POLICY AND PRACTICE

Global magnitude of visual impairment caused by uncorrected refractive errors in 2004

Prévalence mondiale des déficiences visuelles dues à des défauts de réfraction non corrigés en 2004

Magnitud mundial de las discapacidades visuales por defectos de refracción no corregidos en 2004

Serge Resnikoff¹; Donatella Pascolini; Silvio P Mariotti; Gopal P Pokharel

Chronic Disease Prevention and Management, WHO, 20 avenue Appia, 1211 Geneva 27, Switzerland

ABSTRACT

Estimates of the prevalence of visual impairment caused by uncorrected refractive errors in 2004 have been determined at regional and global levels for people aged 5 years and over from recent published and unpublished surveys. The estimates were based on the prevalence of visual acuity of less than 6/18 in the better eye with the currently available refractive correction that could be improved to equal to or better than 6/18 by refraction or pinhole.
A total of 153 million people (range of uncertainty: 123 million to 184 million) are estimated to be visually impaired from uncorrected refractive errors, of whom eight million are blind. This cause of visual impairment has been overlooked in previous estimates that were based on best-corrected vision. Combined with the 161 million people visually impaired estimated in 2002 according to best-corrected vision, 314 million people are visually impaired from all causes: uncorrected refractive errors become the main cause of low vision and the second cause of blindness.
Uncorrected refractive errors can hamper performance at school, reduce employability and productivity, and generally impair quality of life. Yet the correction of refractive errors with appropriate spectacles is among the most cost-effective interventions in eye health care.
The results presented in this paper help to unearth a formerly hidden problem of public health dimensions and promote policy development and implementation, programmatic decision-making and corrective interventions, as well as stimulate research.

RÉSUMÉ

Des estimations mondiales et régionales pour l'année 2004 de la prévalence des déficiences visuelles dues à des défauts de réfraction non corrigés chez les plus de 5 ans ont été établies à partir d'enquêtes récentes publiées et non publiées. Ces estimations ont été obtenues d'après la prévalence d'une acuité visuelle inférieure à 6/18 pour le meilleur des deux yeux avec la correction réfractive actuellement disponible, susceptible d'être ramenée à une valeur supérieure ou égale à 6/18 par une correction réfractive ou un trou sténopéique.
On estime à 153 millions (plage d'incertitude : 123 à 184 millions) le nombre total de personnes souffrant d'une déficience visuelle due à un défaut de réfraction non corrigé, dont huit millions d'aveugles. La cause de la déficience visuelle a été laissée de côté dans les estimations antérieures reposant sur la meilleure vision corrigée. Si l'on combine ce chiffre à l'estimation de 2002 du nombre d'individus déficients visuels établie d'après la meilleure vision corrigée, 314 millions de personne présentent une déficience visuelle toutes causes confondues, les défauts de réfraction non corrigés devenant la principale cause de mauvaise vision et la seconde cause de cécité.
Les défauts de réfraction non corrigés peuvent nuire aux résultats scolaires, réduire la capacité à occuper un emploi et la productivité et, de manière générale, détériorer la qualité de vie. La correction des défauts de réfraction par des lunettes adaptées reste l'intervention la plus rentable en termes de soins ophtalmologiques.
Les résultats présentés dans cet article contribuent à faire ressortir un ample problème de santé publique formellement masqué, à favoriser le développement et la mise en œuvre de politiques, ainsi que la prise de décisions programmatiques et de mesures correctives, et à stimuler la recherche.

RESUMEN

Se ha estimado la prevalencia de las discapacidades visuales causadas por los defectos de refracción no corregidos en 2004 a nivel regional y mundial en la población de 5 o más años a partir de encuestas recientes publicadas o inéditas. Las estimaciones se basaron en la prevalencia de una agudeza visual inferior a 6/18 en el mejor ojo con la corrección refractiva del momento, pero mejorable hasta 6/18 o más con medidas de corrección de la refracción o con un agujero estenopeico.
Se estima que un total de 153 millones de personas (intervalo de incertidumbre: 123 - 184 millones) sufren discapacidad visual como consecuencia de defectos de refracción no corregidos, incluidos ocho millones que padecen ceguera. Esta causa de discapacidad visual no se ha tenido debidamente en cuenta en estimaciones anteriores basadas en la mejor visión corregida. Si se suman a ello los 161 millones de personas con discapacidad visual estimados en 2002 de acuerdo con el criterio de la mejor visión corregida, se obtiene un total de 314 millones de personas con discapacidad visual por todas las causas: los defectos de refracción no corregidos se convierten así en la principal causa de disminución de la agudeza visual y la segunda causa de ceguera.
Los defectos de refracción no corregidos pueden reducir el rendimiento escolar, la empleabilidad y la productividad, y por lo general merman la calidad de vida. Sin embargo, la corrección de esos defectos con unas gafas apropiadas es una de las intervenciones más costoeficaces de la atención oftalmológica.
Los resultados aquí presentados pueden contribuir a hacer aflorar un problema hasta ahora oculto de gran trascendencia en el campo de la salud pública, y promover la formulación y ejecución de políticas, la toma de decisiones programáticas, las intervenciones correctivas y la realización de investigaciones en ese terreno.

