Opportunities for a New Study on Myopia

Invited speaker Bill Geisler (University of Texas at Austin Center for Perceptual Systems) observed that as myopia has continued to increase rapidly across the world, most efforts have focused only on developing medical interventions, such as specialized glasses and contact lenses and atropine eye drops. He underscored the need for efforts that focus on determining and treating the disease's underlying causes. Furthermore, he pointed out that most myopia research has been conducted in ophthalmology and optometry departments and in medical industries, but a wider range of expertise—in visual behavior and neurophysiology, physical and physiological optics, display technology, experimental methodology, human factors, behavior genetics, and communication—is crucial to address the problem.

Invited speaker David Williams (University of Rochester Institute of Optics) echoed the need for a fresh perspective on myopia with a study scope that extends well beyond a summary of current technologies. He asserted that, in some ways, myopia is a more challenging problem than COVID-19—while effective vaccines are available for the latter, the best intervention (e.g., optical, behavioral, or pharmacological) to prevent or mitigate the effects of myopia remains unknown.

Geisler and Williams expressed their hope that the consensus study will raise awareness about the severity of myopia, not only among public school administrators and the general public but also among federal and state governments, foundations, philanthropies, and, especially, technology industry leaders who are well positioned to provide vital resources to help address the problem.

Innovative Research Directions

Invited speaker Martin Banks (Herbert Wertheim School of Optometry & Vision Science, University of California, Berkeley) pointed out that there is a strong evidence base for an environmental role in the onset of myopia. He explained that the spectrum of light and light intensity, as well as the distance of objects, differ in indoor and outdoor environments, and the more time that children spend outdoors, the better their chances of preventing myopia.²

Banks described a 2012 study on the environmental factors that contribute to myopia.³ That study described that the retinal image depends on the three-dimensional structure of the visual environment, the optics of the eye, the shape of the retina, and accommodation. He observed that this dataset had two problems: the images were computer generated and the fixation point had to be assumed. Thus, to confirm the study's assertions, he and his team developed an eyes and scenes tracker, a head-mounted device that includes a binocular eye tracker and a stereo scene camera. They computed relative distances and retinal disparities as adults engaged in everyday tasks in real environments.⁴ Analysis of these data⁵ revealed that depending on the specific task, participants tended to see farther points in the upper field and nearer points in the lower field. Furthermore, the retinal images from the indoor scenes had much larger dioptric values and variation in distance than those of the outdoor scenes.

Banks stressed that data like these should be collected from children to determine which aspects of the environment are predictive for developing myopia. As an example, he referenced Agostino Gibaldi's lightweight prototype headset to measure the illumination and relative distance environments as children engage in various activities. For future work, Banks suggested combining measurements of distance statistics in the central visual field with an eye model to generate statistics of retinal images.

During a brief question-and-answer session with Banks, a committee member expressed concern that a myopic child might prefer to spend more time looking at near objects indoors than at distant objects outdoors, and Banks reiterated the need to develop technologies for studying children and beginning interventions early. He championed the idea of projecting such data onto animal models but noted that it requires substantial work to develop an appropriate eye model over the central visual field. Another committee member inquired whether an eye model that is statistically accurate for populations is sufficient and, if not, whether eye models could be built for individuals. Banks replied that his colleagues have a small set of model eyes and that all aberrations are considered to generate expected retinal images. He suggested developing an inexpensive, easy way to measure axial length for more individualized assessment.

Myopia Risk and Prevention in Australia

Invited speaker David Mackey (University of Western Australia Medical School, Centre for Ophthalmology and Visual Science) explained that in part because much of Australia has high sunlight intensities, it has the highest skin cancer rate in the world. At the same time, however, it also has one of the lowest rates of myopia. Reviewing the results of four studies conducted in Australia, he found that the rate of myopia in adults over two decades was 30% or less—far lower than the rates in cities across East Asia.⁶ In one study (known as the Raine Study), the prevalence of myopia was initially ~25% among young adults; after 8 years, myopia prevalence had increased in one-third of those who had myopia at age 20 and appeared in 14% of those who had not been myopic at age 20.⁷ The increased risk factors for these young adults included being female of East Asian ethnicity, having parents with myopia, and less sun exposure.

