Trends in the prevalence of common ocular conditions and comparison of ophthalmic outpatient utilisation related to these conditions in children with and without various types of disabilities: analysis of nationwide population-based data from Taiwan, 2014–2019

Abstract

Objective

The study objectives were to investigate trends in the prevalence of common ocular conditions among children with and without disabilities; to compare the prevalence of these conditions in children with various disabilities; and to compare ophthalmic outpatient utilisation related to these ocular conditions in children with and without disabilities.

Design

Repeated cross-sectional nationwide population-based study.

Setting

Nationwide analysis in Taiwan based on National Health Insurance (NHI) claims data and the National Disability Registry, from 2014 to 2019.

Participants

All children (aged under 18 years) with a disability in any given year between 2014 and 2019 were included in our analysis. All children with a disability (experimental group) were matched 1:1 with a child of the same age without a disability (control group). Data regarding the children’s disability type and status and ocular conditions were obtained from the National Disability Registry and NHI database of Taiwan.

Outcome measures

(1) The prevalence of myopia, strabismus, astigmatism, amblyopia and hyperopia over time; (2) the prevalence of myopia, strabismus, astigmatism, amblyopia and hyperopia in children with various disabilities; and (3) the association between disability and the use of outpatient vision care. All outcome measures were assessed using data from 2014 to 2019.

Results

Among children with disability, the prevalence of myopia increased from 15.97% in 2014 to 18.07% in 2019. The prevalence of strabismus (2.06–3.90%), astigmatism (8.25–9.24%), amblyopia (4.13–4.95%) and hyperopia (3.36–4.58%) also increased over the study period in children with disabilities. The prevalence of strabismus, astigmatism, amblyopia and hyperopia was significantly higher in children with disabilities than in those without disabilities in all years. For example, in 2019, the prevalence values for strabismus, astigmatism, amblyopia and hyperopia were 3.90%, 9.24%, 4.95% and 4.58%, respectively, among children with disabilities, and 0.67%, 5.84%, 1.33% and 1.70%, respectively, for those without disabilities. The prevalence of these conditions varied considerably across disability types. For example, in 2019, the prevalence of strabismus was highest in children with visual disabilities (10.66%; p<0.001); these children also exhibited a high prevalence of amblyopia (24.34%; p<0.001). The prevalence of myopia was high in children with autism (24.77%), but the prevalence of other ocular conditions was not elevated in this group. Regression results indicated that for myopia, children with disability had 0.48 fewer outpatient clinic visits than those without disability (p<0.001).

Conclusions

The prevalence of common ocular conditions and the corresponding vision care required vary across types of disabilities among children. Healthcare policies must account for children at high risk of various ocular conditions, including those with less common disabilities.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• This study used national claims data, which prevent selection bias.

• We included all types of disability included in Taiwan’s National Disability Registry in the analysis.

• Disability status was verified by physicians instead of self-reported.

• The prevalence rate was calculated using treated patients, and there are limitations with the information available from claims data.

Introduction

Visual health is crucial for the development of children,^{1 2} and the loss of vision results in major distress.³ Approximately 240 million (per data obtained in 2021) children worldwide have some form of disability,⁴ and children with disabilities tend to exhibit a higher prevalence of ocular problems than those without.¹

Few studies have focused on assessing common ocular conditions among children with various types of disabilities. However, such assessments could usefully promote effective data-driven policymaking aimed at improving the quality of life of individuals with disabilities.⁵ For example, if children with disabilities exhibit a higher prevalence of ocular conditions than children without disabilities, this would justify the allocation of more resources to vision care for this vulnerable group.

