The impact of vision screening in preschool children on visual function in the Swedish adult population

Artificial intelligence-based screening for amblyopia and its risk factors: comparison with four classic stereovision tests

Introduction

The development of costs-effective and sensitive screening solutions to prevent amblyopia and identify its risk factors (strabismus, refractive problems or mixed) is a significant priority of pediatric ophthalmology. The main objective of our study was to compare the classification performance of various vision screening tests, including classic, stereoacuity-based tests (Lang II, TNO, Stereo Fly, and Frisby), and non-stereoacuity-based, low-density static, dynamic, and noisy anaglyphic random dot stereograms. We determined whether the combination of non-stereoacuity-based tests integrated in the simplest artificial intelligence (AI) model could be an alternative method for vision screening.

Methods

Our study, conducted in Spain and Hungary, is a non-experimental, cross-sectional diagnostic test assessment focused on pediatric eye conditions. Using convenience sampling, we enrolled 423 children aged 3.6-14 years, diagnosed with amblyopia, strabismus, or refractive errors, and compared them to age-matched emmetropic controls. Comprehensive pediatric ophthalmologic examinations ascertained diagnoses. Participants used filter glasses for stereovision tests and red-green goggles for an AI-based test over their prescribed glasses. Sensitivity, specificity, and the area under the ROC curve (AUC) were our metrics, with sensitivity being the primary endpoint. AUCs were analyzed using DeLong's method, and binary classifications (pathologic vs. normal) were evaluated using McNemar's matched pair and Fisher's nonparametric tests.

Results

Four non-overlapping groups were studied: (1) amblyopia (n = 46), (2) amblyogenic (n = 55), (3) non-amblyogenic (n = 128), and (4) emmetropic (n = 194), and a fifth group that was a combination of the amblyopia and amblyogenic groups. Based on AUCs, the AI combination of non-stereoacuity-based tests showed significantly better performance 0.908, 95% CI: (0.829-0.958) for detecting amblyopia and its risk factors than most classical tests: Lang II: 0.704, (0.648-0.755), Stereo Fly: 0.780, (0.714-0.837), Frisby: 0.754 (0.688-0.812), p < 0.02, n = 91, DeLong's method). At the optimum ROC point, McNemar's test indicated significantly higher sensitivity in accord with AUCs. Moreover, the AI solution had significantly higher sensitivity than TNO (p = 0.046, N = 134, Fisher's test), as well, while the specificity did not differ.

Discussion

The combination of multiple tests utilizing anaglyphic random dot stereograms with varying parameters (density, noise, dynamism) in AI leads to the most advanced and sensitive screening test for identifying amblyopia and amblyogenic conditions compared to all the other tests studied.

The Burden of Pediatric Visual Impairment and Ocular Diagnoses in Barbados

Visual impairment (VI) negatively affects a child's quality of life. The prevalence of VI in the Caribbean is nearly three times higher than in the United States, but the causes remain uncertain. This study leverages Barbados' unique eye care system to survey the eye diseases and VI prevalence in Barbadian children. Medical records of all patients aged <19 years who received ophthalmic care in Barbados' two public eye care centers between January and December 2019 were reviewed, capturing the entirety of public pediatric eye care within the study period. Age at the first visit to the clinic and at the final visit in 2019, sex, best-corrected visual acuity (BCVA), past medical history, and clinical diagnoses were extracted and analyzed. VI was defined as a BCVA of 6/12 or worse in the better-seeing eye. There were 3278 patient records with a mean age at the first visit of 7.8 ± 3.9 years. There were 80 (2.4%) children with VI, 62.5% of which were attributed to amblyopia. A total of 94% of VI was preventable or treatable. The most common diagnoses were refractive error (87.5%), strabismus (27.5%), and allergic eye disease (20.0%). Amblyopia is the major cause of pediatric VI in Barbados and is largely avoidable.

Child health screening program in French nursery schools: Results and related socioeconomic factors

Objectives

The study aims to describe the output of routine health screening performed in French nursery schools by the maternal and child health services among children aged 3-4 years and to quantify the level of early socioeconomic health disparities.