Mackey noted that there are three environmental risk factors for myopia—lack of time spent outdoors, increased duration and level of education, and longer periods of near work—and that they are intertwined. Although not directly involved in the recognized mechanism of myopia, several markers of time spent outdoors are important to consider in understanding myopia risk: vitamin D levels, arm nevus (or mole) counts, bone density, hand wrinkling, pterygium,⁸ conjunctival ultraviolet autofluorescence,⁹ and skin cancer rates. For example, the Busselton Healthy Ageing Study found that 7.7% of participants born between 1946 and 1964 had skin cancer when examined in 2010, but only 11.6% of them had myopia; of those who did not have skin cancer, 21.9% had myopia.¹⁰ Thus, he said, Australian parents are faced with a “balancing act”: more time outdoors could increase their children's risk for skin cancer but decrease their risk for myopia. He encouraged the use of public health measures that are tailored to specific populations to increase outdoor time for children.

Reflecting on the other environmental risk factors for myopia, Mackey indicated that the Raine Study also analyzed television, computer, and smartphone screen time among its young adult participants (aged 20, 22, 27, and 28 years old). Myopia progressed faster in those with “consistently high” or “consistently very high” (measured in hours per day) screen time on computers but not on smartphones.¹¹ Mackey and his colleagues also studied the role of genetics in understanding myopia risk and found that the performance of the polygenic risk score (PRS) approached the level required for clinical utility in Europeans but not in people with other ancestries. Also, a PRS for refractive error was not predictive of myopic maculopathy risk once spherical equivalent was accounted for.¹²

Mackey supported strategies to prevent myopia over efforts to cure myopia. Thus far, he said, increased supervised time spent outdoors at school and a low concentration atropine regimen have been the only preventive approaches demonstrated to be effective in clinical trials in China and Hong Kong, although U.S. studies have mixed results on the benefits of atropine. He advocated for combining people's PRS with pre-myopia levels, family history, at-risk behaviors, and axial length trajectory to identify pre-myopic children and provide personalized interventions. He added that the ability to measure axial length trajectories would be especially valuable because axial length change is greater in children developing myopia.¹³

Using Artificial Intelligence

Invited speaker Daniel Ting (Singapore National Eye Centre) described myopia as a highly complicated disease because of the influence of both genetics and the environment. He noted that more people are affected by myopia (5 billion) than by hypertension, diabetes, age-related macular degeneration, and glaucoma combined. High myopia, defined as –5 diopters¹⁴ (D) or greater, which affects ~20% of myopic people, is particularly problematic in Asia but is increasing globally.

Ting discussed ways that technology could be used to improve health care, as artificial intelligence (AI) applications are emerging for segmentation, classification, and prediction tasks. For example, deep learning enables the analysis of multiple data formats (e.g., text, speech, images) with multiple levels of abstractions. AI can be used to detect lung cancer, assess cardiac health, and classify skin lesions. In ophthalmology in particular, he noted, AI has the potential to be used for retina imaging to detect disease and predict disease progression. He described the use of deep learning with baseline photos from fundus imaging¹⁵ to predict children's risk of developing high myopia within 5 years.¹⁶ Approximately 1,800 eyes were used to train and validate the software, and 200 eyes were used to test algorithms: the photos were found to have high predictive power (~90% accuracy). Ting and his colleagues conducted another study¹⁷ using more than 200,000 images from across six countries to demonstrate that combining AI with fundus imaging had more than 90% accuracy in detecting high myopia and myopic macular degeneration (a myopia complication that might cause vision loss as a patient ages). This system outperformed six retinal specialists in detecting these diseases.

Ting also presented opportunities for and challenges of using large language models in medicine, which have been successful in organizing and summarizing complex and unstructured electronic health records, proofreading manuscripts (using ChatGPT), and supporting clinicians in decision making. He considered ways for AI to play a role in addressing myopia in the future, such as by combining highly complicated datasets (e.g., personal bio-data, environmental data, retina imaging, axial length, genetics); screening for myopia progression; determining myopia treatment; managing atropine treatment; and creating technologies to encourage people to go outdoors. However, he cautioned that AI should be used responsibly, with consideration for data safety, bias, ethics, generalizability, explainability, and transparency. Before using AI to address myopia, he proposed that one should define the intended use of data on the environment and disease prevalence, choose the health economic model based on the intended use, evaluate diagnostic performance, and evaluate direct and indirect costs.