INTRODUCTION

Refractive errors (myopia, hyperopia and astigmatism; presbyopia is not included in this study given the present paucity of data, but it is recognized that uncorrected, it could lead to an impaired quality of life) affect a large proportion of the population worldwide, irrespective of age, sex and ethnic group. Such refractive errors can be easily diagnosed, measured and corrected with spectacles or other refractive corrections to attain normal vision. If, however, they are not corrected or the correction is inadequate, refractive errors become a major cause of low vision and even blindness (for a selection of studies, see http://ftp.who.int/nmh/references/RE-estimates-references.pdf).

Visual impairment from uncorrected refractive errors can have immediate and long-term consequences in children and adults, such as lost educational and employment opportunities, lost economic gain for individuals, families and societies, and impaired quality of life. Various factors are responsible for refractive errors remaining uncorrected: lack of awareness and recognition of the problem at personal and family level, as well as at community and public health level; non-availability of and/or inability to afford refractive services for testing; insufficient provision of affordable corrective lenses; and cultural disincentives to compliance.

The definition of visual impairment in the International statistical classification of diseases, injuries and causes of death, 10th revision (ICD-10), H54, is based on "best-corrected" vision, i.e. visual acuity obtained with the best possible refractive correction.¹ However, to assess the extent of visual impairment caused by uncorrected refractive errors, estimates need to be based on "presenting" vision, i.e. visual acuity obtained with currently available refractive correction, if any. Thus, presenting vision, as opposed to best-corrected vision, provides the prevalence of visual impairment that could be improved simply by appropriate corrective refraction. Basing the definition of visual impairment on presenting vision extends the current definition to one that characterizes visual impairment faced by people in day-to-day activities.

Using best-corrected vision, visual impairment was estimated to affect 161 million people globally in 2002, of whom 37 million were blind.² The main cause of blindness and low vision was cataract; however, it was recognized that unless uncorrected refractive errors were included among the causes, visual impairment at global level was significantly underestimated.

This paper presents the estimate of the prevalence of visual impairment from uncorrected refractive errors for all ages over 5 years at regional and global levels, based on recent published and unpublished surveys. Some results from this paper were reported in a WHO press release on 11 October 2006 to mark World Sight Day.³

Methods

Definitions

Presenting vision is defined by the visual acuity in the better eye using currently available refractive correction, if any. Best-corrected vision is the visual acuity in the better eye achieved by subjects tested with pinhole or refraction.

Visual impairment caused by uncorrected or inadequately corrected refractive errors is defined as visual acuity of less than 6/18 in the better eye that could be improved to equal to or better than 6/18 by refraction or pinhole, thus spanning the low vision and blindness categories as currently defined in the ICD-10.

It should be noted that in the revision of the ICD-10 categories of visual impairment proposed in 2003 by a WHO consultation on the development of standards for characterization of vision loss, low vision is replaced by two categories: moderate visual impairment (presenting visual acuity less than 6/18 but equal to or better than 6/60) and severe visual impairment (presenting visual acuity less than 6/60 but equal to or better than 3/60).⁴

Population estimates and WHO subregions

Estimates of population size and structure were based on the latest estimates of world population (for 2004) in the World population prospects: the 2004 revision; estimates of demographics were based on the World urbanization prospects 2003 – both sources from the United Nations Population Division.^5,6

For the classification of WHO Member States into 17 epidemiological subregions, see Murray & Lopez, 1996.⁷

General inclusion criteria

The following criteria were used to select studies.