Methods

In 30 participating départements, data on screening for vision and hearing impairments, overweight and thinness, dental health, language, psychomotor development, and immunizations were collected for children born on specific dates in 2011 and enrolled in nursery school in 2014-2016. Information was collected on the children, their socioeconomic characteristics and on the school attended. Odds of abnormal screening results were compared for each socioeconomic factor by logistic regressions adjusted for age, sex, prematurity and bilingualism.

Results

Among the 9,939 children screened, prevalence of disorders was 12.3% for vision, 10.9% for hearing, 10.4% for overweight, 7.3% for untreated caries, 14.2% for language and 6.6% for psychomotricity. Newly detected visual disorders were more frequent in disadvantaged areas. Children with unemployed parents were three time more likely to have untreated caries and twice as likely to present language or psychomotor impairments; 52% were referred to a health professional following screening compared to 39% of children with employed parents. Except for children in disadvantaged areas, vaccine coverage was lower among disadvantaged groups.

Conclusion

The prevalences of impairments, which are higher among disadvantaged children, highlight the potential preventive impact of systematic screening under the comprehensive maternal and child healthcare program. These results are important to quantify early socioeconomic inequalities in a Western country known for its generous social welfare system. A more holistic approach to child health is needed with a coherent system involving families and aligning primary care, local child health professionals, general practitioners, and specialists. Further results are needed to evaluate its impact on later child development and health.

Worsening vision at age 4-5 in England post-COVID: Evidence from a large database of vision screening data

PURPOSE

Myopia prevalence has increased in the UK at age 10-16y, but little is known about younger children. We hypothesise that if the 'myopia epidemic' is affecting young children, then there will be increasing rates of bilateral reduced unaided vision (V) at vision screenings of children 4-5 years of age.

METHODS

Retrospective anonymised data from computerised vision screening at age 4-5 years were analysed from serial cross-sectional data. Refractive error is not assessed in UK vision screening, so vision was investigated. Data were only included from schools that screened every year from 2015/16 to 2021/22. The criterion used was unaided monocular logMAR (automated letter-by-letter scoring) vision >0.20 in both the right and left eyes, so as to maximise the chances of detecting bilateral, moderate myopia rather than amblyopia.

RESULTS

Anonymised raw data were obtained for 359,634 screening episodes from 2075 schools. Once schools were excluded where data were not available for every year and data were cleaned, the final database comprised 110,076 episodes. The proportion (percentage and 95% CI) failing the criterion from 2015/16 to 2021/22 were 7.6 (7.2-8.0), 8.5 (8.1-8.9), 7.5 (7.1-7.9), 7.8 (7.4-8.2), 8.7 (8.1-9.2), 8.5 (7.9-9.0) and 9.3 (8.8-9.7), respectively. The slope of the regression line showed a trend for increasing rates of reduced bilateral unaided vision, consistent with increasing frequency of myopia (p = 0.06). A decreasing linear trendline was noted for children 'Under Professional Care'.

CONCLUSIONS

For children 4-5 years of age, there were signs of reduced vision over the last 7 years in England. Consideration of the most likely causes support the hypothesis of increasing myopia. The increase in screening failures highlights the importance of eye care in this young population.

Design, methodology, and baseline of whole city-million scale children and adolescents myopia survey (CAMS) in Wenzhou, China

Background

Myopia is the most common visual impairment in children and adolescents worldwide. This study described an economical and effective population-based screening pipeline and performed the project of a million scale children and adolescents myopia survey (CAMS), which will shed light on the further study of myopia from the level of epidemiology and precision medicine.

Methods

We developed a novel population-based screening pattern, an intelligent screening process and internet-based information transmission and analysis system to carry out the survey consisting of school children in Wenzhou, China. The examination items include unaided distance visual acuity, presenting distance visual acuity, and non-cycloplegic autorefraction. Myopia and high myopia were defined as spherical equivalent (SE) ≤ − 1.00 diopters (D) and SE ≤ − 6.00 D, respectively. Next, the reports of the vision checking were automatically sent to parents and the related departments. The CAMS project will be done two to four times annually with the support of the government. An online eyesight status information management system (OESIMS) was developed to construct comprehensive and efficient electronic vision health records (EVHRs) for myopia information inquiry, risk pre-warning, and further study.