Discussion

Mackey indicated that a lack of time outdoors does influence myopia progression in young adults. He stressed that prevention is still the “best investment” and advised targeting children aged 6–8 years old for increased outdoor time—a group that was particularly affected by COVID-19 lockdowns in China and Hong Kong. Mackey continued that Australian data suggest that children spend 1–2 hours outside per day; however, parents have to ensure that their children wear hats, shirts, and sunscreen and are not outdoors in the middle of the day. He suggested non-dark, ultraviolet-blocking sunglasses to protect against the harmful effects of the sun without voiding the myopia prevention benefits. He also cautioned against providing contradictory health advice to communities, an important lesson learned during the COVID-19 pandemic.

One committee member questioned whether AI could be used as an intervention tool to offer real-time advice to people about adequate lighting levels. Ting championed this path but noted that no consensus exists about the level of luminance that determines myopia progression. However, he said, the more data that are captured, the more patterns of different risk levels that can be identified. Mackey added that quantifying sunlight exposure is difficult and requires better measures than are currently available.

Recalling Mackey's discussion of the PRS, a committee member wondered if different phenomena are occurring in Europe and Asia and whether diagnosis, treatment, and prevention should vary by population. Mackey pointed out that studies on more diverse populations are required not only to avoid bias in the data but also to identify both universal genes and genes common to particular subgroups. He added that cultural behaviors in different parts of the world influence outdoor time (e.g., staying inside to study or to avoid developing freckles), which might shift in future generations and affect people's risks. A committee member then inquired about myopia control efforts in Singapore. Ting replied that more than 90% of myopic children in Singapore are on atropine, which is available in different concentrations depending on the age of onset and the progression of myopia.

Another question was posed about the economic impact of myopia treatments. Ting said that it could be multifaceted; for example, the amount of medical staff time could be reduced, along with the number of children prevented from becoming myopic and experiencing complications. He added that choosing the right health economic tool is important when designing a health economic study based on a particular disease, as is determining how AI will be used (i.e., as an assistive or predictive tool).

A committee member observed that ChatGPT tends to confabulate. Because it could become less reliable over time, he asked how to ensure that it does not “make up” medical recommendations. Ting said that commercially available ChatGPT is not appropriate for health care use; before leveraging ChatGPT to build algorithms, data issues have to be addressed and training prompts reengineered.

Public Health Approaches in Taiwan

Invited speaker Pei-Chang Wu (Chang Gung University) explained that high myopia could lead to blindness and significantly increases one's risk of developing cataracts, glaucoma, retinal detachment, and myopic macular degeneration. After onset, myopia can progress –1 D per year until the end of adolescence, making childhood myopia a chronic and progressive disease. Reflecting on the many complications of myopia, he highlighted the value of prevention through three public health approaches: awareness and recognition, incentives, and policy interventions.

First, awareness and recognition are critical to the successful implementation of myopia interventions. Education for parents, teachers, stakeholders, and the public is crucial so that people recognize protective factors, such as spending time outdoors. Addressing concerns about the harm of ultraviolet exposure with increased time outdoors, he stressed that even when children wear sunglasses and hats, myopia can still be prevented. He added that moderate light intensity (e.g., under a tree, in hallways) is effective in achieving both myopia prevention and partial control. In a school-based cluster randomized trial, children who spent 11 or more hours outdoors per week had 54% lower odds of rapid myopia progression and 30% efficacy of myopia control.¹⁸ Similarly, another study found that ~120 minutes per day with outdoor light exposure at school can reduce myopia incidence by 63.7%.¹⁹

Second, Wu noted that promotion through rewards and reminders is an effective approach to prevent childhood myopia. For example, after children aged 6–12 years participated in an incentive-based outdoor physical activity program in Singapore for 6 months (which included myopia education, good eye care habits, structured weekend outdoor activities, and incentives to increase daily steps), their time outdoors increased significantly. He also described a randomized clinical trial in which parents were sent SMS messages twice per day for 1 year with reminders to take their children outdoors. Results indicated greater light exposure and increased time outdoors, as well as a significant decrease in myopia progression and axial length elongation.²⁰

