Epidemiology and world-wide impact of visual impairment in children

Introduction

This chapter first presents a number of key considerations and issues relevant to understanding and applying evidence from epidemiological studies of childhood visual impairment (VI), severe visual impairment (SVI), or blindness ( Box 1.1 and Box 1.2 ) – collectively “VI” for brevity. Thereafter, the impact of childhood VI is considered from different perspectives. This is followed by consideration of childhood VI in the broader context of child health and childhood visual disability. A synthesis of current epidemiological data, grouped by the GBD super regions, is presented as the key data required for planning clinical services, and for policies. Finally, an overview is presented of primary, secondary and tertiary prevention strategies with consideration of the role of ophthalmic professionals. Areas where critical evidence is lacking are highlighted throughout and potential future directions for research are discussed. Readers are referred to the online extensive supplementary material for a full bibliography.

Specific Issues in the Epidemiological Study of Visual Impairment in Childhood

• Case definition: A standard definition applicable to all children remains challenging, see below “Who is a visually impaired child?” .

• Rarity: As childhood VI is uncommon, large-scale studies are required to achieve representative samples of affected children for precise and unbiased analysis.

• Complex, multidisciplinary management: For a complete picture, information must be sought from all professionals involved in the care of VI or blind children, as many, in some settings the majority, have additional significant non-ophthalmic impairments or chronic disorders.

• Life course approach: Within child health, life course approaches are now common, to understand the complex interplay between biological, environmental, and lifestyle/social influences at all life stages (preconceptional, prenatal, perinatal, and childhood), and how they combine to set and change health trajectories into adult life. Life course epidemiological and epigenetics approaches are increasingly applied to the study of VI and eye disease affecting children or originating in childhood.

• Developmental perspectives: In all research on children, developmental issues (as distinct from age-related issues per se ) must be taken into account in assessing outcomes and their relationship with risk factors.

• Long-term outcomes: Assessment of meaningful outcomes, such as final visual function or educational placement, requires long-term follow-up, into adult life for some outcomes. This is challenging and is increasingly addressed through data science and health informatics approaches. These use routinely collected data, often as electronic or “e” records, in health care (e.g. personal e health records, EHR/EM) and other health or education or welfare administrative systems with record (data) linkage using established methods to minimize errors to create complete datasets for analysis.

• Ethics: Issues of proxy consent (by parents) and children's autonomy increasingly impact participation in ophthalmic epidemiological research.

Framing the Question

It is useful to think about the evidence needed to make clinical, service provision or policy decisions using a “four-part question” based on the PICO mnemonic P opulation I ntervention/Indicator C omparator and O utcomes). A good question incorporates the reference population (e.g. children under 2 years with infantile esotropia), the risk factor or the intervention and, where appropriate, a comparator (e.g. preterm versus term birth, or strabismus surgery versus no surgery), and the outcomes (e.g. parent-reported improvement in cosmesis and objective improvement in alignment and stereopsis).

The importance of framing the question well lies in the fact that the question will determine the type of study (study design) that can answer it, e.g. a descriptive, cross-sectional prevalence survey, or an analytical study of which of the two broad types are “observational,” e.g. case–control or cohort studies, or “interventional,” e.g. randomized controlled trials.

Potential Sources of Information About Visual Impairment

There are a number of sources of epidemiological information about childhood VI or blindness but, in reality, only a few are available in most countries. This explains the currently incomplete picture (see Box 1.3 ).

• Population-based prevalence studies: These represent a source of precise, representative estimates of burden (frequency) and causes. However, the few studies of whole populations of children with VI, such as prior and upcoming national birth cohort studies in the UK, need to be very large (a study of 100,000 children is required in an industrialized country to identify 100–200 children with VI or blindness): these are costly and difficult to do.

• Population-based incidence studies : Studies of all-cause incident (newly occurring) VI are even more difficult, explaining their rarity.

• Special needs/disability registers and surveillance: Specific studies and/or surveillance systems or registers of childhood disability can provide information about VI, but it is important to recognize the potential for bias, as certain visually impaired children may be over-represented in these sources, e.g. those with multiple impairment.

• Studies of intervention/service-based populations: In developing countries, studies of children in special education provide information on causes, but these may be biased because many affected children (particularly those with additional non-ophthalmic impairments) do not have equal access to special education. With other service-based studies, e.g. from clinic attendees, the interpretation of findings and their extrapolation to other populations needs to take these biases into account, as children with treatable conditions are likely to be over-represented.

• Visual impairment registers: These exist in many industrialized countries but, if registration is voluntary and not essential for accessing special educational or social services, registers may be incomplete as well as biased, as they will reflect differences in parental preferences and professionals' practices regarding registration of eligible children.

• Visual impairment teams: Increasingly, children in industrialized countries are evaluated by multidisciplinary teams. They can provide useful information if these teams serve geographically defined populations.

• Disorder-specific ophthalmic surveillance schemes: Uncommon eye conditions in children can be studied using population-based surveillance schemes, and have enabled studies, for example, of congenital eye anomalies, congenital glaucoma and adverse drug reactions. Under-ascertainment can occur. In the United Kingdom, the national active surveillance scheme includes all senior ophthalmologists (the British Ophthalmological Surveillance Unit). This unit facilitates studies of uncommon disorders, including the first population-based incidence study of SVI and blindness in childhood and subsequently the first study of VI, SVI, and blindness. This is an important resource for pediatric ophthalmic epidemiological research in the UK and a potential model for other settings.

• Community-based rehabilitation programs: In many developing countries, rehabilitation of blind and VI children occurs within the community. If the size of the catchment population is known, it is possible to estimate prevalence and obtain population-based data on causes.

• Case ascertainment using key informants: In many developing countries, it may be possible to identify key community and religious leaders, healthcare workers, and others who know their communities well and thus can identify children believed to have VI or ocular disorders. This can be combined with the size of the population at risk, to estimate prevalence and provide population-based data on the causes.

• Household surveys are commonly used in many low-income countries to collect data on a range of health indices. This approach can also be used to identify children who are blind.

• Data modeling can also be used for specific conditions, as has been used to estimate the global incidence of blindness and visual impairment from retinopathy of prematurity.

Regardless of the sources, ascertainment is often incomplete and/or biased. For example, in industrialized countries, families from socially disadvantaged groups or ethnic minorities are less likely to participate in research on health services for visually impaired children. Participation/selection bias affects our ability to generalize findings, especially in research on rare disorders. Also, in some communities in low-income countries, having a disabled child is a source of stigma, which can lead to under-ascertainment in community-based key informant or household surveys. Using multiple sources generally provides a more complete and reliable picture of causes and frequency of childhood VI.

Impact of Visual Impairment

Visual impairment in childhood impacts on all aspects of the child's development and shapes the adult he/she becomes, influencing health, well-being and education, particularly in low-income countries, and employment, social prospects, and lifelong opportunities. Visual impairment also impacts the families of affected children, both in terms of their own general health, mental health and well-being, and on their resources. Although the prevalence and incidence of VI are lower in children than in adults, the years of life lived with VI (“person-years of visual impairment”) are considerable. Personal and social costs are important, but difficult to measure (see Box 1.3 ). The economic costs of childhood VI in terms of loss of economic productivity are significant, a quarter of the costs of adult blindness in some countries. For example, in India an estimated annual cumulative loss of gross national product attributable to childhood VI was US$22 billion.