Prevalence of vision conditions in children in a very remote Australian community

ABSTRACT Clinical relevance Understanding the prevalence of vision conditions in a population is critical for determining the most appropriate strategies for detecting and correcting eye conditions in a community. This is particularly important in very remote regions where access to vision testing services is limited. Background Although recent studies have provided detailed analyses of the prevalence of vision conditions in Aboriginal and/or Torres Strait Islander children in urban and regional areas of Australia, there is a paucity of research examining vision conditions in children in remote regions. Importantly, a significant proportion of the population in remote and very remote regions identify as Aboriginal and/or Torres Strait Islander people. Methods Comprehensive eye examinations were provided to 193 primary school children in a very remote Australian region. Ninety eight percent of children identified as Aboriginal and/or Torres Strait Islander. The eye examination included measures of visual acuity, cycloplegic autorefraction, binocular vision and accommodative function, ocular health and colour vision. Previous history of eye examinations and refractive correction were assessed through parental questionnaire. Results Although the average unaided vision in the population was good (mean: 0.02 ± 0.13 logMAR) and the prevalence of reduced unaided visual acuity (>0.3 logMAR in either eye) was low (4%), vision conditions were detected in 32% of children. The most common conditions were clinically significant refractive errors (18% of children) and binocular vision or accommodative disorders (16%). Of the total population of children tested, 10% had previously had an eye examination, and 2% were reported to have previously been prescribed spectacles. Conclusions In this population of children in a very remote Australian region, up to 1 in 3 children had a vision condition, with many of these conditions being uncorrected and undetected. These findings highlight the important need for additional resources to be made available to very remote communities for the detection and correction of vision conditions in childhood.


Introduction
Good vision in childhood is important, since vision plays a key role in facilitating children's educational outcomes, [1][2][3][4] and poor vision and eye conditions can negatively impact upon quality of life, confidence, and self-esteem. [5][6][7] Worldwide, the most common vision conditions present in childhood are refractive errors. [8][9][10] If uncorrected, refractive errors can negatively impact on academic outcomes and learning-related tasks such as reading performance. 11 Other vision conditions, such as binocular vision and accommodative dysfunction are also common in childhood and have been linked to reduced academic performance. 12 Understanding the prevalence of vision conditions in a population is critical for determining the most appropriate strategies for the provision of eyecare services in a community. This is particularly important in remote (defined as a region with very restricted accessibility by road of goods, services and opportunities for social interaction) 13 and very remote (a region with very little accessibility of goods, services and opportunities for social interaction) 13 Australian regions where access to vision testing services is typically limited. A significant proportion of the population in remote (20%) and very remote (47%) regions identify as Aboriginal and/or Torres Strait Islander people, as compared to 3.3% of the overall population. 14 Keel et al., 15 reported that Aboriginal and/or Torres Strait Islander adults participating in the National Eye Health Survey and residing in very remote regions, were significantly more likely to require referral for further eyecare (e.g. due to the presence of untreated eye conditions) compared to adults residing in less remote regions. This finding was attributed to the barriers to access and limited availability of eyecare services in very remote regions.
The prevalence of vision conditions in Aboriginal and/or Torres Strait Islander adults has been documented in studies across urban, regional, remote and very remote communities, with Aboriginal and/or Torres Strait Islander adults found to have a significantly higher prevalence of preventable/correctable vision loss compared to non-Indigenous Australian adults. [16][17][18] However, the vision of Aboriginal and/or Torres Strait Islander children has been less well studied, with a paucity of studies examining the prevalence of vision conditions affecting Aboriginal and/or Torres Strait Islander children in remote and very remote regions.
In 2010, the prevalence and causes of vision loss of 1,694 Aboriginal and/or Torres Strait Islander children, sampled from urban, regional, remote and very remote regions was assessed as part of the National Indigenous Eye Health survey. 17 A low prevalence of vision impairment (1.5%) was reported, with the major cause of vision loss being refractive error. However, this study was limited by the fact that refractive error was only measured in those children with unaided vision worse than 0.3 logMAR and without cycloplegia, which may have underestimated the prevalence of refractive error.
More recently, two studies have provided a more comprehensive assessment of the visual status, refractive error (cycloplegic) and binocular vision and accommodative function of Aboriginal and/or Torres Strait Islander children in urban and regional areas of Queensland, Australia. 19,20 Both studies reported low rates of vision impairment (VA > 0.3logMAR) in Aboriginal and/or Torres Strait Islander children, but a significant proportion of children were found with refractive error and binocular vision or accommodative dysfunction (e.g. convergence insufficiency, lag of accommodation). A high proportion of vision conditions were reported to be uncorrected, and Aboriginal and/or Torres Strait Islander children were found to be significantly less likely to have previously had a vision examination compared to non-Indigenous children. 19,20 Although these recent studies suggest a significant prevalence of vision conditions in Aboriginal and/or Torres Strait Islander children, neither included children in remote or very remote regions. Understanding the prevalence of vision conditions in Aboriginal and/or Torres Strait Islander children in more remote regions is important, given that there is limited availability of eyecare services in these regions, 21 and that 47% of people residing in very remote areas identify as Aboriginal and/or Torres Strait Islander people. 14 This study aimed to determine the prevalence of vision conditions in Aboriginal and/or Torres Strait Islander children living in a very remote community, to provide evidence to inform the design of improved vision testing service models for remote regions.