Children may differ in their vision care needs depending on their type of disability. For example, fast-paced video games, which place a high demand on the user’s visual acuity or eye-tracking ability, have often been used as an intervention to promote physical activity and reading ability in children with intellectual disabilities and dyslexia.^{6 7} However, this approach may not work for children with other types of disabilities. Children with impaired hearing, for instance, may have a relatively high need for vision care because they rely heavily on vision for communication and development.⁸

A study⁹ conducted in Denmark revealed a high incidence of refractive errors and strabismus among 923 children with developmental delays; of them, 26.8% had strabismus and 10.8% had myopia. Hyperopia is highly prevalent in children with Down’s syndrome.¹⁰ A hospital-based study conducted in Turkey¹¹ reported that 77% of children with intellectual disabilities had nystagmus, strabismus, astigmatism or hypermetropia; the rate among children without intellectual disabilities was only 42.4%. A review study that included individuals with Down’s syndrome revealed a high prevalence of refractive errors, particularly hyperopia and strabismus.¹²

A number of studies have focused on intellectual disabilities or Down’s syndrome,9 10 12–14 but only a few have centred on other types of disabilities. Children with non-intellectual disabilities, such as physical disabilities, often have limited access to healthcare for various reasons, including a lack of social support.¹⁵ Limited access to healthcare can lead to the neglect of vision care, which may ultimately worsen their eye health. Further, children with hearing disabilities are at risk of autism.¹⁶ Providing vision care to children with autism is difficult. Thus, comprehensive vision care programmes that target the unique needs of children with a given disability are urgently required.

Studies on ocular conditions in children with disabilities have often included a limited sample—for example, individuals from only a single centre or school.^{11 17} Thus, the findings that have been obtained do not accurately reflect the real-world prevalence of ocular conditions in children without disabilities and in those with one or multiple types of disabilities. This highlights the need for nationwide studies aimed at determining the relative prevalence of ocular problems in this population.

This study investigated the prevalence of common ocular conditions in children with versus without disabilities over time and compared the prevalence between children with various types of disabilities over time. We also evaluated the association between having a disability (or multiple disabilities) and receiving care for common ocular conditions from an ophthalmologist.

Methods

Study design and data sources

This study was an analysis of repeated cross-sectional data from 2014 to 2019. The National Disability Registry of Taiwan contains the disability-related data of individuals in Taiwan with an official disability certification. In Taiwan, a patient must apply for disability status. The patient is assessed by a physician from a designated medical institution, and upon the physician’s recommendation, the patient can obtain a disability certification. Those with hearing, verbal, limb-related, intellectual, organ-related or facial disabilities; multiple disabilities; dementia; autism; congenital anomalies; metabolic abnormalities; mental disorders; balance disorders; epilepsy; and rare diseases were considered in this study. In Taiwan, individuals with disabilities receive several welfare benefits, such as tax exemptions; therefore, parents or guardians are unlikely to fail to register their children with disabilities. The National Disability Registry provides robust data on all children with disabilities in the country.

Taiwan’s National Health Insurance (NHI) programme is a universal health coverage programme that is compulsory for all of Taiwan’s residents. The NHI database contains the data of all individuals, regardless of whether the individual has ever made a medical claim. All children aged under 18 years with a disability in any given year between 2014 and 2019 were included in our analysis. All children with a disability (experimental group) were matched 1:1 with a child of the same age without a disability (control group). Children in the control group were those not included in the Disability Registry. These children may have had ocular conditions, but the severity thereof was not high enough for them to be certified as having a visual impairment. Disability Registry enrollees can be linked to records in the NHI claims database by using individual-specific anonymous identification numbers.

Outcome variables (NHI data)

The primary outcome variables were the following five ocular conditions diagnosed in accordance with the criteria outlined by the International Classification of Diseases (ICD), 9th Revision (ICD-9) and the ICD, 10th Revision (ICD-10): myopia (ICD-9: 367.1 and 360.21; ICD-10: H52.1* and H44.2*), strabismus (ICD-9: 378*; ICD-10: H49*, H50* and H51*), astigmatism (ICD-9: 367.2*; ICD-10: H52.2*), amblyopia (ICD-9: 368.0; ICD-10: H53.0*) and hyperopia (ICD-9: 3670; ICD-10: H52.0*). In this study, ocular conditions were confirmed on the basis of diagnoses made by ophthalmologists per the relevant ICD codes. The analysis was conducted at the individual level.

