The prevalence of visual impairment and blindness has been assessed in population-based studies on elderly populations while there are relatively few population-based assessments of visual acuity of children (Dandona et al. 2002; Friedman et al. 2009; Gilbert et al. 2008; Goh et al. 2005; He et al. 2004; Multi-Ethnic Pediatric Eye Disease Study Group 2010; Murthy et al. 2002; Naidoo et al. 2003; Nirmalan et al. 2003; Rahi et al. 2010; Salomão et al. 2008). Sometimes, the rate of visual impairment in children was determined by surveys of blind schools or reviewing national registries and surveillance schemes. Investigations with such study designs can deliver valid figures only for those countries with a complete surveillance of all children affected. Because it has remained unclear, to which degree such surveys are representative and indicate the true prevalence of childhood vision impairment, and because India as the second largest nation does not have a complete registry of all visually impaired children yet, we performed the present study to examine the prevalence of visual impairment in children in a population-based manner in a city in Central India.
The Central India Children Eye Study was performed in all government schools run by the local government authority in Nagpur, a city of about 3 million inhabitants in Central India. Children attending government schools usually came from low to lower middle socioeconomic strata. The study included 11 829 school children. In each class, in general, all children participated in the examination. Mean age was 13.0 ± 1.9 years (range: 7–21 years). The Medical Ethics Committee of the Suraj Eye Institute, Nagpur, approved the study. All examinations were carried out in the schools. Social workers of the Suraj Institute particularly trained in refractometry measured visual acuity, before automatic refractometry was performed (Nidek Co, Aichi, Japan). The social workers were regularly controlled for the quality of their examinations. We used an retro-illuminated Snellen chart at a distance of 20 feet. Cycloplegia was not performed. Using the cut-off values of the World Health Organization (WHO) standard and the United States (US) standard, blindness was defined as visual acuity in the better-seeing eye of <20/400 and of <20/200, respectively, and visual impairment was defined as vision of <20/60 and ≥20/400, and of <20/40 and ≥20/200, respectively. Statistical analysis was performed using a commercially available statistical software package (spss for Windows, version 17.0; SPSS, Chicago, IL, USA). Data of frequency were given as mean ± standard error, of all other parameters as mean ± standard deviation. Ninety-five per cent confidence intervals (CI) and odds ratios (OR) were presented. All p values were 2-sided and were considered statistically significant when the values were <0.05.
Visual acuity measurements were available for 11 797 (99.7%) children. Using the WHO definition, 94 (0.8 ± 0.08%; 95% CI: 0.6, 1.0) children fulfilled the criteria for visual impairment and 8 (0.07 ± 0.02%; 95% CI: 0.02, 0.11) children were blind. Combining both definitions, 102 (0.86 ± 0.09%; 95% CI: 0.7, 1.0) children were visually impaired or blind. Using the US definition, 213 (1.8 ± 0.1%; 95% CI: 1.6, 2.1) children had visual impairment and 8 (0.07 ± 0.02%; 95% CI: 0.02, 1.1) children were blind. Combining both definitions, 221 (1.87 ± 0.13%; 95% CI: 1.63, 2.12) children were visually impaired or blind. In multiple logistic regression analysis, prevalence of low vision/blindness (WHO definition) was significantly associated with age (p < 0.001; Odds Ratio (OR):1.29; 95% CI: 1.16, 1.45), female gender (p = 0.001; OR: 2.10; 95% CI: 1.35, 3.27) and refractive error (p < 0.001; OR: 0.70: 95% CI: 0.66, 0.75).
Our findings agree with data from other children studies from India. In a study on 4074 school-aged children in rural Southern India, Dandona et al. (2002) reported on a prevalence of 2.6% of visual acuity of ≤20/40 in the better eye. In another study on 6447 school-aged children from an urban population in New Delhi/India, Murthy et al. (2002) found a prevalence of 4.9% of visual acuity of ≤20/40 in the better eye. In the interpretation of our findings, potential limitations of our study should be considered. First, it remains unclear how representative study population was for the whole population of the age group, because children attending government schools usually come from the low to lower middle socioeconomic population groups. Second, refractometry was not performed in cycloplegia. This potential weakness in the study design may, however, not have affected the assessment of visual acuity and the rate of visual impairment or blindness as defined by the WHO. Third, the cause for visual impairment was explored in detail. Fourth, perimetry was not performed although the visual field is an integral part of quality of vision. It may have led to an underestimation of the prevalence of blindness.
In conclusion, our study showed that the prevalence of visual impairment or blindness in government schools in Nagpur was 0.86 ± 0.09% (WHO definition) and 1.87 ± 0.13% (US definition). Associated factors were higher age, female gender and more myopic refractive error.