Blindness and visual impairment due to uncorrected refractive error in sub-Saharan Africa: review of recent population-based studies
- 1Department of Public Health and Primary Care, University of Cambridge, Cambridge, UK
- 2Centre for Eye Research Australia, Royal Victorian Eye and Ear Hospital, University of Melbourne, Melbourne, Victoria, Australia
- 3Kilimanjaro Centre for Community Ophthalmology, Moshi, Tanzania
- Correspondence to Dr Paul Courtright, Kilimanjaro Centre for Community Ophthalmology, Good Samaritan Foundation, PO Box 2254, Moshi 1, Tanzania;
Contributors JS extracted the data and also wrote the first draft of the manuscript. SL designed the study and supervised the analysis. PC also designed the study and supervised the analysis, and is the guarantor of the article. All authors were responsible for revising and approving subsequent drafts of the article prior to submission.
- Accepted 18 December 2011
- Published Online First 8 February 2012
Aim The authors aimed to review published data on uncorrected refractive error (URE) as a cause of blindness and visual impairment in adults aged ≥40 years in sub-Saharan Africa (SSA).
Methods Data were extracted from population-based prevalence surveys measuring presenting visual acuity (PVA). Results from 11 surveys performed in 10 countries in SSA, encompassing 39 458 people aged ≥40 years and older, are presented.
Results The prevalence of blindness (PVA<3/60 in better eye) ranged from 1.1% in an urban district of Cameroon to 7.9% in a rural district in Ethiopia. More than half of studies (6/11) reported no blindness due to URE. The proportion of moderate visual impairment (PVA ≤6/60 and >6/18) due to URE ranged from 12.3% to 57.1%. Excluding two studies that included uncorrected aphakia as part of URE, the highest proportion of blindness and severe visual impairment due to uncorrected aphakia was found in Gambia (15.2%) and Nigeria (15.8%), respectively.
Conclusion Although URE is a leading cause of visual impairment, it does not represent a major cause of blindness in SSA.
Uncorrected refractive error (URE) constitutes a public health problem worldwide and is included as one of the priority diseases in the ‘VISION 2020: The Right to Sight’ initiative.1 URE has gained increasing attention in recent years as an important cause of avoidable blindness and visual impairment (VI).2 URE is associated with reduced quality of life, as well as loss of economic productivity and educational and employment opportunities.3 4
In population-based blindness surveys, estimation of URE relies upon the assessment of presenting visual acuity (PVA), which records one's visual acuity (VA) with usual correction only. In 2002, WHO estimated that there were 37 million blind and 161 million with VI.5 However, this estimate did not include the contribution to the global burden of blindness and VI due to UREs. Subsequently, WHO estimated an additional 153 million people were visually impaired (PVA <6/18) due to URE, of whom 8 million were blind, thus making URE the leading cause of VI worldwide and the second most prevalent cause of blindness.6
Specifically, for sub-Saharan Africa (SSA; Afr-D and Afr-E combined), it was estimated that in people aged ≥50 years the prevalence of blindness due to URE in SSA was 1.64% and the prevalence of VI was 5.94%.6 However, these estimates were based upon two unpublished studies (Mali and Mauritania) for Afr-D and one published study7 in children aged 5–15 years in South Africa (Afr-E). In light of the recent increase in population-based studies in adults in SSA we aimed to reassess the prevalence of blindness and VI due to URE in SSA.
Our literature search was conducted for the years 1966–April 2011 using Medline. Keywords used included: uncorrected refractive error, undercorrected refractive error, correctable visual impairment, unmet, preventable, avoidable, refractive errors (MeSH), aphakia, blindness, visual impairment, presenting visual acuity, presenting vision, prevalence and population. Studies were selected for inclusion if they were population-based with a sample size >1000, reported PVA <6/18 with its causes, had a high participation rate >85% and provided the standard WHO categories of VA. We also searched reference lists of studies meeting inclusion criteria. Only published studies were included.
Unless mentioned elsewhere, blindness was defined as PVA <3/60 in the better eye. Moderate VI was defined as PVA <6/18 but ≥6/60. Severe visual impairment (SVI) was defined as PVA <6/60 but ≥3/60. URE was defined as PVA <6/18 that improved to ≥6/18 with pinhole or refraction. Presbyopia was not considered in the definition of URE. Unless otherwise specified, the estimates of URE do not include uncorrected aphakia.
All-cause prevalence (and 95% CIs) of blindness, moderate VI and SVI was extracted from each study, as well as the proportion of blindness, moderate VI and SVI due to URE; then, the prevalence of blindness, SVI and moderate VI due to URE was calculated from these estimates.
We found 13 published studies that met the inclusion criteria. Aside from one study among children aged 5–15 years,7 and a study investigating all ages,8 the remainder investigated adults only. Each adult-only study had a minimum age ≥40 years, with five studies ≥50 years only. All studies examining adults limited to ≥50 years used the Rapid Assessment of Avoidable Blindness methodology.9
The 11 studies included covered 10 separate countries and are described in table 1. Three of the studies were national surveys.9–11 Among the included studies, there were 39 458 people aged ≥40 years. The survey populations ranged from districts of approximately 1 million inhabitants to a national survey of Nigeria, the most populous country in SSA. Two studies, both from Cameroon, included uncorrected aphakia in their definition of URE.12 13 The prevalence of blindness (PVA<3/60) ranged from 1.1% in an urban district in Cameroon to 7.9% in a rural district in Ethiopia (table 2).