• The prevalence of best-corrected and presenting visual acuity of less than 6/18 had to be reported or, alternatively, the distribution of causes of presenting visual impairment.
• In children, refractive diagnostics had to be determined by objective refraction under cycloplegia plus subjective refraction.⁸
• The studies had to be population-based, representative of the area sampled, with definitions of visual impairment clearly stated. Studies with inadequate sample sizes and response rate were not included.
• Data reported only for eyes or for the worse eye could not be included in the estimates calculated for people and the better eye.

For further discussion of selection criteria, see Resnikoff et al.,² and Pascolini et al., 2004.⁹

Sources of epidemiological data

Literature sources were searched systematically in Medline up to April 2006. Most surveys meeting the selection criteria were conducted within the past five years; the earliest surveys date from 1995. Unpublished data were provided by academic institutions and national programmes for the prevention of blindness.

Table 1 shows the 31 countries for which surveys that met the selection criteria were available, the bibliography can be found at http://ftp.who.int/nmh/references/RE-estimates-references.pdf and in the WHO Prevention of Blindness and Deafness Programme's global data on visual impairment.¹⁰

For the age group 5–15 years, 16 surveys were found to fit the selection criteria. Of these, 10 were conducted in different countries using a specially designed protocol to estimate the prevalence of visual impairment from uncorrected refractive errors (the refractive error study in children (RESC; see Négrel et al., 2000, for the details of the protocol).¹¹ The RESC studies provided extensive information on visual acuity, refractive errors and use of spectacles.

For the age group 50 years and older, 38 surveys met the inclusion criteria. Of these, 30 were surveys for the rapid assessment of cataract surgical services (RACSS), which also provide prevalence of presenting and best- corrected visual acuity.¹²

An additional 14 surveys reported age-specific prevalence of presenting visual impairment and its causes in other age groups.

Estimation of prevalence of visual impairment from uncorrected refractive errors

For the age group from 5 to 15 years, the prevalence is estimated by the difference between the prevalence of presenting and best-corrected visual acuity of less than 6/18 with refraction under cycloplegia: this difference corresponds to the prevalence of presenting visual acuity that could be improved to equal to or better than 6/18 by appropriate correction. In the case of studies reporting only the prevalence of presenting visual acuity, the prevalence of visual impairment due to refractive error was determined from the distribution of causes determined in the surveys.

The prevalence for people aged 16–39 years was estimated to be the same as that for those aged 5–15 years, on the assumption that from the ages of 16 years to 39 years, the refractive status generally does not undergo changes that require further correction.¹³

The prevalence for people aged 40–49 years was either estimated from the results of surveys that reported age-specific data for this age group or calculated by a linear fit between the prevalence at age 39 and 55 years.

For the population aged 50 years and older, the prevalence was estimated from the difference between visual acuity of less than 6/18 with the available correction and visual acuity of less than 6/18 with best correction determined using refraction or pinhole, assuming that pinhole approximates complete refraction.

Estimation of prevalence of blindness from uncorrected refractive errors

Uncorrected refractive errors in adults aged 50 years and older have been shown to lead to blindness in some regions: the corresponding prevalence of blindness was determined from the difference between prevalence of presenting and best-corrected visual acuity of less than 3/60.

Blindness from uncorrected refractive errors was also reported in some surveys for the age group 40–49 years. Since there were insufficient data worldwide, it was assumed that the global number of people blind from this cause in this age group was 5.13 times lower than the corresponding number in people aged 50 years and over, based on the ratio of the total number of people visually impaired 50 years and over and those aged 40–49 years.

Estimation of refractive services coverage

The RESC studies also report the prevalence of uncorrected visual acuity in the age group 5–15 years: the prevalence of uncorrected, presenting and best- corrected visual acuity (VA <6/18) provides an estimate of the percentage coverage of refractive services using the formula:

Since percentage coverage is based on presenting visual acuity, it is an estimate of both the provision of refractive services and the compliance to prescription.

Extrapolations

Since data were not available for every country, extrapolations were made to estimate the global prevalence of visual impairment from uncorrected refractive errors. The rationale for the extrapolations was the similarity of the epidemiology of refractive errors, the availability and/or affordability of refractive services and compliance. Various kinds of extrapolations were made, based on the data selected:

• the prevalence in urban and rural areas within a country was extrapolated to all urban and rural areas, respectively, of the country; the country prevalence was determined by weighting the prevalence by the rural–urban distribution of the population;
• in subregions with data from several countries, an average prevalence was determined and applied to all other countries in the subregion. The average was calculated by weighting the prevalence from the countries by their share of the population in the subregion and taking into account the urban and rural distribution of the population;
• in the case of whole subregions lacking data, the prevalence was extrapolated from other subregions with similar epidemiology of refractive errors and with similar WHO epidemiological classification.⁶

For China and India, estimates were made separately because of the population size. Some subregions estimated to have similar prevalence of visual impairment from uncorrected refractive errors and provision of refractive services were combined.

