Prevalence and Major Causes of Visual Impairment in Iranian Adults: A Systematic Review.

PURPOSE: The purpose of this study is to provide a pooled estimate of moderate-to-severe visual impairment (MSVI) and blindness in Iran for people 50 years and over and to identify the major causes through systematic review.
MATERIALS AND METHODS: International (PubMed, ISI Web of Science, and Scopus) and national databases (Scientific Information Database, Barakat Knowledge Network System, Iran Databank of Ophthalmology Research, and Magiran) databases were searched. Following relevance assessment and critical appraisal, eight studies were included. A funnel plot was drawn to explore the stability for estimation. Single-variable meta-regression analysis was applied for heterogeneity assessment, and a random effect model was used (but no significant source for the observed heterogeneity was found).
RESULTS: Age-standardized pooled estimate of MSVI was 4.24% (95% confidence interval [CI]: 2.92-5.56); 3.98% (95% CI: 2.37-5.59) for men, and 4.08% (95% CI: 2.95-5.21) for women. Blindness (visual acuity <3/60) prevalence was 1.31% (95% CI: 1.23-1.39); 0.96% (95% CI: 0.89-1.03) for men, and 1.13% (95% CI: 1.06-1.20) for women. Causes of visual impairment (VI) were cataract (40.23%), amblyopia (12.03%), corneal opacity (9.63%), age-related macular degeneration (9.31%), diabetic retinopathy (4.94%), and glaucoma (3.67%).
CONCLUSION: VI prevalence in the 50 years and older population in Iran seems significantly better than the corresponding global estimates. A rough 60% rate of treatable VI was estimated, mostly attributable to unoperated cataract.

Introduction

Visual impairment (VI) is a prototype noncommunicable disease with physical and psychological consequences in the aging population and increased prevalence which is arising from demographic and epidemiologic transition. In 2010, 0.5% (95% confidence interval (CI): 0.4%–0.5%) of the world population was blind and 2.8% (CI: 2.5%–3.3%) had moderate-to-severe VI (MSVI).[1] Eighty-five percent of the blind and 78% of MSVI cases are over 50 years old, and 60% are women.[1] The global age-standardized prevalence of blindness and MSVI among adults 50 years of age and older are 1.9% and 10.4%, respectively.[1] About 8.2% of the visually impaired people are living in the Eastern-Mediterranean Region (EMR), where the prevalence of VI is about 6%.[2] Causes of VI account for about 70% of sense organ disorders in females[3] and 1.24% (95% uncertainty interval: 0.9%–1.68%) of global burden of diseases.[4]

Globally, the leading cause of VI is cataract, followed by glaucoma, age-related macular degeneration (ARMD), and corneal opacities (including those attributable to trachoma).[5] Later, when VI definition was redefined to include uncorrected refractive error, it became the top reason.[6] Several regional population-based surveys have been done in Iran; they are heterogeneous with respect to geography, age range, rural-urban coverage, etc., and the estimates they provide are varied. Recent reports on cataract surgical rates (CSRs) throughout the country and distribution of human resources have highlighted pervasive disparities which could explain inconsistencies in the reported estimates.[78]

Estimation of VI prevalence has implications for priority setting as well as public health strategy decisions. However, national surveys are costly and complex undertakings. Instead, the World Health Organization and the International Agency for the Prevention of Blindness have developed and advocate the use of an efficient protocol called Rapid Assessment for Avoidable Blindness (RAAB) which can still be a challenge when conducted at the national level. Therefore, it would be wise to first make the best conjecture by calculating a pooled estimate from available evidence. Although this approach not a replacement for RAAB studies gives us the best attainable estimate at the current time, also provides indices on the adequacy of data and the need for further surveys and RAAB projects in the country.

Materials and Methods

Definition of the pooled estimates

According to the International Classification of Diseases-10 (ICD-10), there are four levels of visual function: normal, moderate VI, severe VI, and blindness. The approximations and comparisons of metric, Snellen, and decimal notations, and cutoffs are presented elsewhere.[9] The two categories of MSVI (Presenting Distance Visual Acuity in Better Eye worse than 6/18 but equal to or better than 6/60 and worse than 6/60 but equal to or better than 3/60 in the better eye) are combined in this study, and they are referred to as “low vision” (worse than <6/18 and equal to or better than 3/60) (ICD-10 code: H54.1 and H54.2). Blindness is defined as a visual acuity of <3/60. VI includes low vision and blindness combined (ICD-10 code: H54.0).

