Refractive Errors and Amblyopia Among Children Screened by the UCLA Preschool Vision Program in Los Angeles County


To summarize the results of 5 years of vision screening with the University of California, Los Angeles (UCLA) Preschool Vision Program (UPVP).


Retrospective evaluation of a screening program.


The UPVP performed visual acuity and undilated noncycloplegic refractive screening using an autorefractor on 93,097 children between 2012 and 2017. Of these, 79,451 children, who were between 3 and 5 years old, were screened for the first time, and 14,259 were referred for full cycloplegic examination if they met specific refractive criteria for myopia, hyperopia, astigmatism, or anisometropia. UPVP performed 6779 cycloplegic examinations on this population. Data from the right eye only were included in this analysis.


Of the examined population, hyperopia was found in 61% (4018), myopia in 20% (1336), and astigmatism in 93% (6122) of children. Latino children had higher rates of astigmatism and worse visual acuity compared to all other races/ethnicities. An astigmatism cutoff of ≥1.50 diopters (D) in either eye correctly predicted the need for glasses 93% of the time; increasing this cutoff to ≥1.50 D in both eyes increased the positive predictive value to 96%. Refractive amblyopia was found in 780 children (1.0% of the screened population and 11.5% of the examined population), and of these, 211 (27%) were bilaterally amblyopic.


These data represent the largest published sample of vision screening results on preschool-aged children, provide additional insight on the proportion of common refractive errors and their association with race/ethnicity, and can inform screening criteria to more accurately identify children who need intervention to prevent permanent vision loss.

It is estimated that in 2015, the prevalence of visual impairment in children 3 to 5 years old was 1.5%, or 174,600 individuals, and the incidence is projected to increase substantially in the near future. Unlike strabismus, refractive errors, especially unilateral cases, can often go unnoticed in children, as this patient population may be unaware or unable to adequately describe their condition. Vision screening in this population allows for early diagnosis and treatment to prevent the long-term effects of visual impairment. For these reasons, the United States Preventive Services Task Force and the American Academy for Pediatric Ophthalmology and Strabismus recommend that all children between the ages of 3 and 5 years undergo vision screening.

Amblyopia is a reduction of visual acuity that results from a disruption of the normal development of the visual pathway. Refractive amblyopia is the most common, the most difficult to detect, and the most straightforward to correct. The benefit to prompt correction of refractive errors has been reported on previously: children have improved academic performance and quality of life.

Between 2012 and 2017, the University of California, Los Angeles (UCLA) Preschool Vision Program (UPVP) screened over 90,000 preschoolers in the Los Angeles area using a Retinomax autorefractor (Righton, Tokyo, Japan), which has been shown to be sufficiently sensitive and specific for this purpose. The details of the UPVP program have been reported on previously, as have preliminary results from the first year. ,

This comprehensive 5-year summary represents the largest sample size of vision screening in a population of children aged 3-5 years. The data enable a more precise analysis of the distribution of refractive errors and proportion of amblyopia in a vulnerable population. Further, it allows for comparisons to prevalence levels previously published in other large studies such as the Multiethnic Pediatric Eye Disease Study (MEPEDS) and the Baltimore Pediatric Eye Disease Study (BPEDS). Importantly, this represents the largest sample of Latino children, who have been previously reported to have higher rates of astigmatism; the data can expand on this disproportionate proportion.


This retrospective cross-sectional evaluation of a screening program was approved by the UCLA Institutional Review Board and conformed to the requirements of the United States Health Insurance Portability and Accountability Act. Between 2012 and 2017, UPVP performed 93,097 screening procedures among preschool-aged children at subsidized preschools in Los Angeles County. Of these records, 3,305 were excluded for being either younger than 3 or older than 5 years old, and 10,341 were excluded for being repeat visits of the same individual child; for each of these children, only their first visit with UPVP was included in the analysis. The remaining 79,451 records met inclusion criteria.