Results

Prevalence of visual impairment from uncorrected refractive errors by age and subregion

It is estimated that globally 153 million people over 5 years of age are visually impaired as a result of uncorrected refractive errors, of whom 8 million are blind. Table 2 shows the number of people in the WHO subregions with visual impairment from this cause and the corresponding prevalence by age. There is no evidence of visual impairment caused by uncorrected refractive errors in children aged less than 5 years.

From the data reported in surveys it was not possible to distinguish conclusively between the prevalence of male and female cases of uncorrected refractive errors for any of the age groups.

Some 12.8 million in the age group 5–15 years are visually impaired from uncorrected or inadequately corrected refractive errors, a global prevalence of 0.96%, with the highest prevalence reported in urban and highly developed urban areas in south-east Asia and in China.

The number of people aged 16–39 years visually impaired from uncorrected refractive errors is 27 million, a prevalence of 1.1% globally. This could, however, be an underestimate, being derived directly from the prevalence in the age group 5–15 years, although the prevalence of refractive errors, especially myopia, is higher between the ages of 13 and 18 years.

The prevalence in people aged 40–49 years globally is 2.45%; it is high in subregions or countries where the prevalence for people aged 50 years and older is also high. Almost 95 million people aged 50 years and older are visually impaired from uncorrected refractive errors: the prevalence is between 2% and 5% in most regions of the world, but is almost 10% in China and almost 20% in India and in Sear-D (WHO subregions defined in Murray & Lopez, 1996).⁷

Of the 95 million people aged 50 years and older visually impaired from uncorrected refractive errors, 6.9 million are blind (Table 3). Based on this, it is estimated that 1.3 million people in the age group 40–49 years are blind from uncorrected refractive errors. There was no evidence in any surveys of significant blindness in the age groups 5–15 years and 16–39 years.

The average coverage of refractive corrections calculated from the RESC studies for visual acuity cut-off point of less than 6/18 is shown in Table 4.

Discussion

Limitations: uncertainties of the data and extrapolations

The sampling and examination methods in the RESC studies were designed to produce results that could be directly compared between countries: for presenting visual acuity of less than 6/18, the uncertainties given were between 20% and 25%. These uncertainties in turn affect the estimates for the age group 16–39 years, which are based on the results for the age group 5–15 years. The uncertainties reported in the RACSS studies are between 15% and 25% for the prevalence of visual acuity of less than 6/18; other studies reported uncertainties from 15% to more than 20%. In all the studies, the uncertainties become higher for the prevalence of visual impairment of less than 3/60.

The assumption that in adults the measurement of visual acuity with pinhole approximates the results obtainable with full refraction brings some additional bias to the estimates.¹⁴

Significant limitations are introduced by the need to extrapolate the prevalence and provision of services data from one rural or urban area to all rural and urban areas in a country or subregion, as well as from one country to other countries, or from one country to whole subregions. Some extrapolations could be particularly prone to error, as in the case of countries such as China or India, for which extrapolations are made for very large populations, or in the case of subregions with scarce data, such as Eur-B1, Eur-B2 and Eur-C (WHO subregions defined in Murray & Lopez, 1996).⁷

To check the consistency of the extrapolations, the estimates were verified using studies that did not fit the inclusion criteria due to the visual acuity ranges or the testing and reporting methods used. Studies on the epidemiology of refractive errors have also been taken into consideration in all extrapolations.

The assumptions and extrapolations were reviewed by a group of independent experts convened by the WHO Prevention of Blindness Programme.¹⁵

Extent of visual impairment from uncorrected refractive errors worldwide

Uncorrected refractive errors are a major cause of blindness and low vision: it is estimated that 8 million people are blind and 145 million have low vision because of lack of adequate refractive correction (Table 5). The uncertainties associated with the data and the extrapolations can lead to overestimates as well as underestimates of these figures: if the uncertainties are estimated at 20%, the total of 153 million could vary from 123 million to 184 million.

The estimate of visual impairment caused by uncorrected refractive errors presented in this paper confirms that the problem is of public health concern, as emphasized previously.^16,17 This finding is significant considering that refractive errors could be easily diagnosed and that spectacle correction is among the most cost-effective interventions in eye care.