Results

The CAMS completed the first-round of screening within 30 days for 99.41% of Wenzhou students from districts and counties, in June 2019. A total of 1,060,925 participants were eligible for CAMS and 1,054,251 (99.37% participation rate) were selected through data quality control, which comprised 1305 schools, and 580,609, 251,050 and 170,967 elementary, middle, and high school students. The mean age of participants was 12.21 ± 3.32 years (6–20 years), the female-to-male ratio was 0.82. The prevalence of myopia in elementary, middle, and high school students was 38.16%, 77.52%, and 84.00%, respectively, and the high myopia incidence was 0.95%, 6.90%, and 12.98%.

Conclusions

The CAMS standardized myopia screening model involves automating large-scale information collection, data transmission, data analysis and early warning, thereby supporting myopia prevention and control. The entire survey reduced 90% of staff, cost, and time consumption compared with previous surveys. This will provide new insights for decision support for public health intervention.

Supplementary Information

The online version contains supplementary material available at 10.1186/s40662-021-00255-1.

Scope and costs of autorefraction and photoscreening for childhood amblyopia-a systematic narrative review in relation to the EUSCREEN project data

Background

Amblyopia screening can target reduced visual acuity (VA), its refractive risk factors, or both. VA testing is imprecise under 4 years of age, so automated risk-factor photoscreening appears an attractive option. This review considers photoscreening used in community services, focusing on costs, cost-effectiveness and scope of use, compared with EUSCREEN project Country Reports describing how photo- and automated screening is used internationally.

Methods

A systematic narrative review was carried out of all English language photoscreening literature to September 10th 2018, using publicly available search terms. Where costs were considered, a CASP economic evaluation checklist was used to assess data quality.

Results

Of 370 abstracts reviewed, 55 reported large-scale community photoscreening projects. Five addressed cost-effectiveness specifically, without original data. Photoscreening was a stand-alone, single, test event in 71% of projects. In contrast, 25 of 45 EUSCREEN Country Reports showed that if adopted, photoscreening often supplements other tests in established programmes and is rarely used as a stand-alone test. Reported costs varied widely and evidence of cost-effectiveness was sparse in the literature, or in international practice. Only eight (13%) papers compared the diagnostic accuracy or cost-effectiveness of photoscreening and VA testing, and when they did, cost-effectiveness of photoscreening compared unfavourably.

Discussion

Evidence that photoscreening reduces amblyopia or strabismus prevalence or improves overall outcomes is weak, as is evidence of cost-effectiveness, compared to later VA screening. Currently, the most cost-effective option seems to be a later, expert VA screening with the opportunity for a re-test before referral.

Amblyopia Outcomes Through Clinical Trials and Practice Measurement: Room for Improvement: The LXXVII Edward Jackson Memorial Lecture

PURPOSE

To describe amblyopia prevalence and outcomes using results from randomized studies and a clinical registry.

DESIGN

Review of published studies, analysis of data in Intelligent Research in Sight (IRIS) Registry from 2013 to 2019, personal perspective.

METHODS

Literature review, analysis of IRIS Registry data and IRIS-50, a visual acuity quality measure.

RESULTS

Clinical trials have reduced the treatment burden of amblyopia by reducing hours of patching and frequency of atropine eye drops with clinical success of about 83%. There is no appreciable age effect if treatment is started before 5 years of age, outcomes are stable to at least 15 years of age, and treatment can be somewhat effective until 12 years of age. The IRIS Registry identified 1,760,066 individuals with amblyopia for a prevalence of 2.47%. Refractive error alone accounted for 68.9% of childhood cases. Mean amblyopic eye visual acuity improved 1.8 lines for children 3-6 years of age and 0.8 lines for 7-12 years, but mean residual amblyopia was more than 2 lines. Among 18,841 children aged 3-7 years eligible for IRIS-50, 77.3% were successful. The odds ratios for success were significantly lower for African-American (0.67; 95% confidence interval [CI] = 0.58 to 0.78) and Hispanic or Latino (0.84; 95% CI = 0.75 to 0.94) children compared with white children.

CONCLUSIONS

Clinical trials provided evidence of a beneficial effect from several treatments, with substantially reduced doses than previously recommended. Registry data from clinical practice found residual visual acuity impairment among all ages and races, especially among minorities.