To analyse whether children with disability were more likely to visit ophthalmologists for these conditions, we defined another outcome variable: the number of outpatient visits for these conditions.

Statistical analysis

Data on the basic sociodemographic characteristics of children with or without disabilities were assessed. Subsequently, we analysed the crude annual prevalence of each of the ocular conditions using data from 2014 to 2019. The prevalence of the ocular conditions was stratified by disability type. Because various sociodemographic factors influence the aforementioned prevalence, logistic regression was performed, and age, sex, location and insurable family income were controlled for. To account for the possible non-linear effects of the continuous variables (age and income), we included the squared terms for these variables. The probability of each disability was predicted. Because the data for a child spanned multiple years, we accounted for clustered variance by using generalised estimating equation models based on negative binomial and log link functions to estimate the number of ophthalmology visits. For this part of the analysis, data for all years (2014–2019) were used. A dummy variable indicating a specific year was included in the regression. The adjusted marginal effects of disability were calculated for each ocular condition. All statistical analyses were performed using STATA (V.15; StataCorp, College Station, Texas, USA).¹⁸

Patient and public involvement

None.

Results

Table 1 summarises the sociodemographic characteristics of the study population for 2019. Of all children, 1.37% had a disability. Figure 1 presents the prevalence trends by year for the five ocular conditions, stratified by disability status.

Table 1.

Sociodemographic characteristics (2019) of the study sample

Variables	Children with disability			Children without disability
	N=51 586	Mean	SD	N=51 586	Mean	SD	P value
Age (mean/SD)		10.6	4.6		10.6	4.6	1
Sex
Boys	34 023			26 554			<0.001
Girls	17 562			24 720
Missing	1			312
Insurable income (NT$, mean/SD)	51 586	28 606	24 767	51 586	33 356	25 387	<0.001
Insurable income (NT$)
Q1 (mean/SD)	18 435	7913	9916	13 971	11 197	10 875	<0.001
Q2 (mean/SD)	11 866	24 000	2	12 003	24 000	0
Q3 (mean/SD)	10 157	32 032	4653	12 044	32 194	4701
Q4 (mean/SD)	11 128	64 672	24 849	13 568	65 480	26 324
Low-income status (yes)	5196			1466			<0.001
Regions
Taipei	8153			9071			<0.001
Other northern	16 911			16 591
Central	9698			9652
South	13 591			13 301
East and off-island	2858			2255
Missing	375			716

US$1≈NT$30.

NT$, New Taiwan dollar.

Figure 1.

Trends in crude annual prevalence of each of the five common ocular conditions in children with and without disability. Prevalence was calculated as the total number of cases divided by the total number of children in each category. Children with disabilities were matched 1:1 with children of the same age without disabilities.

The prevalence of most ocular conditions was markedly higher in children with disabilities than in those without. In children without disabilities, the prevalence of astigmatism increased from 5.03% in 2014 to 5.84% in 2019. During the same time interval, the prevalence increased from 8.25% to 9.24% in children with disabilities. The prevalence of strabismus almost doubled in children with disabilities; the prevalence was fivefold to sixfold higher in children with disabilities than in those without. The average prevalence of hyperopia in children with disabilities was 3.83% over the study period, whereas it was only 1.32% in children without disabilities. The prevalence of amblyopia was approximately fourfold higher in children with disabilities than in those without.

Table 2 presents the estimated prevalence (2019) of the five ocular conditions stratified by disability type. The estimated prevalence rates for other years (2014–2018) are presented in online supplemental appendix A. In an analysis stratified by disability type, significant differences in the prevalence of all five ocular conditions were observed. Children with certain types of disabilities exhibited a higher prevalence of myopia than those without disabilities, which highlights the importance of considering the type of disability to estimate prevalence. Children with visual disabilities exhibited the highest prevalence of strabismus (10.66%; p<0.001). These children also exhibited a high prevalence of amblyopia (24.34%; p<0.001). Children with autism exhibited high prevalence of myopia (24.77%), but the prevalence of other ocular conditions was not elevated in this group.