Third, Wu discussed the role of policy intervention in myopia prevention in four school policy studies. One of the interventions that used intermittent daily increases in time outdoors to 80 minutes per day resulted in reduced myopic incidence by 50%.²¹ He said that curriculum changes have also been made in Taiwan's primary schools, with more frequent outdoor activities throughout the day. Furthermore, since 1980, school nurses in Taiwan have performed unaided visual acuity (VA) screenings every semester to detect myopia in children, with referral to an ophthalmologist for those found to have reduced VA. He noted that in 1999, the Ministry of Education initiated a 5-year program that included changing lights in classrooms, changing desk heights, and adding near-work breaks and eye exercises; however, the prevalence of reduced VA continued to worsen and increase. After additional studies, in 2009 Wu and his colleagues persuaded the government to promote more outdoor activity for children, which became the priority of the campaign for myopia prevention, and the prevalence of reduced VA began to reverse and decrease.²² The prevalence of reduced VA declined from 50% in 2001 to 35% in 2010.²³

Increasing Global Access to Eye Care

Invited speaker Andrew Bastawrous (International Centre for Eye Health at the London School of Hygiene & Tropical Medicine) remarked that millions of people need to be connected to eye care services: some cannot find care and others do not know that solutions are available. He noted that the effects of uncorrected vision and eye health issues could be “catastrophic”: children with poor vision are two to four times less likely to complete their education, and poor eye health in general increases mortality risk, degrades social connections, and decreases quality of life. For example, cataracts (a leading cause of blindness) could be easily and inexpensively corrected with surgery, and many workers could increase their productivity simply by wearing glasses.

Bastawrous, who is the cofounder and chief executive officer of Peek Vision, said that the company's technology enables anyone anywhere to be screened for vision issues. He explained that Peek Vision redesigned the vision test because the standard Snellen chart is problematic for those without literacy skills and is complicated for examiners to administer and score. The new test is available through a mobile app, Peek Acuity, which has been downloaded in 190 countries. The person administering the test only needs to know how to use a smartphone: the person being tested points in the direction the letter on the screen is facing, and the lay person administering the test swipes the screen in that direction. An algorithm then determines if the person gives the right answers. He indicated that building the app required many iterations—for example, the developers learned that ambient light affects one's accuracy, and the person administering the test has to be 2–3 meters from the person taking the test. He confirmed that after review, Peek Acuity was found to be reliable, reproducible, and in alignment with reference standards for adults and children over age 6. He noted that it can also be used on younger children.

After identifying people in need of care through the vision test, Peek Vision works to increase access to vision services and prospectively monitoring care. Bastawrous described working with an ophthalmologist in Kenya to design a trial in which teachers were equipped to use Peek Acuity to conduct vision tests in the classroom. The trial also included notifications to local hospitals about the number of children referred for care, as well as automated messages sent to parents with information about next steps. The randomized controlled trial screened 21,000 children and identified 900 visually impaired children in only 9 days, with 25 teachers administering the vision test via the app. The trial also compared how many children visited the hospital for an eye exam within 28 days after either the Peek Acuity and an SMS message or the Snellen chart and a letter: the results were 54% for the Peek Acuity protocol and 21% for the Snellen protocol. After realizing that many of the children did not require hospital-level services and could be treated elsewhere, 96% of the children who needed care were able to obtain care.

Bastawrous discussed a parallel study conducted in India that identified children with poor vision, triaged for refractive error, distributed glasses, and followed up after 3 months: at that time, only 50% were wearing their glasses. Now, he continued, the goal is to determine at which stage children are “falling out” of the care pathway and why and to build on the “impact pipeline.”

Peek's software and data intelligence platform creates different capabilities and design configurations based on user need in an app called Peek Capture, which identifies the patient prospectively monitoring care. He explained that once people are screened, they are referred to a treatment location. All of the data are seen live in Peek Admin. Data on how people move through the pathway and how many obtain treatment are tracked, and programs can be designed specifically for the people who have not received treatment (e.g., by addressing cost, transportation, or language barrier issues). As of September 2023, Peek Vision supported 59 programs across 12 countries and had reached 5 million people across all age groups.