Major causes of VI were considered as follows: Cataract, corneal opacities (including cicatricial trachoma), glaucoma, AMRD, diabetic retinopathy, and moderate to deep amblyopia for further analysis on the entities.

Search strategy

In line with the National and Subnational Burden of VI project in Iran,[910] which is part of the National and Subnational Burden of Diseases, Injuries, and Risk Factors study from 1990 to 2013[9] and the ongoing review of literature, we searched for relevant published literature between January 2002 and December 2014 indexed in international sources including PubMed, ISI Web of Science, Scopus, and national electronic databases including the Scientific Information Database (www.sid.ir), Barakat Knowledge Network System (http://health.barakatkns.com), Iran Databank of Ophthalmology Research (www.iraneyedoc.com), and Magiran (www.magiran.com). The search covered Farsi and English literature.

We used and combined the following search strings:

([monocular blindness OR complete blindness OR total blindness OR functional blindness OR avoidable blindness OR treatable blindness OR preventable blindness OR unavoidable blindness OR needless blindness] AND [Iran OR IR OR Iranian OR Iranians])

([reduced vision OR subnormal vision OR diminished vision OR vision loss OR low vision] AND [Iran OR IR OR Iranian OR Iranians])

(vision impaired OR VI OR visual defect OR visual loss) AND (Iran OR IR OR Iranian OR Iranians)

(amblyopia OR corneal opacity OR cataract OR glaucoma OR refractive error OR diabetic retinopathy OR lens opacity OR IOL implantation OR cataract extraction OR presbyopia OR ARMD OR AMD OR retinal vascular disorder OR ocular trauma OR trachoma) AND (Iran OR IR OR Iranian OR Iranians).

Study selection and quality assessment

By performing a comprehensive search, 4984 documents were identified. The flow of literature review is summarized in Figure 1. After removing duplicates, two of the authors (EA and GhS) independently screened the titles and abstracts to select cross-sectional or population-based studies containing data regarding vision status specifying VI levels. Any disagreement was resolved by consensus or a third reviewer (sustainable forest management). Studies with restricted age groups (cohort studies)[11] not representative of the total population or eye health screening data[12131415] and those poorly defined) nonconsistent with standard definitions and scales)[16] were excluded. Two independent reviewers conducted critical appraisal of methodological characteristics, study population framework, sampling method, sample size, VI measurement methods, and response rate (score 0–10). Studies with a lower score than 6 were excluded. Reference lists of the eligible articles (eight papers) were reviewed to identify additional eligible studies next. Tables 1 and 2 summarize the features of the included studies.

Figure 1

Flow chart of the systematic search and review

Table 1

List of studies included in the meta-analysis and their point estimates of the prevalence of moderate-to-severe visual impairment and blindness

Table 2

Appraisal scores by the two reviewers on all appraisal criteria for each of the eight studies

Data abstraction from included papers

Data were collected independently by two reviewers (EA and GhS/MRL) to ensure validity and reproducibility of the methods. A standardized data abstraction sheet was employed. Collected data included background information, such as citations, publication year, study period, publishing journal, and the respective corresponding author; methodological aspects of study design, data source, study scope (urban, rural, slum, or nomadic; provincial, district, or community level), measurement tools, sample size, sampling method; response rate and the proportion of cases to the study population; and age- and sex-specific prevalence rates with 95% CI. Corresponding authors were contacted for clarifications and obtaining more information about the missing aspects, when necessary. When participants with lower than 50 years were included in studies, the prevalence estimation was recalculated for age group more than 50 years.

Statistical analysis

For statistical analysis, the STATA statistical software (version 11; STATA Corporation, College Station, TX, USA) was used. Standardized prevalence (adjusted for the age distribution) for the 50 years and older population with 95% CI of each study point estimate was calculated. Cochran's test was used to determine the heterogeneity index among studies. Compatible with heterogeneity results, a random effect was assumed to estimate the overall and sex-specific VI prevalence in the Iranian adult population. The funnel plot was used to evaluate publication bias and data adequacy for estimation. Meta-regression was used to explore the potential sources of heterogeneity among studies. Point estimation of VI prevalence was delivered in forest plots wherein the diamond area represents study weights and lines demarcate 95% CI.

Results

Eight studies were entered. The flow chart of the systematic search and review process is illustrated in Figure 1.

In terms of age and sex distribution, five studies covered all age groups, and in the other three, participants were over 40 or 50 years. Three of the eight studies were carried out in Tehran (province or the capital), two in Yazd district, one in Sari (center of Mazandaran province) and two were conducted in provinces of Khuzestan, and Sistan-o-Balouchestan.

