Visual impairment prevalence, causes, and role of healthcare access: a systematic review and meta-analysis in Iran

Authors: Afarid M, Vardanjani HM, Mahdaviazad H, Alamolhoda M, Farahangiz S.

Geographical coverage: Iran

Sector: Burden of disease, biomedical and health systems

Sub-sector: Epidemiology, causes and access

Equity focus: None specified

Study population: General population

Review type: Other review

Quantitative synthesis method: Systematic review and meta-analysis

Qualitative synthesis method: Not applicable

Background:

Visual impairment (VI) is a prevalent public health issue, affecting an estimated 285 million people worldwide, including 39 million who are blind. The primary causes, such as refractive errors, cataracts, macular degeneration, and glaucoma, are largely preventable or manageable with appropriate medical care. Despite numerous epidemiological studies on VI prevalence and causes, there is significant variation in the reported prevalence across different regions of Iran.

Objectives: To assess the VI prevalence, related causes, and its relationship with access to physicians (as one of the main indices of health care access) in Iran.

Main findings:

Authors included a total of eight observational studies in the review, all conducted in Iran. Three studies were conducted in urban areas, two studies in rural areas and three were conducted in rural/urban areas. Sample sizes varied among included studies from 1,185 to 11,975. Overall, five studies were conducted on all age groups and three on participants older than 40 years.

Authors reported the pooled prevalence of blindness using the random-effects model of 0.80% (95%CI: 0.61-0.99%, Q = 291.09, I² = 87.98%). They reported that the prevalence of blindness in men was 0.71% (95% CI: 0.46-0.97%) and in women was 0.92% (95% CI: 0.63-1.21%). Based on the results from the meta-regression, authors found that a greater number of GPs (t = -3.45, p = 0.001) and pharmacists (t = -2.30, p = 0.024) was associated with a lower prevalence of blindness, while a greater number of medical specialists was associated with a higher prevalence of blindness (t = 3.58, p = 0.001).

The overall pooled prevalence of low vision was 2.92% (95% CI: 2.40-3.44%, Q = 1092.58, I² = 96.52%); low vision prevalence rate was reported at 2.50% (95% CI: 1.85-3.14%) in men and 3.38% (95% CI: 2.54-4.22%) in women. There was considerable variation in the prevalence of low vision across age groups in both genders. The meta-regression model showed that the number of GPs was independently associated with a lower prevalence of low vision (t = -2.80, p = 0.006), while a higher density of specialists was associated with a higher prevalence of low vision (t = 2.54, p = 0.013).

Pooled prevalence of VI was 5.57% (95% CI: 4.71-6.43%, Q = 1803.98, I² = 97.89%). It was 5.02% (95% CI: 3.87-6.17%) in men (Figure 6) and 6.26% (95% CI:4.91-7.61%) in women (Figure 7). Based on the results obtained by the meta-regression model, authors found a negative independent association between higher density of pharmacists and prevalence of VI (t = −3.74, p = 0.001).

Authors included five studies in analysis based on the best-corrected visual acuity (BCVA) and found that the pooled prevalence of cataracts was 37.4% (95% CI: 29.5-45.3%) as the most common cause of VI. The pooled prevalence of macular degeneration was 9.7% (95% CI: 2.2-17.2%), amblyopia 8.2% (95% CI: 4.6-11.9%), corneal opacity 6.6% (95% CI: 1.9-11.2%), and glaucoma 4% (95% CI: 2.4-5.6%), based on BCVA. Moreover, the pooled prevalence of diabetic retinopathy was 17% (95% CI: 10.8-23.2%), which was reported in only two studies. The I² was 67%, 56.3%, 79.6%, 77.8% and 0.0% for the following causes, respectively: cataracts, amblyopia, macular degeneration, corneal opacity and glaucoma.

Based on presenting visual acuity (PVA), authors included five studies in the analysis. Authors reported refractive errors as the most common causes of VI, with a pooled prevalence of 54.6% (95% CI:43.4-65.8%).Pooled prevalence of cataract, amblyopia, macular degeneration, corneal opacity and glaucoma based on PVA was 23.5% (95% CI: 19.1-27.9%), 4% (95% CI:1.5-6.4%), 5.1% (95% CI: 1.7-8.4%), 1.3% (95% CI:0.4-2.2%), and 1.8% (95% CI: 0.6-3%), respectively. The I² in the heterogeneity test for the most common causes of VI based on PVA were as follows: refractive errors 89.4%, cataracts 48.6%, amblyopia 60.8%, macular degeneration 75.2%, corneal opacity 0.4%, and glaucoma 0.0%.

Authors conclude that access to health care services may lead to diagnosis of preventable causes of VI, and that well-designed studies and national surveys should be conducted to provide accurate data from different regions.

Methodology:

Authors included observational studies exploring the prevalence of VI, low vision and blindness from the general population, from all geographical regions of Iran.

The authors conducted a literature search on the following databases: PubMed, Web of Science, Scopus, Google scholar, Scientific Information database, and MagIran for observational studies. A manual search of the reference lists of review articles and practice guidelines were executed to identify any additional studies. Two investigators screened all retrieved studies independently in two phases including titles and abstracts and then full-texts. Included studies were critically appraised using the Newcastle-Ottawa quality assessment scale for observational studies and its starring system was used for quality and bias assessment. Authors extracted reference data (study year, location, etc) of included studies and outcome data.

Authors assessed heterogeneity among included studies using the Q-statistic and I² index. The overall pooled prevalence of blindness, low vision and VI were estimated by using a random-effects model with the DerSimonian and Laird method. A random-effects meta regression analysis was used to assess the associations between the geographic distribution of health care access and the overall pooled prevalence estimate. P values less than 0.05 were statistically significant.

Applicability/external validity:

Authors note that the prevalence estimates in Iran are lower than most other countries and regions, according to a WHO study.

Geographic focus:

This study focuses on Iran only; therefore, findings may only be applicable to Iran. However, as authors did not identify population-based studies from different geographical regions of Iran, the results cannot be generalised for the entire Iranian population.

Summary of quality assessment:

The systematic review has clear inclusion criteria, the search covered a sensible time period and full-texts of included studies were independently screened by two reviewers independently. The authors used appropriate methods to pool data into a meta-analysis and acknowledged some of the important limitations of the review. However, important limitations were identified, not acknowledged by the review authors. Searches were conducted in relevant databases, however, authors did not consider inclusion of unpublished work, and included English and Parisian studies only, which means some important literature may have been missed. While the authors report assessing study quality, they do not report the quality of each study included in the meta-analysis. Moreover, it is unclear if there was independent data extraction by two reviewers and they do not analyse the included studies by risk of bias status. Authors acknowledge that findings are not generalisable for the entire Iranian population, and do not draw strong policy conclusions. Taking all important limitations identified into account, low confidence was attributed to this review.