Methods
This was a mixed methods study that involved the provision of comprehensive eye examinations to children in a very remote region to characterise vision and eye health status, as well as a secondary qualitative study component that aimed to understand the current children's eye health services in the region and the barriers to accessing these services, through interviews conducted with local service providers in the community. All children attending two primary schools in a very remote region in Queensland (n = 395) were invited to participate in the eye examinations. Letters regarding the study were sent home from school to the families of all children, and the study was further promoted to the community through notices in school newsletters, flyers placed around the community and through the local radio station. Ninety eight percent of children enrolled at the schools identify as Aboriginal and/or Torres Strait Islander. These schools are the only two primary schools in the region, and according to the Accessibility/Remoteness Index of Australia, the region is classified as being very remote. 13 This region in Far North Queensland has a population of 3,200 people (population density 1.9/km 2 ) and is located more than 10 hours' drive from the nearest city. Healthcare services in the region are provided by a hospital and 4 primary health services. There are no permanent optometry or ophthalmology services, however outreach optometry services visit the region and ophthalmology services visit a nearby region. The region is ranked in the lowest decile of the Australian Bureau of Statistics Socio-Economic Indexes for Areas.
Consultation with the school community, local community-controlled health services and members of the local Regional Council informed the development of this project. The study was conducted in accordance with the Tenets of the Declaration of Helsinki and was approved by the Queensland University of Technology Human Research Ethics Committee, and Queensland Government Department of Education. The parent or guardian of each child provided written informed consent, and all children provided written assent, prior to participation in the study.
Children participating in the study were provided a comprehensive eye examination, conducted at the school by experienced research optometrists. Parents or guardians completed a questionnaire outlining their child's Indigenous status (whether they identified as Aboriginal, or Torres Strait Islander or Aboriginal and Torres Strait Islander) as well as their child's previous ocular history, including previous eye examinations, prescription of glasses, and past ocular surgery or treatments.
The eye examination included a series of tests assessing distance and near visual acuity, binocular vision and accommodation and refractive error. The criteria used to define the presence of a significant vision condition was consistent with that used by Cox et al. 20 Unaided monocular distance visual acuity was assessed in both eyes using a logMAR chart at 4 metres. 22 A Lea symbols logMAR chart was used at 3 metres for children who could not be tested with a letter chart. Near visual acuity was assessed monocularly in each eye using a Lea symbols logMAR near chart at 40 cm. Impaired distance and near vision were defined as habitual visual acuity worse than 0.3 logMAR in either eye. 8,9,17 Cover testing was performed at distance (4 m) and near (40 cm) to detect strabismus. Howell Dwyer phoria cards were used to assess horizontal heterophoria. A significant distance heterophoria was defined as exophoria or esophoria ≥5 Δ, and near heterophoria as esophoria ≥6 Δ or exophoria ≥10 Δ. 23 Near point of convergence was assessed with an accommodative target, and a significantly receded near point of convergence was defined as a break point of ≥7.5 cm. 24 Monocular estimation method retinoscopy was used to assess the accuracy of the accommodative response, 25 with children viewing developmentally appropriate words or symbols/pictures (for pre-literate children) of approximately 6/12 size, on paediatric near retinoscopy cards attached to the retinoscope. A significant lag of accommodation was defined as ≥ +1.50 D.
Ocular health was investigated through assessment of pupil reactions, ocular motility, external ocular examination and fundus imaging. Colour vision was assessed using the Ishihara test (24 plate edition), with a colour vision deficiency being defined as three or more errors.
Cycloplegic autorefraction was performed using the Nidek Retinomax handheld autorefractor, 30 minutes after the instillation of 1 drop of 1% tropicamide. Cycloplegia was confirmed when there was minimal pupil response to light. Clinically significant refractive error was defined using the criteria of French et al. 26 Clinically significant myopia was defined as a spherical equivalent refraction (SER) of ≤ −0.50D, clinically significant hyperopia as SER of ≥ +2.00 D, and clinically significant astigmatism as cylinder refraction ≥1.00 D. Clinically significant anisometropia was defined as an interocular difference in SER of ≥1.00 D (calculated as the absolute difference between the SER of each eye).
Spherical refractive errors were further categorised as mild myopia (SER ≤ −0.50D and > −3.00 D), moderate myopia (SER ≤ −3.00 D and > −6.00 D), high myopia (SER ≤ −6.00 D), mild hyperopia (SER ≥ +0.50 D and < +2.00 D), moderate hyperopia (SER ≥ +2.00 D and < +6.00 D) or high hyperopia (SER ≥ +6.00 D). The astigmatic refractive errors were further categorised as mild astigmatism (≥0.50 D and <1.00D), moderate astigmatism (≥1.00 D and <1.50 D) or high astigmatism (≥1.50 D). Astigmatic refractive errors were converted into power vectors 27 to allow further analysis, where J0 represents with-the-rule (WTR, positive values) or against the rule (ATR, negative values), and J45 represents oblique astigmatism. Refractive errors were categorised based upon the refractive error present in one or both eyes. For cases where a different category of refractive error was present in each eye, the eye with the more clinically significant refractive error was used to classify that participant. For example, if a participant had mild hyperopia in one eye and moderate hyperopia in the other eye, they were classified as having moderate hyperopia.
A report of the vision testing outcomes for each child was provided to the parent or guardian. Appropriate clinical management was also provided, including provision of spectacles where these were considered to be clinically necessary and referral for further examination with visiting optometry or ophthalmology services if required through local primary health care providers.
A series of semi-structured, in-depth interviews were conducted with local service providers involved in primary health care, education, and criminal justice. The interviews were included to provide a social context to service delivery in the region and to identify possible barriers to detecting and correcting eye conditions in the region. Participants were selected purposively on the basis of their ability to provide information rich data. Recruitment was facilitated through a snowball sampling method, including participants who had expressed an interest in the research or were referred to the research team because of their professional and community experience.
All participants involved in the interview component of the study provided written informed consent. Interview questions examined the local community context for service delivery, health and health care in the region, and barriers to service delivery (see supplementary material). All interviews were conducted by a trained and experienced researcher who had worked with this community previously. While a high level of saturation was achieved through the data, the number of interviews was limited by the relatively short period of time (three days) budgeted and scheduled for the qualitative component of the project. All interview data were transcribed, coded and thematically analysed, 28 to distil indepth information from the data. 29 Each transcription was analysed in turn for passages and/or content of a type which referenced issues, experiences and opinions that might inform relevant research elements. Recurring themes were established by coding transcripts and regrouping into conceptual constructs. In this way, a series of analytic codes were developed. 30