The Future of School Eye Health: Making a Difference with School Eye Health Rapid Assessment

Invited speaker Priya Morjaria (International Centre for Eye Health at the London School of Hygiene & Tropical Medicine), who is also associated with Peek Vision, elaborated that the company works with eye health, governmental, and international nongovernmental organizations to make “large-scale, sustainable improvements to eye health systems” by “collaborating on a patient journey that reflects local needs and resources, adapting a software workflow to support pathways to treatment, and supporting data analysis to help programs increase impact.” She explained that the Rapid Assessment of Avoidable Blindness (RAAB), an adult-based survey, and the School Eye Health Rapid Assessment (SEHRA), a child-based survey, are two initiatives translated into Peek-powered school and community programs.

Morjaria remarked that SEHRA is based on the principles of RAAB (i.e., with a restricted population, simplified process, and software for data entry and analysis) but was developed specifically to understand the eye health needs of children. It was initially supported in 2019 by a grant from the U.S. Agency for International Development and various pilots to develop and validate a planning tool began in February 2023 in several countries. Global experts in school and general eye health, policy implementers, educators, and program funders convened to better understand the needs for such a tool.

Morjaria shared that after collecting this input, a framework was developed for SEHRA, including the Minto Method Scoping Module and the School Survey Magnitude Module. The Minto module helps determine how suitable an environment is for a school eye health program and identifies issues that might arise if a program is implemented. This module has several components to evaluate, including sectoral legislation, institutional and service delivery environment, human resources, supply chain, and barriers. The School module then helps inform how many children need eye care services in a particular location and what types of services they need. To begin this module, the location for the survey is chosen; a sample of participating schools that is representative of the area is determined; and the survey team selection and mobile app training are completed, with predefined case studies to encourage reliable data collection and confidence in results.

Morjaria noted that the SEHRA workflow begins when children are screened at school and ends when a diagnosis and a referral to a clinical unit are provided to the children who need them. While these data are being collected via the mobile app, they are pushed into the cloud, she explained, and real-time updates track the survey to highlight any issues that need to be addressed. Reports on outcomes emerge at the end of the survey and further analysis is done. She clarified that because these detailed reports are automated, survey implementation teams do not need to include statisticians and analysts.

Morjaria reiterated that the key goal for SEHRA is the development of a planning tool: using data to understand the size and specific aspects of the problem and any potential bottlenecks before planning a school eye care program is critical to meet local needs. For example, she said that if the data show that only 50% of school-age children in a region are enrolled in school, then an additional program outside of school might be beneficial. Furthermore, the survey sample is categorized by how many children have each kind of eye health issue—data that help determine which care facilities and care levels are required to manage each issue and better use resources. With SEHRAs already completed in parts of India and Pakistan, she indicated that Peek Vision has committed to doing eight more SEHRAs within 18 months.

Discussion

Committee members suggested focusing on whether children should be spending time outdoors in one long recess or over the course of several shorter sessions. Wu responded that in some animal studies, interventions with more frequent, shorter periods of light exposure were more effective. This approach also works better for the school day, he continued, as it provides more breaks from long-duration near work.

A committee member then inquired about strategies to address the situation, noting that Bastawrous reported that only 50% of children were wearing their glasses after only 3 months. Morjaria replied that allowing children to choose their frames increases their willingness to wear glasses. However, she pointed out that children who wear glasses are often bullied. Her team educates children about why certain celebrities wear glasses in the hopes that they might imitate them, and they also work with parents and communities to raise awareness about the importance of wearing glasses. Responding to a question about whether charging a nominal fee would increase compliance, Morjaria explained that her team has tried several approaches: providing glasses at no cost, offering a voucher to get glasses at another location, and requiring a small payment for glasses. She emphasized that the effectiveness of the approach depends on the cultural setting. A question was raised about whether any differences exist in terms of compliance based on diagnosis. Morjaria said that if children have significantly improved vision with the glasses, they might be more likely to wear them; therefore, considering the value that children perceive from the glasses is critical.

