important intellectual content: Jonas, Amick, Feltner, Vander Schaaf, Brown.

Statistical analysis: Jonas, Wallace, Feltner, Vander Schaaf, Brown.

Obtained funding: Jonas.

Administrative, technical, or material support: Jonas, Amick, Feltner, Baker.

Supervision: Jonas, Amick, Feltner.

Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported.

Funding/Support: This research was funded under contract HHSA-290-2012-00015-I, Task Order 6, from the Agency for Healthcare Research and Quality (AHRQ), US Department of Health and Human Services, under a contract to support the USPSTF.

Role of the Funder/Sponsor: Investigators worked with USPSTF members and AHRQ staff to develop the scope, analytic framework, and key questions for this review. AHRQ had no role in study selection, quality assessment, or synthesis. AHRQ staff provided project oversight, reviewed the report to ensure that the analysis met methodological standards, and distributed the draft for peer review. Otherwise, AHRQ had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; and preparation, review, or approval of the manuscript findings. The opinions expressed in this document are those of the authors and do not reflect the official position of AHRQ or the US Department of Health and Human Services.

Additional Contributions: We gratefully acknowledge the following individuals for their contributions to this project, including AHRQ staff (Iris Mabry-Hernandez, MD, and Tracy Wolff, MD) and RTI International–University of North Carolina EPC Staff (Carol Woodell, BSPH; Christiane Voisin, MSLS; Sharon Barrell, MA; and Loraine Monroe). The USPSTF members, expert consultants, peer reviewers, and federal partner reviewers did not receive financial compensation for their contributions. Ms Woodell, Ms Voisin, Ms Barrell, and Ms Monroe received compensation for their role in this project.

Additional Information: A draft version of the full evidence report underwent external peer review from 3 content experts (Michael P. Clarke, MB, FRCOphth, Newcastle Upon Tyne; Steven J. Goldstein, MD, Weill Cornell Medical College; Sean P. Donahue, MD, PhD, Vanderbilt University Medical Center) and 2 federal partner reviewers from the National Institute of Child Health and Development and the National Eye Institute. Comments from reviewers were presented to the USPSTF during its deliberation of the evidence and were considered in preparing the final evidence review.

Editorial Disclaimer: This evidence report is presented as a document in support of the accompanying USPSTF Recommendation Statement. It did not undergo additional peer review after submission to JAMA.