Global causes of visual impairment

If blindness and low vision from uncorrected refractive errors (this paper) and from all other causes (2002 estimate) are combined, 314 million people are visually impaired globally (Table 5). Uncorrected refractive errors are the second cause of blindness after cataract (Fig. 1) and the main cause of low vision: overall, they are the cause of almost half of all visual impairment.

Given the magnitude of the problem, uncorrected refractive errors need to be assessed and reported as a cause of visual impairment. It is expected that the ICD-10 definition of visual impairment will include, from the next revision in 2009, presenting vision along with the currently used best-corrected vision.

Reasons why refractive errors remain uncorrected

In the age group 5–15 years, non- correction of refractive errors is due to several factors: the lack of screening, and the availability and affordability of refractive corrections are the most important. However, cultural disincentives also play a role, as shown in surveys from countries where routine screening and provision of corrections are free of charge or easily accessible, but compliance remains low (S Wedner, unpublished observations, 2006).^18,19 Perhaps one of the most remarkable findings in this study is that even in economically advantaged societies, refractive errors can go undetected or uncorrected in children.²⁰

The estimated number of people aged 50 years and older visually impaired from uncorrected refractive errors is over 94 million, a figure that could be an underestimate, being based in part on studies that used only pinhole in place of full refraction.¹⁴ In countries where the prevalence is very high, important underlying causes are index myopia caused by cataract, uncorrected aphakia and insufficient intra-ocular lens correction.^21,22 This is particularly true in rural areas.

For the age group 5–15 years, the prevalence of visual impairment from uncorrected refractive errors in some regions appears to be higher in urban areas than in rural areas, despite the reported better access to services. This may be due to a high incidence of myopia in these populations: it is suggested that there may be a direct cause–effect relation between increased access to education and myopia, but other secular changes could be contributing factors. In this age group the prevalence of myopia reported in studies that used the same definitions and cut-off levels ranges from 3% to 35%, hypermetropia from 0.4% to 17%, astigmatism from 2.2% to 34% depending on the region and on the urban/rural setting.

The coverage of refractive corrections determined from the RESC studies is less than or around 50% in most regions of the world; urban areas have, as expected, better service coverage than rural areas (Table 4).

CONCLUSIONS

These findings warrant the urgent implementation of the following fundamental policies.

• Screening of children for refractive errors should be conducted at community level and integrated into school health programmes, accompanied by education and awareness campaigns to ensure that the corrections are used and cultural barriers to compliance are addressed and removed.
• As the cost of refractive corrections is still high compared with the personal and family resources in many regions, corrections must be accessible and affordable for people of all ages.
• Eye-care personnel should be trained in refraction techniques. Training and information programmes should also be designed for teachers and school health-care workers.
• Reliable and affordable equipment for refractive assessments should be developed.
• Refraction services need to be integrated with eye-care systems and included as a part of cataract surgery services.
• Impairment from uncorrected refractive errors, provision of refractive services and outcomes of the provisions should be monitored at national level to identify communities in need and evaluate the most cost-effective interventions.

Another aspect of visual functioning that has not been discussed in this paper is near vision: the unmet need of correction of presbyopia is currently unknown and should be assessed and included in future estimates of visual impairment.

Acknowledgements

The authors would like to acknowledge the contribution of the WHO Refractive Errors Working Group. This study was supported financially by a grant from the World Optometry Foundation of the World Council of Optometry. Particular appreciation is due to all experts from around the world who have generously provided survey reports, journal articles and unpublished results.

Competing interests: None declared.

References

1. International statistical classification of diseases, injuries and causes of death, tenth revision. Geneva: WHO; 1993.        

2. Resnikoff S, Pascolini D, Etya'ale D, Kocur I, Pararajasegaram R, Pokharel GP, et al. Global data on visual impairment in the year 2002. Bull World Health Organ 2004;82:844-51. PMID:15640920        

3. Sight test and glasses could dramatically improve the lives of 150 million people with poor vision [WHO news release, 11 October 2006]. Geneva: WHO; 2006. Available at: http://www.who.int/mediacentre/news/releases/2006/pr55/en/index.html        

4. Prevention of Blindness and Deafness. Consultation on development of standards for characterization of vision loss and visual functioning. Geneva, 4–5 September 2003. Geneva: WHO; 2003 (WHO/PBL/03.91).        