Early Detection of Myopia: What Is Known, What Needs to Be Known

Invited speaker Donna Fishman (director, National Center for Children's Vision and Eye Health [NCCVEH]), noted that ~25% of U.S. children have a vision disorder requiring treatment. She emphasized that disparities persist in children's vision and eye health because of a lack of standardization in vision screening mandates and funding across the country. Although local initiatives might be successful, she emphasized that they are not sustainable or scalable without funding. She championed a holistic approach to “ensure those most at risk of vision disorders have access to a strong system of care to reduce the incidence of unnecessary eye disease and improve learning opportunities, vision health, and future success.”

To achieve this goal, Fishman asserted that a wide range of actions are needed: strong policies, parent education, evidence-based vision screening, referrals to eye care, eye care and treatment, coordination between providers and screening programs, coordination between health and education sectors, data collection and surveillance, and continuous quality improvement and program evaluation at state and local levels. Most importantly, she stressed, these strategies have to work for the most medically underserved children (e.g., migrants, immigrants, refugees, rural residents, racial and ethnic minorities, and those with language or cultural challenges). Because children interact with many people and institutions, she continued, families; family engagement and support organizations; schools; early childhood education and care programs; and primary care, eye care, public health, policy makers, and community organizations have to be involved to promote eye health equity.

Fishman provided an overview of policies, guidelines, and programs for children's vision screening that contribute to early detection of myopia. Consensus recommendations for the medical community are available from eye care organizations and NCCVEH offers recommendations and guidelines for public health and lay screeners. The federal government only requires vision screening for Head Start and Medicaid's Early and Periodic Screening, Diagnostic and Treatment (EPSDT) benefit, so states and school districts set their own guidelines for screenings. Public health goals come from the National Academies, U.S. Preventive Services Task Force, Healthy People 2030,²⁴ and the Centers for Disease Control and Prevention (CDC), among others. She stressed that the lack of uniformity across these sectors, as well as the “unwinding” of Medicaid, which required states to begin redetermining eligibility for Medicaid enrollment postCOVID-19, reduce access and increase disparities.

Fishman noted that NCCVEH tracks state-level vision screening regulations for preschool- and school-age children (see Figure 1). She remarked that more elementary-age children are getting screened than children in middle and high school, which, owing to the trajectory of myopia, is problematic for those who are missed initially and not screened later. She added that even though a screening might be done in school, the lack of time, ability, and requirements to ensure eye care and the lack of funding for referral follow-up means that many children do not receive care; however, the data on this issue are incomplete. The barriers to receiving care vary and need to be identified.

FIGURE 1

School-based vision screening by state. SOURCE: https://preventblindness.org/vision-screening-requirements-by-state.

Fishman explained that data are collected in several ways, including clinical data through the American Academy of Ophthalmology's Intelligent Research in Sight system (which is not comprehensive for children of all ages or for all insurance carriers), the National Health and Nutrition Examination Survey (which is out of date), individual states (which are inconsistent), and the annual National Survey of Children's Health (which is the best current source but relies only on parents' reports). To improve surveillance, she offered the following recommendations: address gaps in data collection; unbundle vision from EPSDT and require uniform data reporting; collect comprehensive information in screening programs and unify vision screening data from disparate sources; conduct research on vision screening efficacy, social determinants of health effects on vision screening, and gaps in data; and conduct research on effective interventions to improve rates of eye care after screening. In closing, Fishman advocated for establishing accountability and performance measures for vision health in health insurance programs, children's developmental screening, well-child checks, and school screenings.

Perspectives on Vision Screening and Access to Care

Invited speaker Megan Collins (Johns Hopkins University Wilmer Eye Institute and Berman Institute of Bioethics) explained that the purpose of vision screening shifts as children age. Young children (aged 0–5) are screened primarily for amblyopia (reduced vision in one eye caused by abnormal visual development early in life) risk factors because amblyogenic, vision-threatening conditions can often be treated most effectively with early intervention. For older children (aged 5–18), vision screenings can help detect children who may have uncorrected refractive errors. Recent though limited population-based health studies for school-age children in the United States show that myopia prevalence, in particular, is on the rise. In addition, the effects of vision impairment caused by uncorrected refractive error begin to have more detrimental consequences for the students' daily lives during higher grades. She indicated that 80% of states require vision screening for school-age children at some time, but only five states require that screening to occur annually or biannually (see Figure 1).