Table 2.

Estimated prevalence (2019) of common ocular conditions in children with different types of disabilities†, ‡

Type of disability		Myopia	Astigmatism	Strabismus	Hyperopia	Amblyopia
	Total	Prevalence	Prevalence	Prevalence	Prevalence	Prevalence
Visual	797	20.2**	12.55**	10.66**	7.4**	24.34**
Hearing	2913	20.53**	8.2**	1.58**	4.12**	3.78**
Verbal	1109	20.02**	10.37**	3.34**	4.51**	4.87**
Limb	4118	18.53**	10.93**	6.97**	5.76**	5.42**
Intellectual	15 151	19.07**	9.89**	3.38**	4.78**	5.54**
Multiple	9016	10.68**	8.02**	5.46**	4.77**	4.64**
Organ	2019	23.43**	7.43**	2.03**	2.92**	2.72**
Facial	103	21.36**	14.56	0.97	4.85*	5.83*
Dementia	8	37.5*	0
Autism	8323	24.77**	8.21**	1.89**	3.14**	2.74**
Congenital anomaly	1124	17.53**	12.37*	4.98**	5.69**	7.38**
Mental	230	21.3**	3.48	1.74*	0.87	0.87
Balance abnormality	25	20.0*	20.0**	12		8.00
Epilepsy	215	22.33**	14.42**	3.26*	6.51**	6.05**
Rare diseases	885	12.2**	8.36**	4.97**	6.21**	4.97**
Others	5550	13.7**	9.67**	4.34**	5.09**	5.10**
No disability	3 704 608	22.02**	5.07**	0.64**	1.67**	1.31**

**P<0.001; *p<0.05

†Controlled for age, area and insurable income.

‡Blank cells indicate that the corresponding values could not be estimated because of a small sample size.

Children with certain types of disabilities exhibited a considerably lower prevalence of ocular conditions than that of children without disabilities. For example, in children with multiple disabilities and rare diseases, the estimated prevalence of myopia was 10.68% (p<0.001) and 12.20% (p<0.001), respectively; the prevalence of myopia was markedly lower in these groups than in children without disabilities (22.02%).

Table 3 presents the adjusted marginal effects of disability on the use of outpatient care for ocular conditions. In relation to myopia and amblyopia, children with disabilities visited outpatient clinics less often than those without disabilities. We also excluded children with visual disabilities to analyse whether children with other disabilities visited outpatient clinics less often than those without disabilities (lower part of table 3). The results demonstrated that the utilisation pattern remained similar when children with visual disabilities were excluded.

Table 3.

Adjusted marginal effects of disability on use of outpatient care for ocular conditions

All disabilities
Ocular conditions	Marginal effect (dy/dx)	SE	z	95% CI		P value
Myopia	-0.484	0.013	36.77	0.509	0.458	<0.001
Amblyopia	-0.024	0.007	3.28	0.039	0.010	0.001
Strabismus	0.156	0.003	58.05	0.151	0.161	<0.001
Hyperopia	0.096	0.003	29.44	0.090	0.103	<0.001
Amblyopia	0.172	0.004	39.24	0.164	0.181	<0.001

Excluding visual disability
Ocular conditions	Marginal effect (dy/dx)	SE	z	95% CI		P value
Myopia	-0.474	0.013	35.81	0.500	0.448	<0.001
Amblyopia	-0.023	0.007	3.15	0.038	0.009	0.002
Strabismus	0.151	0.003	56.34	0.146	0.156	<0.001
Hyperopia	0.093	0.003	28.52	0.087	0.100	<0.001
Amblyopia	0.153	0.004	35.15	0.145	0.162	<0.001

SE: Standard error

CI: Confidence interval

Discussion

We investigated the prevalence of five common ocular conditions, namely myopia, strabismus, astigmatism, amblyopia and hyperopia, in children with disabilities. The main findings of this study are as follows. The prevalence of most ocular conditions assessed in our study was considerably higher in children with disabilities than in those without. The prevalence of all five conditions varied markedly between disability types. Children with certain types of disabilities exhibited a higher prevalence of myopia than those without disabilities. In relation to myopia and amblyopia, children with disabilities visited outpatient clinics less often than those without.