Age-standardized prevalence of MSVI (H54.1 and H54.2) for the 50 and older age groups varied from 1.4% in 2009 in Varamin[21] to 7.14% in 2006 in Tehran province.[19] Considering a random effect (chosen due to heterogeneity among studies: Q = 56.62, P < 0.001), respective pooled estimate of MSVI was calculated as 4.24% (95% CI: 2.92–5.56); 3.98% (95% CI: 2.37–5.59) for males and 4.08% (95% CI: 2.95–5.21) for females Figure 2. Age-standardized prevalence of blindness (H54.0) in this age group varied from 0.9% in 2009 in Yazd[22] to 3.42% in 2006 in Khuzestan.[20] Again, considering a random effect assumption (Q = 63.21, P < 0.001), blindness prevalence was estimated at 1.31% (95% CI: 1.23–1.39); 0.96% (95% CI: 0.89–1.03) for males, and 1.13% (95% CI: 1.06–1.20) for females Figure 3.

Figure 2

Age-standardized prevalence of moderate-to-severe visual impairment by gender; each line represents results of one of the eight studies, center line of diamonds represent summary measures (ES: Effect size) and lateral tips of diamond show CIs and area of each square is proportional to the study's weight in the meta-analysis. CI: Confidence interval

Figure 3

Age-standardized prevalence of blindness by gender; each line represents results of one of the eight studies, centere line of diamonds represent summary measures (ES: Effect size) and lateral tips of diamond show CIs and area of each square is proportional to the study's weight in the meta-analysis. CI: Confidence interval

Furthermore, in the single-variable meta-regression analysis for heterogeneity considering year of study, province, and sample size, no significant source of variance was found among studies with respect to MSVI and blindness (P > 0.05).

Causes of visual impairment

The major causes of VI, in descending order, were cataract (40.23%; 95% CI: 39.58–40.88), amblyopia (12.03; 95% CI: 11.45–12.60), corneal opacities including cicatricial trachoma (9.63%; 95% CI: 9.20–10.07), ARMD (9.31%; 95% CI: 8.72–9.89), diabetic retinopathy (4.94%; 95% CI: 4.61–5.27), and glaucoma (3.67%; 95% CI: 3.41–3.93); the remaining 20.19% were due to other or unknown causes.

Discussion

The calculated blindness and MSVI prevalence rates were 1.31 and 4.24, respectively. This blindness estimate is very similar to that reported in Eastern Europe and South America in 2002 and lower than the estimates for the EMR as a whole; somehow on par with the estimates for blindness in Bulgaria and Turkey[5] but lower than that in Lebanon, Oman, Saudi Arabia, and Tunisia.[5]

Despite having a relatively better status as compared with that in the EMR,[25] based on the analysis provided in the introduction, our estimate mean that a good deal of treatable blindness does exist. As stated in the results, visually significant cataract is the leading cause of blindness in Iran, and 40% of VI was attributed to unoperated cataract. Since the rural population is underrepresented in most of the studies, one may safely attribute more than 40% VI to cataract. In fact, by excluding amblyopia from the causes of VI-considering the fact that it is mostly unilateral-the share of cataract increases to 45% which is even more comparable with the global rate of 50% for cataract blindness[5] and the overall rate in EMR (49%).[5] It should be noted that amblyopia is not normally listed along with VI causes as it mostly causes unilateral mild-to-moderate VI. In addition, coincident amblyopia cannot be differentiated from other major causes of VI.

We know that CSR in Iran has increased significantly in the past two decades. The estimation of 526 surgeries per million populations in 2000 rose to 1331 in 2005 and reached a value of 6328 in 2010.[78] This improvement has been partly attributed to transition to phacoemulsification in Iran.[25] This period coincides with our systematic review span. Having significant cataract-related VI despite acceptable or high CSR means that cataract surgical coverage (CSC) is insufficient and points to disparities in access in one hand and induced demand on the other hand. This analysis is substantiated by the inequity observed in the distribution pattern of ophthalmologists which ranges between 0.5 and 2.8/million population throughout different provinces, excluding the capital.[26]

The cataract backlog is not likely to be attributable to lack of surgical equipment, diagnostic services, or surgeons' availability either. Access seems responsible for the observed inequality; the problem is specifically due to human resource distribution, referral system, and secondary eye care centers. Poor access is being addressed partly through an outreach program by Nooravaran Salamat Mobile Hospital. They report that they have provided 5122 cataract surgeries in the past 5 years in 17 most underserved zones of the country. Grossly, one out of 20 visited cases needed cataract surgeries.[27]

This study has some limitations; uncorrected refractive error was identified as a major cause of treatable blindness. However, 5 of the 8 articles reported best-spectacle-corrected visual acuity and not presenting visual acuity (PVA), and thus, so the aggregate estimate for uncorrected or undercorrected refractive error prevalence cannot be made by current meta-analysis.