Results
One hundred and ninety-three children aged between 5 and 11 years (mean age 8.4 ± 1.8 years, 51% girls) provided consent to participate in the study and had eye examinations performed. This represented 49% of the 395 students enrolled at the two schools. Ninety seven percent (n = 187) of the children identified as Aboriginal and/or Torres Strait Islander, with 4% (n = 7) identifying as Aboriginal, 30% (n = 58) identifying as Torres Strait Islander and 63% (n = 122) identifying as Aboriginal and Torres Strait Islander.
Of the children tested, 10% (95% CI: 6%-15%) (n = 18) were reported by their parent/guardian to have previously had an eye examination (vision screening or comprehensive vision test with an optometrist or ophthalmologist). Two percent (95% CI: 0%-5%) (n = 3) of children were reported to have previously been prescribed spectacles, however only one child currently had spectacles. Of those reporting a previous eye examination, the majority reported having previously had a vision screening (n = 12) and few reported having had an eye examination with an optometrist or ophthalmologist (n = 6).

Visual acuity
The mean unaided monocular distance visual acuity was 0.01 ± 0.13 for the right eye and 0.02 ± 0.14 logMAR for the left eye (range: −0.16 to 1.34 logMAR). The mean unaided near visual acuity was similar at 0.00 ± 0.17 logMAR for the right eye and −0.01 ± 0.11 logMAR for the left eye (range −0.12 to 1.32). No children exhibited bilateral vision impairment. Four percent (n = 7) of children exhibited monocular distance vision impairment, and 4% (n = 7) exhibited monocular near vision impairment (visual acuity >0.3 logMAR). Monocular distance vision impairment that was not correctable with spectacles was evident in 3% (n = 6) of children. The causes of monocular vision impairment in these cases were either amblyopia (due to high refractive error, n = 3), or associated with past ocular trauma (n = 3). Table 1 and Figure 1 illustrates the distribution of spherical equivalent refractive error in the children tested. The majority of children (n = 164) were classified as being either emmetropic in both eyes, or as having mild hyperopia in one or both eyes. A smaller number of children had mild myopia (n = 23), moderate hyperopia (n = 4) or high hyperopia (n = 2) in one or both eyes. No children exhibited moderate or high myopia.