The UPVP has been described in detail elsewhere. In brief, the visit protocol at each school comprised an initial screening day for all children and a subsequent examination day for any child who failed the screening day. On the screening day, the Retinomax 3 autorefractor (Rmax) was used to screen children for refractive errors. The criteria for failing the initial screening were as follows: sphere ≤−3.25 diopters (D) or ≥ +1.75 D, cylinder ≥ 1.50 D, or an interocular difference of ≥ 1.50 D of sphere or ≥ 1.50 D of cylinder error ( Figure 1 ). (In the first year of the program, the screening failure criteria included spherical anisometropia of ≥ 2.00 D [≥ 1.50 if antimetropic] and cylindrical anisometropia of ≥ 1.0 D; the subsequent change occurred to align our screening criteria with the UCSD Shiley Eye Institute EyeMobile Program.) For those who failed the initial screening, a secondary screening was performed on the examination day using Rmax with the same criteria. If the child failed the secondary screening and consent had been obtained, a comprehensive ophthalmologic examination was performed. For some schools, consent for vision screening was obtained at the time of school enrollment. For others, a separate consent form for vision screening was provided to the parent or legal guardian to return signed. All children for whom an ophthalmic examination was recommended were provided another consent form for their parent or legal guardian to sign.

Figure 1

Criteria during initial noncycloplegic screening with Retinomax autorefractor (Righton, Tokyo, Japan) that would prompt referral for re-screening and possibly cycloplegic examination.

The comprehensive examination consisted of testing with Allen pictures with occlusion glasses at a 10-foot distance, color vision, stereoacuity with the Titmus fly test, ocular motility, ocular alignment performed with cover test, slit-lamp examination, cycloplegic retinoscopy, and indirect ophthalmoscopy. Cycloplegic refraction was performed at least 30 minutes after instillation of cycloplegic eye drops, and the best-corrected visual acuity (BCVA) was assessed. Children who would benefit from refractive correction selected a frame style and were fitted with frames. UPVP then arranged to have the glasses made and delivered to the child within a few weeks. Children with more complex ocular pathology were referred to an ophthalmologist, and all children were recommended to follow up with an eye care professional at a time interval dependent on the child’s diagnosis.

An initial analysis of the data from the first year of the program was published previously; the analyses performed here are similar. Snellen visual acuity (VA) format was converted to logMAR in order to facilitate the calculations. Hyperopia and myopia were calculated based on cycloplegic refraction spherical equivalent (SE). SE was calculated as the sum of the spherical plus half the cylindrical error. On cycloplegic exam, hyperopia was defined as ≥ 0.50 D, myopia as ≤ −0.50 D and emmetropia in between −0.50 D and +0.50 D. Refractive error in minus cylinder, as reported by Rmax, was converted to plus cylinder; astigmatism was defined as having a cylinder ≥ 1.50 D. Anisometropia was defined as ≥ 1.50 D and ≥ 1.00 D in the difference in spherical or cylindrical error, respectively.

Amblyopia was defined as unilateral if there was a ≥2-line interocular difference in the BCVA, with at least 1 refractive risk factor (anisohyperopia ≥ 1.00 D, anisoastigmatism ≥ 0.50 D, or anisomyopia ≥ 2.00 D), and bilateral if BCVA in each eye was <20/50 for children 3 years old and <20/40 for children ≥4 years old with at least 1 refractive risk factor (hyperopia > 4.00 D for 3-year-olds and > 3.50 D for 4- to 5-year-olds, astigmatism of > 2.00 D for 3-year-olds and >1.50 D for 4- to 5-year-olds, myopia >3.00 D for 3-year-olds and > 1.50 D for 4- to 5-year-olds). Amblyopia classification is detailed in Figure 2 . Based on BCVA of the amblyopic eye, unilateral amblyopia was categorized as mild (≥20/30), moderate (20/40 to 20/80), and severe (≤20/100).

Figure 2

Amblyopia classification. Patients were classified as amblyopic according to this flow diagram. To qualify for bilateral amblyopia, while a maximum best-corrected visual acuity was specified for both eyes, only 1 eye had to meet refraction requirements. SE = spherical equivalent.

Results are reported as the mean (standard deviation) with a 95% confidence interval (CI) where applicable. Statistical analyses include the Student t test, ANOVA with Bonferroni correction, and Kruskal-Wallis (KW) testing for variables that are not normally distributed, performed in Microsoft Excel (Redmond, Washington, USA) and R Studio (Boston, Massachusetts, USA). Because the results between the left eye and right eye were comparable, we report results for the right eye only.