A variable share of 2.3%[28] to 33.6%[17] of VI has been attributed to uncorrected refractive error in different studies; the global estimate is 21%.[29] By comparison, we can project a VI of about 20% for our own status. Cataract and the projected VI due to uncorrected refractive error sum up to 60%–65%, which is comparable with the almost 68% global rate of avoidable blindness.[29]

Corneal opacities were responsible for about 10% of the cases of blindness. This is a high estimate and is due to cicatricial trachoma whose active form used to be endemic and highly prevalent in the past decades. It has been reported as the cause of VI in 3%[30] to 35.3%[31] in other studies.

ARMD share in blindness was about 9.5% in our analysis compared to the 6.6% in the world.[29] It is by far the leading cause of blindness in the Western Europe and North America as it accounts for 17% of the cases;[29] on the other hand, it causes <3% blindness in south Asia.[29] Our relatively higher estimate may be attributed to the epidemiologic transition in the country.[32] The current life expectancy in Iran is 74 years (it was 69 in 2000 and 65 in 1990).[33] The increasing rate of macular degeneration has also been reported in the Global Burden of Diseases dataset wherein the burden increased from a disability-adjusted life year of 84/100,000–102/100,000 at 1990–2010 [unpublished data].[34]

The rate of VI due to diabetic retinopathy is comparable with Asia Pacific – high income countries like Singapore and Western Europe.[3536] Our nation has more than 150 vitreo-retinal subspecialists and a large number of centers are equipped with the needed diagnostic and therapeutic modalities for diabetic retinopathy care but we need targeted and comprehensive evidence on diabetic retinopathy.

Estimates for glaucoma in population-based surveys are not generally dependable. They set visual acuity cutoffs for full eye examination, and we know that glaucoma affects acuity only in moderate to advanced stages; so glaucoma is underestimated. In addition, glaucoma is occasionally diagnosed along with other ocular comorbidities such as cataract; this is highly relevant to RAAB which trades off precise diagnosis with protocol efficiency and primarily targets treatable blindness estimation.

We had 20% estimates as “others.” This covers childhood and congenital sequelae (dystrophies and amblyopia) and optic nerve pathologies.

The intergender difference in the prevalence of blindness was 0.17% and statistically significant. This is in line with a recent meta-analysis that indicated more women than men were visually impaired and blind in all regions globally; women account for 64.5% of all blind people and the overall women to men blindness ratio is 1.43%, ranging from 1.39% in Africa, and 1.41% in Asia.[37] This figure was 1.18 in our study which shows a relatively better status, but as we could not make appropriate age adjustments in gender-specific prevalence in the eight reviewed studies, we cannot draw accurate comparisons with other studies.

A systematic review on CSC in low- and middle-income countries has shown that women's uptake of cataract surgery is about half of that for men.[38] Moreover, it has been shown that female patients receive less postoperative care.[39]

We determined age standardized prevalence rates of VI with a well-defined disease definition in 10 years, but at the same time, prevalence rates were varied in these eight studies, and we observed significant heterogeneity. While six studies included rural samples, the rural-urban combinations in the studied populations were not proportionate to that of the nation; separate estimates for rural and urban populations were provided only in two studies. In addition, some provinces were represented more than once (Tehran and Yazd), and most provinces are not represented.

It should also be mentioned that reviewed studies were published during the past decade; it is likely that VI and its causes have varied in the time course of a decade. These underscore the necessity for supplemental studies in which PVA should be the primary index, proportionate rural populations be included, and the geographical gap be addressed.

Conclusions

We extrapolated a stable estimate of 1.3% and 4.2% for blindness and MSVI prevalence, respectively, for the 50 years and older age group in Iran, through a meta-analysis of eight original population-based surveys in the past decade which are significantly better than the global corresponding estimates. The study findings, in light of global evidence, point to a rough 60% treatable VI – mostly attributed to visually significant cataract. Estimates need to be updated for the 2nd half of the vision 2020 time course, addressing PVA (significant uncorrected refractive error), rural population, and updated attributable reasons. This can be in the form of a national survey or regional RAAB projects.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.