Cycloplegic refractive error
The distribution of astigmatic refractive errors is shown in Table 2 and Figure 2. Most children exhibited no astigmatism or mild astigmatism (n = 185), with a smaller proportion of children with moderate (n = 3) and high (n = 5) astigmatism in one or both eyes (Figure 2). Of those children with mild and greater levels of astigmatism, 41% had oblique, 31% had ATR and 28% had WTR astigmatism. As highlighted in Figure 2A, most children with moderate and high astigmatism exhibited WTR axes (77%).

Binocular vision and accommodation
Cover testing revealed one child with strabismus. Table 3 presents the mean outcomes from the binocular vision and accommodation testing. Thirty-one children (16%, 95% CI: 11%-22%) had a clinically significant binocular vision or accommodation issue. The most common issue was a significant lag of accommodation (≥1.50D), present in 20 children (10%, 95% CI: 6%-16%). Two percent (95% CI: 1-5%) of children had a clinically significant distance phoria (exophoria or esophoria ≥5 Δ), 4% (95% CI: 2%-7%) had a significant near phoria (exophoria ≥10 Δ or esophoria ≥6 Δ), and 2% (95% CI: 1%-5%) had a significantly receded near point of convergence (≥7.5 cm break point). Of those children Table 1. Mean ± SD spherical equivalent cycloplegic refractive error outcomes (top) and classification of spherical equivalent refractive errors (bottom). Anisometropic refractive error was calculated as the absolute difference between the spherical equivalent refraction of the two eyes.   with a clinically significant binocular vision or accommodation issue, 26% (n = 8) had previously had an eye examination. Ocular health examination revealed three children with clinically significant unilateral eye health conditions, all related to previous trauma (retinal degeneration, cataract and corneal scarring). Two of these children were reported to have previously had an eye examination. Seven children failed the Ishihara screening test for colour vision deficiency (6 boys and 1 girl).
Considering the vision testing outcomes for all children, 32% (95% CI: 25%-39%) of children (n = 61) were identified with vision conditions. The most common conditions were clinically significant refractive errors, which were detected in approximately 1 in 5 children. Of the 61 children identified with vision conditions, 20% (95% CI: 11%-32%) (n = 12) were reported by their parent/guardian to have previously had a vision test.

Community interviews
Interviews were conducted with eight local service providers. All participants had worked in the region for several years and five had always lived in the region. Interviews ranged in length from 20-50 minutes. Interviews were conducted with seven female and one male participant, of whom five were Aboriginal and/or Torres Strait Islander.
Interviews with service providers suggested that eye health conditions primarily affecting adults, and associated with systemic conditions such as diabetes, were regarded as a priority in the region. Several service providers cited a lack of local capacity to provide primary health care education and health promotion, and a lack of capacity for detailed vision screening. This, together with a high turnover of professional staff in the region, were cited as barriers to service provision. Several participants also reported that while local people were supportive of vision testing for children, that the limited access to comprehensive vision testing meant these services were often prioritised for older people with known eye conditions. Nonetheless, eye and ear conditions among children were considered by local service providers to have significant impacts on children's social engagement.