Many professional organizations provide recommendations about vision screening in the school-age population. Most recommend screening every 1–2 years with acuity testing. Instrument-based photo screening is primarily recommended for the pre-school/pre-verbal population. The American Association for Pediatric Ophthalmology and Strabismus guidelines for instrument-based pediatric vision screening have refractive error cutoffs, including a myopia cutoff (in diopters), that are primarily focused on amblyopia risk factors. Collins emphasized that one gap is the lack of evidence-based guidelines for instrument-based screening referrals in the school-age population,

Collins further highlighted the disconnect between policy and practice by sharing the results of a national survey of 184 school-based vision programs that provide vision screening and eye exam services directly in the school setting: 92% screened elementary-age children, 95% used instrument-based vision screening, and 41% used distance VA testing. She pointed out that despite the fact that no guidelines endorse it in the school-age population, many school-based vision programs use instrument-based vision screening, likely because it is quicker than other methods, is objective, does not require a child's cooperation, and is likely more cost-effective.

Collins underscored that this lack of national vision screening policies creates gaps and issues; for example, pediatricians are not incentivized to offer vision screening and reimbursement for vision screening varies. These issues are compounded by disparities among economically disadvantaged and racial and ethnic minority groups for which eye disease prevalence and the need for vision screening are higher but access to vision screening and eye care is lower than advantaged groups. Collins suggested that researchers and policy makers should explore screening policies and practices for other childhood conditions (e.g., depression, obesity, dental problems, and anxiety), as there may be relevant lessons about effective national-level data collection efforts that the vision community could adopt. Despite these challenges, she continued, novel approaches to maximize screening and eye exam access are emerging from school-based vision programs and community health centers, including Federally Qualified Health Centers (FQHCs). These programs could be leveraged to increase collaboration with community eye care providers. In closing, she referenced an article describing how school-based vision programs help overcome access barriers and foster trusted partnership with the school community, improve student health literacy, and provide experiential learning opportunities in community-based initiatives.²⁵

Transforming School Health Services and Access

Invited speaker Jessie Mandle, representing the Healthy Schools Campaign (HSC), noted that HSC works at the “intersection of education, health, and equity.” HSC advocates for national and local policies and initiatives that support health and learning and it prioritizes comprehensive mental, physical, and behavioral health services; green schoolyards; and clean air at school. For example, HSC's Space to Grow Program aims to get more children outside during the school day. A nationally recognized program, it has transformed more than 34 asphalt schoolyards in historically underinvested communities in Chicago into “vibrant outdoor spaces” to play and learn. HSC also works to increase sustainable funding, cover more services in schools, expand the health care workforce, and encourage collaboration between education and health care sectors.

Mandle described different models for delivering and financing school health services. Some schools partner with community-based providers or host mobile vans, some operate school-based health clinics, and others offer telehealth. She stated that almost half of all U.S. children are covered by Medicaid or the Children's Health Insurance Program and Medicaid plays a unique enabling role in providing services through school district–employed providers. Medicaid can pay for school-based physical, mental, and behavioral health services if students are enrolled in Medicaid: the provided services are covered by the state; they are delivered by a qualified provider; states have appropriate billing, documentation, and oversight mechanisms in place; and the Centers for Medicare & Medicaid Services approves a state plan. The types of providers and services eligible for reimbursement in schools are numerous, she continued, including physician, nursing, and vision services.

Mandle explained that the Bipartisan Safer Communities Act has several provisions to support states in expanding Medicaid school-based services for states that expand Medicaid coverage for students. States that have expanded Medicaid school-based services are seeing positive results: a recent HSC analysis found that 25 states have already adopted policies to expand services for students, and she directed workshop participants to HSC's interactive map²⁶ to learn about services covered in each state. She asserted that work remains to ensure that states are maximizing Medicaid dollars; when states expand these programs, students, staff, families, and communities benefit.²⁷ In closing, she emphasized that both the green schoolyard movement and school-based Medicaid programs encourage sectors to work together to improve health and academic outcomes in schools.

Discussion

A committee member noted that greenspace initiatives could have significant impact on children's refractive error and inquired about efforts to extend the amount of time children spend outside during the school day. Mandle replied that HSC is encouraging longer recess times and discouraging punishments that take recess away from children. HSC's Space to Grow Program includes support to help teachers incorporate more outdoor learning. Collins added that because inner-city neighborhoods in particular might not have safe public greenspaces, opportunities to be outside during the school day are imperative.