During this 5-year period, 79,451 children were screened and met eligibility criteria for the first time by the UPVP using the autorefractor with the failure criteria previously mentioned and met inclusion criteria for the study. A summary of the population demographics can be found in Table 1 . The mean age was 4.3 (0.57) years, and 49% of subjects were female. Of the screened population, 79% (63,129 children) self-reported their ethnicity as Latino, and 31% (24,937 children) spoke Spanish as their primary language.

Table 1

Summary of Population Data for the Children Who Were Screened, Sub-grouped by Self-reported Ethnicity, Age, and Sex

Ethnicity Age in Years Sex Total
3 (37%) 4 (54%) 5 (9%) Female (49%) Male (51%) Unknown (1%)
Asian (5%) 1587 1987 466 1924 2080 36 4040
Black/African American (5%) 1624 2203 519 2151 2177 18 4346
Latino (79%) 23,791 34,425 4913 30,705 31,823 601 63,129
Other/Unknown (6%) 1492 2361 618 2124 2295 52 4471
White (4%) 1222 1777 466 1683 1762 20 3465
Total 29,716 42,753 6982 38,587 40,137 727 79,451

Of the 79,451 children who were screened and met eligibility criteria, 18% (14,549 children) failed the initial screening and were offered to return on a subsequent day for a repeat screening and, if indicated, cycloplegic examination. Of the children who failed screening and were offered examination, 56% (8,101) attended their examination appointment after parental consent was obtained, 84% (6,779) of whom actually received a cycloplegic examination and 16% (1,322) of whom did not receive a cycloplegic examination or receive glasses after determination that there was no significant visual pathology based on re-screening of VA and refraction, and per the ophthalmologist’s clinical judgment.

In total, 86% of children (n = 5,883) who received a cycloplegic examination needed glasses due to myopia, hyperopia, and/or astigmatism; an additional 498 children (0.6% of the screened population) received glasses by virtue of already being under the care of an optometrist or ophthalmologist. In sum, 7.9% of the screened population received glasses.

Visual Acuity

Uncorrected VA (UCVA) in logMAR among examined children was 0.38 (0.19) (20/48) (95% CI: 0.01 to 0.75) and BCVA was 0.19 (0.14) (20/31) (95% CI: -0.09 to 0.46). Visual acuity did not vary by sex: the average male BCVA was 0.18 (0.14) (20/30) (95% CI: −0.09 to 0.45), and the average female BCVA was 0.19 (0.14) (20/31) (95% CI: 0.08 to 0.46) ( P = .23) ( Table 2 ).

Table 2

Visual Acuity Results (N, %) for the Right Eyes of the Examined Population

Age (Years) Total
3 4 5
UCVA (logMAR) n = 1598 n = 4039 n = 1098 n = 6735
≤0.2 (20/32) 312 (20%) 1012 (25%) 316 (29%) 1640 (24%)
0.3 to 0.4 (20/40 to 20/50) 618 (39%) 1607 (40%) 446 (41%) 2671 (40%)
0.5 to 0.6 (20/63 to 20/80) 436 (27%) 1032 (26%) 238 (22%) 1706 (25%)
≥0.7 (20/100) 232 (15%) 388 (10%) 98 (9%) 718 (11%)
BCVA (logMAR) n = 1445 n = 3756 n = 1041 n = 6242
≤0.2 (20/32) 895 (62%) 2716 (72%) 799 (77%) 4410 (71%)
0.3 to 0.4 (20/40 to 20/50) 460 (32%) 878 (23%) 203 (20%) 1541 (25%)
0.5 to 0.6 (20/63 to 20/80) 53 (4%) 108 (3%) 30 (3%) 191 (3%)
≥0.7 (20/100) 37 (3%) 54 (1%) 9 (1%) 100 (2%)

Only gold members can continue reading. Log In or Register to continue

Stay updated, free articles. Join our Telegram channel

Mar 14, 2020 | Posted by in OPHTHALMOLOGY | Comments Off on Refractive Errors and Amblyopia Among Children Screened by the UCLA Preschool Vision Program in Los Angeles County

Full access? Get Clinical Tree

Get Clinical Tree app for offline access