Discussion
This study provides a detailed overview of the prevalence of vision conditions amongst children residing in a very remote Australian community, of whom 98% identified as Aboriginal and/or Torres Strait Islander. Using a vision testing protocol including cycloplegic autorefraction and detailed accommodation and binocular vision testing, a significant proportion of children (32%) had a vision condition, with the most common conditions being refractive error or binocular vision and accommodation dysfunction. In a previous study conducted in a regional Australian area and employing a similar testing protocol, Cox et al 20 reported vision conditions in 29% of Aboriginal and/or Torres Strait Islander children, which is consistent with the current findings in a very remote region.  In this population, the mean unaided vision was good (0.0 logMAR) and prevalence of vision impairment was low (4%), consistent with previous studies of Aboriginal and/or Torres Strait Islander children, that have estimated the prevalence of vision impairment (>0.3 logMAR) to range from 1% to 3%. 17,19,20 Of note, many of the vision conditions identified (e.g. mild myopia, astigmatism, hyperopia and increased accommodative lag) can cause significant functional vision difficulties and have been linked in previous studies to reduced academic performance, 11 but may not result in vision impairment >0.3 logMAR.
Another important finding was that only 10% of children were reported to have previously had an eye examination and only 2% had been previously prescribed glasses, indicating that a large proportion of the vision conditions were previously undetected and uncorrected. Previous studies of Aboriginal and/or Torres Strait Islander children in regional Australia have reported relatively low rates of previous vision tests (23%) and prescription of refractive corrections (8%). 20 The lower rates found in the current study, likely reflects the limited availability and access to eyecare for children in this very remote region. Of those children reporting previous vision tests, the majority were reported to have had a vision screening either at school or their primary health care provider, with only a small proportion reported to have had eye examinations with an optometrist or ophthalmologist, underscoring the limited availability of comprehensive eyecare in this very remote region.
The prevalence of vision conditions will depend on the population tested, the testing procedures used, and criteria to define the presence of vision conditions, which makes direct comparisons with other studies challenging. However, previous studies in urban and regional areas of Australia, New Zealand, North America and Europe, employing a variety of testing protocols, have estimated the prevalence of vision conditions to range from 20 to 35% in school-aged populations. 20,[31][32][33][34][35] This suggests the overall prevalence of vision conditions observed in this very remote region is consistent with children in other locations in Australia, New Zealand and some other European and North American regions.
In the current study, 18% of children exhibited clinically significant refractive error, with the most common refractive error being myopia. Although the prevalence of myopia in this study appeared slightly higher than past studies of Aboriginal and/or Torres Strait Islander children, 19,20 all children exhibited mild myopia, and the prevalence of myopia was lower than reported in population-based studies of non-Indigenous children in urban locations in Australia. 26 However, given the documented increase in myopia prevalence globally and likely progression of myopia over time, these findings suggest that further study of children in remote communities is warranted, including longitudinal follow up to better understand the development and progression of myopic refractive error in these communities.
Binocular vision and accommodation dysfunction was detected in 16% of children, with an increased lag of accommodation being the most common issue, which is consistent with another study of Aboriginal and/or Torres Strait Islander children in regional Queensland. 20 Previous studies have reported a significant positive correlation between hyperopic refractive errors and lag of accommodation. 36 The accommodative lag present in the current population is consistent with the finding of more than 40% of the children exhibiting mild or moderate hyperopia. Recent studies suggest that correction of low to moderate hyperopia results in improved accommodative performance and reading speed during sustained near tasks, 37 supporting the potential clinical benefit of the correction in some cases.
The relatively high prevalence of vision conditions, coupled with the low rate of previous vision tests reported in the current study, supports the need for additional strategies and resources to improve the availability and access to comprehensive eyecare for children in this very remote region. Given the remoteness of the location, permanent comprehensive vision testing services in very remote regions may not be feasible. 21 Service provision is also impacted by staff turnover which was reported in the interview component of this study to be high, including not only health services, but also in areas such as education. This can directly impact service coordination activities, particularly referral, and can more broadly impact local community integration, including cultural competency.
Increased visiting eyecare services and expanded vision screening programs in the community to detect children with clinically significant vision conditions may be required. To ensure sustainability, cultural safety and optimum accessibility of services, any future expansion or development of eyecare service models should be community driven. 38 Although further work is required to determine the optimum vision screening protocol for children in remote communities, the fact that many of the detected vision conditions in the current study do not result in substantive impairment of distance vision, suggests that tests of near visual function and binocularity may be required for effective screening performance.
While a strength of this study is the detailed battery of vision tests performed on all children including cycloplegic refraction, a limitation is the relatively small sample size of children. There is some potential that this relatively small sample size could potentially skew results, as children already under optometric or ophthalmological care may have been less likely to choose to participate. However, the sample represented approximately half of the population of primary school children within the region which is a greater or equivalent response rate to a number of previous studies of vision conducted in Australian schools. 20,39,40 The data were collected in a single location, so may not be generalisable to all remote regions. An additional limitation is the reliance on a parental questionnaire to determine the rate of previous eye examinations, which could result in an underestimation of these rates.
In conclusion, the current study reports that up to one in three children in a very remote region of Australia exhibit a vision condition, with many being uncorrected and undetected. These findings highlight the important need for additional resources to very remote communities for the detection and correction of vision conditions in childhood. The development of improved community-driven eye care service models is required to ensure good vision and eye health for Aboriginal and/or Torres Strait Islander children living in very remote communities.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Funding
The work was supported by the Queensland University of Technology [Health-Law Research Fund Grant]