Collins highlighted the difficulty of collecting the needed data at the state level; for example, the number of children being screened and receiving care is unknown. Some states do not have vision screening mandates (see Figure 1) or records indicating whether children in those states have poor vision health. She said that a national registry where all states could report such data would highlight discrepancies that need to be addressed. Fishman stressed that even though states might not have regulations, they might have guidelines and thus still conduct vision screening. She added that some states have funding for vision screening and eye care that is not considered regulation because it is not written in legislation.

Committee co-chair Terri Young (University of Wisconsin–Madison) considered the challenge of implementing vision screening in schools that already provide screenings for other health conditions and inquired about examples of partnerships for holistic screenings—especially given that unchecked refractive error could influence obesity, anxiety, and depression. Collins promoted CDC's Whole School, Whole Community, Whole Child model, which centers on the notion that physical and mental health issues influence a child's success in school. She noted that although coordinated screening efforts could be beneficial for students and may prove easier for schools, it requires system-level coordination and integration that are difficult to achieve on a large scale without adequate support and community buy-in. However, she noted that the Community School Model typically has an on-site nurse and school coordinator who could help facilitate school-based delivery of vision services, including vision screenings, eye exams, and eyeglasses, when indicated. Mandle highlighted the benefit of having a nurse on site every day for increased early detection and referral and reiterated that Medicaid can play a role in funding that position. She recognized Head Start for its successful holistic screening program for all students prior to enrollment. Fishman added that Head Start also tracks whether students obtain care.

Committee co-chair Kevin Frick (Johns Hopkins Carey Business School) invited Fishman to identify her priorities to improve the system of care. She responded that closing the gap between vision screenings and eye exams is critical to ensure that children receive care; however, all sectors have to be involved to achieve that goal. A committee member reflected on programs that bridge the gap between screening and visiting an eye care professional, noting that some national groups have advocated for comprehensive eye exams for all, while others advocate for including vision screening as part of well-child visits. In particular, she wondered why mobile vans are not used more often. Collins said that school-based vision programs that bring the services to the children via mobile vans or visiting optometrists or ophthalmologists can be highly effective, but these programs are not yet widely used because of a lack of funding, resources, and infrastructure. She encouraged leveraging FQHCs and community health centers. Mandle added that accountability frameworks incentivize funders, thus echoing the value of high-quality data and metrics.

A participant asked how many states have mandated comprehensive eye exams prior to school enrollment and Fishman directed participants to the NCCVEH website²⁸ for information about states that require them. Collins estimated that most professional pediatric, ophthalmology, and optometric associations recommend vision screenings to detect children at risk for an eye condition. Only a few states mandate a comprehensive eye exam and Collins expressed her concern about expanding such mandates given the current evidence about the value of screenings. In addition, it would be difficult for all children to be able to take those exams in a timely way because of the national shortage of pediatric optometrists and ophthalmologists, thus creating more bottlenecks in care and possibly delayed entry to school.

A committee member posed a question about how to bridge the gap between the desire for federal regulation to achieve uniformity across states and the need to tailor implementation to local communities. Collins and Fishman emphasized that translating local enthusiasm into action is difficult without funding; federal funds and infrastructure can support states in the local implementation of programs based on evidence-based practices. Mandle suggested that a committee of state and local representatives could help address conflicts that arise. Fishman said that a federal coordinating body could help implement policies and increase data surveillance.

As the workshop drew to a close, Young explained that refractive error has been a “vexing problem” for centuries; prior to COVID-19, it had been estimated that 50% of the world's population would be myopic by 2050. She pondered the unintended consequences of today's more urban, more industrialized, and less agrarian societies with increased indoor and educational demands that require more near work. As tablet and smartphone use becomes the norm for children, she urged the global community to develop strategies to provide eye exams and continued surveillance for all. She proposed that standardized vision screening begin at age 3 instead of at age 6 and noted that further research could help predict which children will progress and might benefit from different types of interventions. She ended the workshop with a quote from poet Amanda Gorman: “For there is always light, if only we're brave enough to see it. If only we're brave enough to be it.”