More than half (6/11) the studies reported no blindness due to URE, with the highest prevalence estimate 0.28% (accounting for 3.5% of blindness) in Gurage Zone, Ethiopia (table 3).14 Most studies (9/11) reported some cases of SVI due to URE, with the highest estimate 0.48% (14.6% of SVI) in Eritrea.11 The proportion of moderate VI due to URE was markedly higher in all studies although there was considerable heterogeneity among studies. In a national blindness survey across all ages in Ethiopia, the proportion of blindness and low vision (PVA <6/18 but ≥3/60 in the better eye) due to URE was 7.8% and 33.4%, respectively. The survey of children conducted in South Africa reported the prevalence of blindness due to URE to be zero.7
Use of intraocular lenses (IOLs) in cataract operations is highly variable (table 4). In one district of Ethiopia, surveyed over 10 years ago, IOLs were not used, whereas a more recent survey in south Malawi reported that >90% of people in the population had IOL surgery. The highest proportion of blindness and SVI due to uncorrected aphakia was found in The Gambia (15.2%) and Nigeria (15.8%), respectively.
We found no published studies in SSA in which the prevalence of blindness due to URE in the population aged ≥50 years was as high as WHO estimate of 1.64%.6 The highest prevalence was 0.16% (95% CI 0.13 to 0.18) in Eritrea11 and many surveys found no one blind from URE. Similarly, the prevalence of SVI and VI due to URE in the published studies, ranging from 0 to 0.48% and from 0.66 to 5.71% respectively, is also lower than WHO estimate of 5.95% for the two together. It is likely that the difference is due to WHO reliance on only two studies in SSA of people over age 50 years.
No published studies were identified from Southern Africa and Francophone countries of West and Central Africa; this is a limitation of our data, and an important future research priority. However, the surveys available reflect a wide array of geographical heterogeneity, and there is good representation from both urban and rural populations. Although URE affects all age groups, our main findings are from studies of adults aged ≥40 years. The burden of refractive error and URE may be higher in other age groups, especially children. WHO advocates that services for refractive error should be targeted at children, the poor and adults aged ≥50 years.1
There are limitations in this analysis. The countries included in this review may not represent all of SSA in terms of URE. As of December 2006, all of the countries corresponding to the studies in this review had signed the VISION 2020 Global Declaration, formed a VISION 2020 national committee or prevention of blindness committee, participated in a VISION 2020 workshop and drafted a VISION 2020 national plan.1 Service provision for the management of refractive error could be higher in these countries than ones in which none of the objectives had been met. Nonetheless, in Nigeria, the spectacle coverage was exceptionally low (4.4% at the VA 6/18 level).15 Correction for refractive error should be affordable, of good quality and culturally appropriate.1
Another limitation is the reliance on pinholes in most of the surveys to diagnose refractive error. Pinholes should be between 1.0 mm and 1.5 mm in diameter to diagnose refractive error.16 If the pinhole is too large, it will not effectively correct refractive error, whereas a pinhole that is too small will markedly increase diffraction and decrease the amount of light entering the eye.
This analysis does not include causes of mild VI (PVA<6/12–6/18), which was most frequently (77.9%) due to URE in Nigeria.15 We also did not assess presbyopia, a common reason for requiring spectacles. For Africa, it is estimated that there may be 118.5 million people with presbyopia, the majority of whom have inadequate optical correction.17
The inclusion of aphakia with URE needs consideration. On the one hand, aphakia may be corrected by spectacles. On the other hand, the real solution for aphakia is to use IOLs with biometry in cataract surgery. Both studies from Cameroon12 13 considered refractive error to be ‘improvement of unaided VA 6/18 to VA>6/18 using pinhole or +10 dioptre glasses for aphakia’. In the Rapid Assessment of Avoidable Blindness, methodology, uncorrected aphakia is considered as a separate cause of blindness or VI. Increasing use of implantable IOLs in cataract surgery in many areas of Africa reduces the prevalence of aphakia, though borderline outcome following IOL surgery without biometry may be common. The high prevalence of uncorrected aphakia in Nigeria may have been due to couching, an antediluvian treatment of cataract, which is still commonly practiced in Nigeria and other areas of SSA. Couching in Nigeria is associated with poor visual outcomes, and adherence to aphakic correction is poor; even with aphakic correction 42.6% of patients surveyed with aphakia were blind.18 Moreover, in the same study, the most common cause of a poor outcome following cataract surgery was URE in eyes with IOLs, and uncorrected aphakia in non-IOL eyes.19
There are few population-based data on URE in children and young adults in SSA. The Refractive Error Study in Children, a multi-country study reported that refractive error was the most common cause of reduced vision (63.6% of cases), but most children (97.3%) had a PVA>20/32 (6/9.5 in metre scale) in both eyes.7 Refractive error prevalence in populations may change over time in response to putative environmental effects;20 it is possible that this might happen in the future in SSA resulting in an increased need for refractive error services.
In conclusion, available evidence from a number of surveys in SSA indicates that URE is a major cause of VI but not of blindness. Although this review only includes data on URE, the heterogeneity in this should serve as a reminder that the prevalence of refractive error itself may vary widely due to genetic and environmental factors and that SSA is a diverse continent.
Competing interests None.
Provenance and peer review Not commissioned; internally peer reviewed.