Global and regional estimates of prevalence of refractive errors: Systematic review and meta-analysis

Methodological quality of the review: Medium confidence

Author: Hassan Hashemi, Akbar Fotouhi, Abbasali Yekta, Reza Pakzad, Hadi Ostadimoghaddam, Mehdi Khabazkhoob

Region: Worldwide

Sector: Refractive errors

Subsector: Epidemiology of refractive errors

Equity focus: No

Study population: All ages

Review type: Other review

Quantitative synthesis method: Systematic review and meta-analysis

Qualitative synthesis method: Not applicable

Background: Refractive errors are the most common ocular problem affecting all age groups. They are considered a public health challenge. Recent studies and WHO reports indicate that refractive errors are the first cause of visual impairment and the second cause of visual loss worldwide, as 43 per cent of visual impairments are attributed to refractive errors. Refractive errors also affect the economy of different societies. Although recent studies suggest an increase in the prevalence of myopia due to lifestyle changes, differences in ethnic groups, measurement methods, definitions of refractive errors, and age groups of the participants hinder a definite conclusion regarding the pattern of the distribution of refractive errors worldwide. The distribution of refractive errors is not equal in different countries. A high prevalence of myopia in East Asian countries is a common finding in most previous studies. However, there are some controversies regarding hyperopia. Although some studies have shown a high prevalence of hyperopia in Europe and western countries, it is difficult to make a conclusion since most of these studies were conducted on the elderly, and the high prevalence of hyperopia in this age group is a normal finding due to lens changes.

Objectives: The aim of this study was to evaluate the prevalence of refractive errors across the world in this meta-analysis.

Main findings: A total of 163 articles were used for the final analysis. However, the number of articles was different for the meta-analysis of myopia, hyperopia and astigmatism. In children, the EPP of myopia, hyperopia and astigmatism was 11.7% (95% CI: 10.5-13.0), 4.6% (95% CI: 3.9-5.2), and 14.9% (95% CI: 12.7-17.1), respectively. The EPP of myopia ranged from 4.9% (95% CI: 1.6-8.1) in South-East Asia to 18.2% (95% CI: 10.9-25.5) in the Western Pacific region, the EPP of hyperopia ranged from 2.2% (95% CI: 1.2-3.3) in South-East Asia to 14.3% (95% CI: 13.4-15.2) in the Americas, and the EPP of astigmatism ranged from 9.8% in South-East Asia to 27.2% in the Americas. In adults, the EPP of myopia, hyperopia and astigmatism was 26.5% (95% CI: 23.4-29.6), 30.9% (95% CI: 26.2-35.6), and 40.4% (95% CI: 34.3-46.6), respectively. The EPP of myopia ranged from 16.2% (95% CI: 15.6-16.8) in the Americas to 32.9% (95% CI: 25.1-40.7) in South-East Asia, the EPP of hyperopia ranged from 23.1% (95% CI: 6.1%-40.2%) in Europe to 38.6% (95% CI: 22.4-54.8) in Africa and 37.2% (95% CI: 25.3-49) in the Americas, and the EPP of astigmatism ranged from 11.4% (95% CI: 2.1-20.7) in Africa to 45.6% (95% CI: 44.1-47.1) in the Americas and 44.8% (95% CI: 36.6-53.1) in South-East Asia. The results of meta-regression showed that the prevalence of myopia increased from 1993 (10.4%) to 2016 (34.2%) (P=0.097).

Authors note that this study showed that astigmatism was the most common of the refractive errors in children and adults, followed by hyperopia and myopia. The highest prevalence of myopia and astigmatism was seen in South-East Asian adults. The highest prevalence of hyperopia in children and adults was seen in the Americas.

Methodology:

The meta-analysis was conducted according to the Preferred Reporting Item for Systematic Reviews and Meta-Analysis (PRISMA) guidelines. After the literature search, all studies were entered into EndNote X6. Duplicate articles were identified and removed using the duplicates command. Relevant articles were selected in three phases. In phases 1 and 2, the titles and abstracts of the studies were screened, and irrelevant articles were excluded. In phase 3, the full texts of the studies were carefully evaluated. All three phases were conducted by two interviewers independently. It should be noted that the reviewers were blind to the process of article selection. To extract articles from 1990 to 2016 on the prevalence of refractive errors for this meta-analysis, international databases including Medline, Scopus, Web of Sciences, Embase, CABI, CINAHL, DOAJ, and Index Medicus for Eastern Mediterranean Region-IMEMR were searched. The literature was reviewed using a combination of words, like population (children, student, adult, and related MeSH terms), outcome [refractive error, myopia, hyperopia, astigmatism, spherical equivalent (SE), cylinder power], and study design (prevalence, ratio, cross-sectional, survey, descriptive, and epidemiology). In addition, the reference lists of all searched studies and reviews were evaluated to find similar studies. The quality of the selected articles was evaluated by the two reviewers using the STROBE checklist that contains 22 questions on the methodologic aspects of descriptive studies, including the sampling method, study variables and statistical analysis. The quality assessment results were classified into low quality (less than 15.5), moderate quality (15.5-29.5) and high quality (32-46). The binomial distribution formula was used to calculate the variance and estimated pooled prevalence. The Q statistic with a significance level of 10% was used to evaluate the presence of heterogeneity, and I² was used to determine the amount of heterogeneity among studies. Meta-regression analysis was used to evaluate the trend of the prevalence of refractive error with the study year and sample size. Analysis was performed on STATA.

Applicability/external validity: This meta-analysis was conducted worldwide. The results gave a prevalence of refractive error in children and adults separately, according to WHO regions for the first time.

Geographic focus: Review conducted worldwide.

Summary of quality assessment:

Medium confidence was attributed to the conclusions about the effects of this study. Although authors conducted appropriate methods to analyse data of included studies, language bias was not avoided and authors/experts were not contacted as part of the search strategy.