5. World population prospects: the 2004 revision [CD-ROM edition – extended dataset]. New York: United Nations; 2005.        

6. World urbanization prospects: the 2003 revision. New York: United Nations; 2004.        

7. Murray CJL, Lopez AD, eds. The global burden of disease: a comprehensive assessment of mortality and disability from diseases, injuries and risk factors in 1990 and projected to 2020 [Global Burden of Disease and Injury Series, Vol. 1]. Cambridge: Harvard School of Public Health on behalf of the World Health Organization and the World Bank; 1996.        

8. Zhao J, Mao J, Luo R, Li F, Pokharel GP, Ellwein LB. Accuracy of noncycloplegic autorefraction in school-age children in China. Optom Vis Sci 2004;81:49-55. PMID:14747761 doi:10.1097/00006324-200401000-00010        

9. Pascolini D, Mariotti SP, Pokharel GP, Pararajasegaram R, Etya'ale D, Négrel AD, et al. 2002 global update of available data on visual impairment: a compilation of population-based prevalence studies. Ophthalmic Epidemiol 2004;11:67-115. PMID:15255026 doi:10.1076/opep.11.2.67.28158        

10. Prevention of Blindness and Deafness. Available data on blindness, update 2006. Geneva: WHO; 2006. Available at: http://www.who.int/blindness/publications/global_data.pdf        

11. Négrel AD, Maul E, Pokharel GP, Zhao J, Ellwein LB. Refractive error study in children: sampling and measurement methods for a multi-country survey. Am J Ophthalmol 2000;129:421-6. PMID:10764848 doi:10.1016/S0002-9394(99)00455-9        

12. Limburg H, Kumar R, Indrayan A, Sundaram KR. Rapid assessment of prevalence of cataract blindness at district level. Int J Epidemiol 1997; 26:1049-54. PMID:9363527 doi:10.1093/ije/26.5.1049        

13. Weale RA. Epidemiology of refractive errors and presbyopia. Surv Ophthalmol 2003;48:515-43. PMID:14499819 doi:10.1016/S0039-6257(03)00086-9        

14. Thulasiraj RD, Nirmalan PK, Ramakrishnan R, Krishnadas R, Manimekalai TK, Baburajan NP, et al. Blindness and vision impairment in a rural south Indian population: the Aravind Comprehensive Eye Survey. Ophthalmology 2003; 110:1491-8. PMID:12917162 doi:10.1016/S0161-6420(03)00565-7        

15. Prevention of Blindness and Deafness. Elimination of avoidable visual disability due to refractive errors. Report of an informal planning meeting. Geneva, 3–5 July 2000. Geneva: WHO; 2001 (WHO/PBL/00.79).        

16. Thylefors B. A global initiative for the elimination of avoidable blindness. Am J Ophthalmol 1998;125:90-3. PMID:9437318 doi:10.1016/S0002-9394(99)80239-6        

17. Dandona R, Dandona L. Refractive error blindness. Bull World Health Organ 2001;79:237-43. PMID:11285669        

18. Preslan MW, Novak A. Baltimore vision screening project. Phase 2. Ophthalmology 1998;105:150-3. PMID:9442791 doi:10.1016/S0161-6420(98)91813-9        

19. Khandekar R, Mohammed AJ, Al Raisi A. The compliance of spectacle wear and its determinants among school children of Dhakhiliya region of the Sultanate of Oman. A descriptive study. Sultan Qaboos University Journal for Scientific Research & Medical Sciences 2002;4:39-42.        

20. Vitale S, Cotch MF, Sperduto RD. Prevalence of visual impairment in the United States. JAMA 2006;295:2158-63. PMID:16684986 doi:10.1001/jama.295.18.2158        

21. Bourne RRA, Dineen BP, Ali SM, Noorul Huq DM, Johnson GJ. Prevalence of refractive error in Bangladeshi adults. Results of the National Blindness and Low Vision Survey of Bangladesh. Ophthalmology 2004;111:1150-60. PMID:15177965 doi:10.1016/j.ophtha.2003.09.046        

22. Murthy GVS, Gupta S, Tewari HK, Jose R, Bachani D. National survey on blindness and visual outcomes after cataract surgery. 2001–2002. Report. New Delhi: National Programme for Control of Blindness, Directorate General of Health Services, Ministry of Health and Family Welfare, Government of India; 2002.        

(Submitted: 11 February 2007 – Revised version received: 28 May 2007 – Accepted: 11 June 2007 – Published online: 16 November 2007)