For children with certain types of disabilities, the high prevalence of ocular conditions may be due to preterm birth and congenital defects.¹ Ocular conditions may lead to cognitive disorders and delayed development in children. Ocular conditions are associated with the development of autism and intellectual disabilities.¹⁹ This may explain the high prevalence of common ocular conditions in children with certain types of disabilities.

The prevalence of myopia was generally low in children with disabilities. This might be because myopia is generally more neglected than other ocular conditions, probably because it is less likely to be identified by the caregivers of children with disabilities. Nonetheless, this assumption must be validated in subsequent studies.

Unlike the findings of a review study, ours revealed a higher general prevalence of myopia in children with and without disabilities.¹ However, the prevalence in our study was lower than that reported by a survey study conducted by Tsai et al in Taiwan.²⁰ This might be because the NHI database contains information regarding the prevalence of treated conditions. Nevertheless, our findings revealed the relative prevalence of common ocular conditions and the care-seeking behaviours of children with various types of disabilities and those without disabilities.

Children with a hearing disability exhibited high prevalence of myopia. Children with hearing disabilities require sharp vision for effective communication and education (sign language, lip reading, facial gestures and reading).²¹ Consequently, it is crucial for policies to prioritise and make efforts to improve the vision of children with hearing disabilities.

We observed that for strabismus, hyperopia and amblyopia, children with disabilities had higher utilisation of ophthalmic care than children without disabilities. This finding held when we excluded children with visual disabilities. This may reflect the protective effect of the NHI system. However, children with disabilities may have unmet vision care needs; NHI only mitigates financial barriers to care, but children with disability may face other barriers to medical access, such as difficulties navigating services or a lack of care coordination.²²

The present study has some limitations. First, although a markedly high prevalence of common ocular conditions was noted in children with disabilities, this prevalence might have been underestimated because of under-reported medical information.²³ Nevertheless, because the public support system in Taiwan for children with disabilities is well developed and medical care is easily accessible through the NHI system, our data were highly likely to include all children with disabilities in Taiwan. This highlights an advantage of analysing prevalence by using data obtained from a universal healthcare system. In addition, we did not rely on self-report disability status—recall and reporting biases may be unavoidable in analyses performed using data collected through questionnaire surveys.^{24 25} Second, we defined an outpatient visit for a given ocular condition as a visit during which the patient received a diagnosis of that particular condition. However, the children might have visited outpatient clinics for another ocular condition. For example, if a child who had received a diagnosis of myopia made a visit for allergic conjunctivitis, the visit may contain both diagnostic codes. Third, although this was a nationwide study, for certain disabilities, the sample size was small. Fourth, the claims data have certain limitations. For example, the claims database does not disclose the basis on which diagnoses were made; we were thus unable to check whether all ophthalmologists made a diagnosis by using the same criteria. The database also does not have information on perceived support or the education level of caregivers. Healthcare utilisation may be affected by these characteristics. Finally, we cannot completely ignore the possibility of having missed children with disabilities because of non-registration (probably because of stigma) with the National Disability Registry.

Conclusions

The prevalence of the ocular conditions of strabismus, astigmatism, amblyopia and hyperopia was markedly higher among children with disabilities than among those without. The vision care needs and family and caregiving background of children may differ depending on their type of disability.²⁶ Healthcare policies must account for children at high risk of various ocular conditions, including those with less common disabilities.

Data availability statement

Data may be obtained from a third party and are not publicly available. The data can be accessed upon approval from the Ministry of Health and Welfare, Taiwan.

Ethics statements

Patient consent for publication

Not required.

Ethics approval

This study was approved by the Institutional Review Board (IRB) of National Yang Ming Chiao Tung University (IRB number: YM109014E).